{ "cells": [ { "cell_type": "markdown", "id": "c698cabe", "metadata": { "papermill": { "duration": 0.002335, "end_time": "2026-08-10T18:23:37.732880+00:00", "exception": false, "start_time": "2026-08-10T18:23:37.730545+00:00", "status": "completed" } }, "source": [ "# ETFs: Feature Engineering\n", "\n", "Every model here sees the world through this matrix, so what is computable now bounds what\n", "can be learned later. Each feature answers one question: at the moment a position is decided,\n", "which observations are already on the tape, and what does the feature make of them?\n", "\n", "## Learning objectives\n", "\n", "- State a feature's timing contract - its lookback and its information lag - before writing\n", " the code that computes it\n", "- Compute trailing and cross-sectional statistics that read no observation dated after the\n", " decision timestamp, and declare a lag where an input is not yet available at it\n", "- Show that withholding later dates leaves every feature value unchanged, which is what\n", " separates a trailing statistic from one fitted over the whole sample\n", "- Read a feature set for scale, dispersion, redundancy and decay before any model sees it\n", "\n", "## Book reference, prerequisites and artifacts\n", "\n", "Chapter 8, Sections 8.1-8.6. Reads split- and dividend-adjusted daily bars via `load_etfs()`,\n", "the Treasury constant-maturity series via `load_macro()`, the tradability gate\n", "`eligibility.csv` from [`01_feasibility_analysis`](01_feasibility_analysis.ipynb), and\n", "`config/setup.yaml`. Writes `features/financial.parquet` with a `.digest.json` sidecar, read by\n", "[`05_evaluation`](05_evaluation.ipynb), which tests fold by fold whether any of it predicts.\n", "[`04_model_based_features`](04_model_based_features.ipynb) builds a second matrix from the same\n", "prices, of features that are themselves model outputs, and `05_evaluation` evaluates the two\n", "together." ] }, { "cell_type": "code", "execution_count": 1, "id": "4eff3c7f", "metadata": { "execution": { "iopub.execute_input": "2026-08-10T18:23:37.738322Z", "iopub.status.busy": "2026-08-10T18:23:37.738083Z", "iopub.status.idle": "2026-08-10T18:23:40.094119Z", "shell.execute_reply": "2026-08-10T18:23:40.093463Z" }, "papermill": { "duration": 2.359673, "end_time": "2026-08-10T18:23:40.094763+00:00", "exception": false, "start_time": "2026-08-10T18:23:37.735090+00:00", "status": "completed" } }, "outputs": [], "source": [ "\"\"\"ETFs: Feature Engineering.\"\"\"\n", "\n", "import logging\n", "import warnings\n", "from datetime import date\n", "\n", "import polars as pl\n", "import yaml\n", "from ml4t.engineer.features.momentum import adx, aroon, cci, macd, rsi, stochastic\n", "from ml4t.engineer.features.regime import choppiness_index, hurst_exponent\n", "from ml4t.engineer.features.trend import ema, sma\n", "from ml4t.engineer.features.volatility import bollinger_bands, natr\n", "from ml4t.engineer.features.volume import obv\n", "\n", "from case_studies.utils.artifact_digest import value_digest, write_artifact\n", "from case_studies.utils.feature_engineering import (\n", " EPS,\n", " assert_values_agree,\n", " assign_families,\n", " clip_within_date,\n", " cross_sectional_percentile,\n", " drawdown_block,\n", " families_from_config,\n", " family_coverage,\n", " momentum_volatility_block,\n", " plot_coverage_through_time,\n", " plot_cross_sectional_dispersion,\n", " plot_feature_distributions,\n", " plot_persistence,\n", " plot_redundancy_clusters,\n", " plot_timing_contract,\n", " register_frame,\n", " rolling_zscore,\n", " trailing_volume_ratio,\n", " warmup_audit,\n", ")\n", "from data import load_etfs, load_macro\n", "from utils.artifact_specs import resolve_label_horizon\n", "from utils.paths import display_path, get_case_study_dir\n", "\n", "warnings.filterwarnings(\"ignore\")\n", "logging.disable(logging.INFO)\n", "\n", "CASE_DIR = get_case_study_dir(\"etfs\")\n", "FEATURES_DIR = CASE_DIR / \"features\"" ] }, { "cell_type": "markdown", "id": "e782020d", "metadata": { "papermill": { "duration": 0.001849, "end_time": "2026-08-10T18:23:40.099927+00:00", "exception": false, "start_time": "2026-08-10T18:23:40.098078+00:00", "status": "completed" } }, "source": [ "The start of the sample is a parameter so the matrix can be rebuilt over a shorter window\n", "without editing the notebook; left unset it reads the whole history." ] }, { "cell_type": "code", "execution_count": 2, "id": "218fc763", "metadata": { "execution": { "iopub.execute_input": "2026-08-10T18:23:40.104351Z", "iopub.status.busy": "2026-08-10T18:23:40.104283Z", "iopub.status.idle": "2026-08-10T18:23:40.106283Z", "shell.execute_reply": "2026-08-10T18:23:40.105858Z" }, "papermill": { "duration": 0.004763, "end_time": "2026-08-10T18:23:40.106543+00:00", "exception": false, "start_time": "2026-08-10T18:23:40.101780+00:00", "status": "completed" }, "tags": [ "parameters" ] }, "outputs": [], "source": [ "START_DATE = None" ] }, { "cell_type": "markdown", "id": "9414f1e1", "metadata": { "papermill": { "duration": 0.001796, "end_time": "2026-08-10T18:23:40.110311+00:00", "exception": false, "start_time": "2026-08-10T18:23:40.108515+00:00", "status": "completed" } }, "source": [ "## Configuration\n", "\n", "Every window, the ranked feature list, the regime threshold, the decision horizon and the\n", "holdout boundary are declared in `config/setup.yaml` and bound here. A value the notebook\n", "invents is a second source of truth for a decision the rest of the pipeline reads from one\n", "place. The horizon fixes how far the persistence figure has to look, because a feature has to\n", "hold its ordering for at least one decision cycle to be usable at that cadence." ] }, { "cell_type": "code", "execution_count": 3, "id": "1cf20002", "metadata": { "execution": { "iopub.execute_input": "2026-08-10T18:23:40.114542Z", "iopub.status.busy": "2026-08-10T18:23:40.114495Z", "iopub.status.idle": "2026-08-10T18:23:40.130083Z", "shell.execute_reply": "2026-08-10T18:23:40.129594Z" }, "papermill": { "duration": 0.018324, "end_time": "2026-08-10T18:23:40.130394+00:00", "exception": false, "start_time": "2026-08-10T18:23:40.112070+00:00", "status": "completed" } }, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "10 feature families are declared, each with its own lookback and lag\n", "Positions are re-decided every 21 sessions, so a feature has to hold its ordering that long to be usable at this cadence\n", "Dates from 2024-01-01 onward are the holdout; D.3 rebuilds the matrix without them and compares the two builds value by value\n" ] } ], "source": [ "setup = yaml.safe_load((CASE_DIR / \"config\" / \"setup.yaml\").read_text())\n", "FAMILIES = families_from_config(setup)\n", "WINDOWS = setup[\"features\"][\"windows\"]\n", "OSCILLATORS = setup[\"features\"][\"oscillators\"]\n", "STATE = setup[\"features\"][\"state\"]\n", "RANKED = setup[\"features\"][\"ranked\"]\n", "REGIME_THRESHOLD = setup[\"features\"][\"regime_threshold\"]\n", "MACRO_LAG_DAYS = setup[\"modeling\"][\"latent_factors\"][\"macro_context\"][\"availability_lag_days\"]\n", "DECISION_CYCLE = int(resolve_label_horizon(\"etfs\", setup[\"labels\"][\"primary\"], setup).rstrip(\"Dd\"))\n", "HOLDOUT_START = date.fromisoformat(setup[\"evaluation\"][\"holdout_start\"])\n", "\n", "print(f\"{len(FAMILIES)} feature families are declared, each with its own lookback and lag\")\n", "print(\n", " f\"Positions are re-decided every {DECISION_CYCLE} sessions, so a feature has to hold its \"\n", " f\"ordering that long to be usable at this cadence\"\n", ")\n", "print(\n", " f\"Dates from {HOLDOUT_START} onward are the holdout; D.3 rebuilds the matrix without them \"\n", " f\"and compares the two builds value by value\"\n", ")" ] }, { "cell_type": "markdown", "id": "53472863", "metadata": { "papermill": { "duration": 0.001941, "end_time": "2026-08-10T18:23:40.134393+00:00", "exception": false, "start_time": "2026-08-10T18:23:40.132452+00:00", "status": "completed" } }, "source": [ "## A. What the thesis says should carry information\n", "\n", "The hypothesis is cross-sectional and it is about persistence: among liquid ETFs spanning\n", "equities, bonds, commodities and currencies, the ones that have gained relative to the rest\n", "continue to over the following month. Three things follow.\n", "\n", "What the ranking is formed on is trailing relative performance, so the matrix holds returns\n", "and risk-adjusted returns at eight horizons rather than one - which horizon the effect lives\n", "at is an empirical question, not a modelling assumption. The skip-recent construction drops the\n", "most recent month from the six- and twelve-month windows, because that month reverses where\n", "the ones before it continue.\n", "\n", "The **conditioning** is regime: a cross-asset rotation only works while the assets disagree,\n", "so the matrix also carries volatility, drawdown, trend strength, the equity-bond correlation\n", "and the shape of the yield curve. These rank nothing against each other; they say which\n", "environment a ranking is being formed in, which is what the register's `role` column records\n", "and what no assertion can recover from the values.\n", "\n", "The **representation** matters as much as the quantity. A raw return has a distribution that\n", "drifts with the volatility of the period, so the three features `config/setup.yaml` lists as\n", "ranked - the six-month return, its risk-adjusted version and the three-month volatility - are\n", "carried as cross-sectional percentiles as well as levels, comparable across dates by\n", "construction.\n", "\n", "The register is declared in `config/setup.yaml`, one row per family: what it reads, how far\n", "back, and with what delay. A `lag` of zero means the input is on the tape at the decision\n", "itself, which every price-derived family here is: the decision is taken at the close and\n", "executes at the next open. The yield curve is the one exception, at one session: a Treasury\n", "series dated t is treated as available from the close of t+1, which is what\n", "`config/setup.yaml` declares as the case study's macro policy. That lag is why it is a family\n", "of its own rather than sharing one with the cross-asset correlation, which reads prices." ] }, { "cell_type": "code", "execution_count": 4, "id": "e644e5ff", "metadata": { "execution": { "iopub.execute_input": "2026-08-10T18:23:40.138744Z", "iopub.status.busy": "2026-08-10T18:23:40.138687Z", "iopub.status.idle": "2026-08-10T18:23:40.166692Z", "shell.execute_reply": "2026-08-10T18:23:40.166236Z" }, "papermill": { "duration": 0.030676, "end_time": "2026-08-10T18:23:40.166946+00:00", "exception": false, "start_time": "2026-08-10T18:23:40.136270+00:00", "status": "completed" } }, "outputs": [ { "data": { "text/html": [ "
\n", "shape: (10, 6)
familyroleinputslookback (bars)lag (bars)frame
strstrstri64i64str
"momentum""signal""adjusted close"2520"time series"
"risk-adjusted momentum""signal""log return"2520"time series"
"volatility""state""log return"2520"time series"
"oscillator and trend""signal""OHLC"2000"time series"
"range and drawdown""state""OHLC"1260"time series"
"volume""state""share volume"630"time series"
"extremes and consistency""signal""adjusted close"2520"time series"
"cross-sectional position""signal""ret_126d, sharpe_126d, vol_63d"1260"cross-section"
"cross-asset regime""state""SPY and TLT log returns"630"cross-asset"
"yield curve""state""10y and 2y constant-maturity y…2521"macro"
" ], "text/plain": [ "shape: (10, 6)\n", "┌─────────────────────┬────────┬────────────────────┬─────────────────┬────────────┬───────────────┐\n", "│ family ┆ role ┆ inputs ┆ lookback (bars) ┆ lag (bars) ┆ frame │\n", "│ --- ┆ --- ┆ --- ┆ --- ┆ --- ┆ --- │\n", "│ str ┆ str ┆ str ┆ i64 ┆ i64 ┆ str │\n", "╞═════════════════════╪════════╪════════════════════╪═════════════════╪════════════╪═══════════════╡\n", "│ momentum ┆ signal ┆ adjusted close ┆ 252 ┆ 0 ┆ time series │\n", "│ risk-adjusted ┆ signal ┆ log return ┆ 252 ┆ 0 ┆ time series │\n", "│ momentum ┆ ┆ ┆ ┆ ┆ │\n", "│ volatility ┆ state ┆ log return ┆ 252 ┆ 0 ┆ time series │\n", "│ oscillator and ┆ signal ┆ OHLC ┆ 200 ┆ 0 ┆ time series │\n", "│ trend ┆ ┆ ┆ ┆ ┆ │\n", "│ range and drawdown ┆ state ┆ OHLC ┆ 126 ┆ 0 ┆ time series │\n", "│ volume ┆ state ┆ share volume ┆ 63 ┆ 0 ┆ time series │\n", "│ extremes and ┆ signal ┆ adjusted close ┆ 252 ┆ 0 ┆ time series │\n", "│ consistency ┆ ┆ ┆ ┆ ┆ │\n", "│ cross-sectional ┆ signal ┆ ret_126d, ┆ 126 ┆ 0 ┆ cross-section │\n", "│ position ┆ ┆ sharpe_126d, ┆ ┆ ┆ │\n", "│ ┆ ┆ vol_63d ┆ ┆ ┆ │\n", "│ cross-asset regime ┆ state ┆ SPY and TLT log ┆ 63 ┆ 0 ┆ cross-asset │\n", "│ ┆ ┆ returns ┆ ┆ ┆ │\n", "│ yield curve ┆ state ┆ 10y and 2y ┆ 252 ┆ 1 ┆ macro │\n", "│ ┆ ┆ constant-maturity ┆ ┆ ┆ │\n", "│ ┆ ┆ y… ┆ ┆ ┆ │\n", "└─────────────────────┴────────┴────────────────────┴─────────────────┴────────────┴───────────────┘" ] }, "execution_count": 4, "metadata": {}, "output_type": "execute_result" } ], "source": [ "register_frame(FAMILIES).select(\n", " [\"family\", \"role\", \"inputs\", \"lookback (bars)\", \"lag (bars)\", \"frame\"]\n", ")" ] }, { "cell_type": "markdown", "id": "de64b0cd", "metadata": { "papermill": { "duration": 0.001964, "end_time": "2026-08-10T18:23:40.171078+00:00", "exception": false, "start_time": "2026-08-10T18:23:40.169114+00:00", "status": "completed" } }, "source": [ "## B. Inputs and their observability\n", "\n", "Three inputs. Daily bars are split- and dividend-adjusted, so a trailing return spans a\n", "corporate action without a jump. The tradability gate is a table of fund-years written by\n", "[`01_feasibility_analysis`](01_feasibility_analysis.ipynb), which admits a fund to a year on\n", "the turnover it averaged in the year before, so membership is decided by what was already\n", "known and a fund listed in 2019 cannot appear in a 2011 ranking. The 10-year minus 2-year\n", "Treasury spread is the third, and the only input not on the tape at the decision; C.4 says\n", "what is done about that." ] }, { "cell_type": "code", "execution_count": 5, "id": "9f343662", "metadata": { "execution": { "iopub.execute_input": "2026-08-10T18:23:40.176204Z", "iopub.status.busy": "2026-08-10T18:23:40.176131Z", "iopub.status.idle": "2026-08-10T18:23:40.224044Z", "shell.execute_reply": "2026-08-10T18:23:40.223591Z" }, "papermill": { "duration": 0.051453, "end_time": "2026-08-10T18:23:40.224618+00:00", "exception": false, "start_time": "2026-08-10T18:23:40.173165+00:00", "status": "completed" } }, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "470,662 daily bars covering 100 of the 100 declared funds, 2006-01-03 to 2025-12-31\n", "The tradability gate admits 1,704 fund-years\n", "The Treasury spread carries 9,495 rows, one per calendar day from 2000-01-03 to 2025-12-31, repeating the last published value on weekends and holidays\n" ] } ], "source": [ "prices = (\n", " load_etfs()\n", " .select([\"symbol\", \"timestamp\", \"open\", \"high\", \"low\", \"close\", \"volume\"])\n", " .sort([\"symbol\", \"timestamp\"])\n", ")\n", "if START_DATE is not None:\n", " prices = prices.filter(pl.col(\"timestamp\") >= pl.lit(START_DATE).str.to_date())\n", "\n", "eligibility = pl.read_csv(CASE_DIR / \"eligibility.csv\")\n", "yield_curve = (\n", " load_macro()\n", " .select(\"timestamp\", ((pl.col(\"dgs10\") - pl.col(\"dgs2\")) / 100).alias(\"slope\"))\n", " .drop_nulls()\n", " .sort(\"timestamp\")\n", ")\n", "print(\n", " f\"{len(prices):,} daily bars covering {prices['symbol'].n_unique()} of the \"\n", " f\"{len(setup['universe']['assets'])} declared funds, {prices['timestamp'].min()} to \"\n", " f\"{prices['timestamp'].max()}\"\n", ")\n", "print(f\"The tradability gate admits {len(eligibility):,} fund-years\")\n", "print(\n", " f\"The Treasury spread carries {len(yield_curve):,} rows, one per calendar day from \"\n", " f\"{yield_curve['timestamp'].min()} to {yield_curve['timestamp'].max()}, repeating the last \"\n", " f\"published value on weekends and holidays\"\n", ")" ] }, { "cell_type": "markdown", "id": "54e1ec91", "metadata": { "papermill": { "duration": 0.002151, "end_time": "2026-08-10T18:23:40.230450+00:00", "exception": false, "start_time": "2026-08-10T18:23:40.228299+00:00", "status": "completed" } }, "source": [ "The longest window below runs over 252 sessions, roughly one trading year, so how much history\n", "a fund brings decides when it can carry a feature at all. Grouping the funds by the year of\n", "their first bar says where that constraint bites. What to read off the table is the split\n", "between the block present from the beginning, which fills every window inside the first year\n", "and never falls out again, and the funds that arrive later, a handful at a time, each of which\n", "spends its own first year producing nulls in the long-window families. That thin tail is what\n", "F1's coverage dips are, what Section E's null policy removes, and why the matrix starts a year\n", "after the price history does." ] }, { "cell_type": "code", "execution_count": 6, "id": "fc1129d8", "metadata": { "execution": { "iopub.execute_input": "2026-08-10T18:23:40.235822Z", "iopub.status.busy": "2026-08-10T18:23:40.235714Z", "iopub.status.idle": "2026-08-10T18:23:40.252632Z", "shell.execute_reply": "2026-08-10T18:23:40.252148Z" }, "papermill": { "duration": 0.020706, "end_time": "2026-08-10T18:23:40.253328+00:00", "exception": false, "start_time": "2026-08-10T18:23:40.232622+00:00", "status": "completed" } }, "outputs": [ { "data": { "text/html": [ "
\n", "shape: (10, 4)
first tradedfundsfewest sessionsmedian sessions
i32u32u32i32
20067148935031
20071145374708
2008344684468
2009140514051
2010140204020
2011335703570
2012433153399
2013431343181
2015125732573
2018118951895
" ], "text/plain": [ "shape: (10, 4)\n", "┌──────────────┬───────┬─────────────────┬─────────────────┐\n", "│ first traded ┆ funds ┆ fewest sessions ┆ median sessions │\n", "│ --- ┆ --- ┆ --- ┆ --- │\n", "│ i32 ┆ u32 ┆ u32 ┆ i32 │\n", "╞══════════════╪═══════╪═════════════════╪═════════════════╡\n", "│ 2006 ┆ 71 ┆ 4893 ┆ 5031 │\n", "│ 2007 ┆ 11 ┆ 4537 ┆ 4708 │\n", "│ 2008 ┆ 3 ┆ 4468 ┆ 4468 │\n", "│ 2009 ┆ 1 ┆ 4051 ┆ 4051 │\n", "│ 2010 ┆ 1 ┆ 4020 ┆ 4020 │\n", "│ 2011 ┆ 3 ┆ 3570 ┆ 3570 │\n", "│ 2012 ┆ 4 ┆ 3315 ┆ 3399 │\n", "│ 2013 ┆ 4 ┆ 3134 ┆ 3181 │\n", "│ 2015 ┆ 1 ┆ 2573 ┆ 2573 │\n", "│ 2018 ┆ 1 ┆ 1895 ┆ 1895 │\n", "└──────────────┴───────┴─────────────────┴─────────────────┘" ] }, "execution_count": 6, "metadata": {}, "output_type": "execute_result" } ], "source": [ "prices.group_by(\"symbol\").agg(\n", " pl.col(\"timestamp\").min().alias(\"first\"), pl.len().alias(\"sessions\")\n", ").with_columns(pl.col(\"first\").dt.year().alias(\"first traded\")).group_by(\"first traded\").agg(\n", " pl.len().alias(\"funds\"),\n", " pl.col(\"sessions\").min().alias(\"fewest sessions\"),\n", " pl.col(\"sessions\").median().cast(pl.Int32).alias(\"median sessions\"),\n", ").sort(\"first traded\")" ] }, { "cell_type": "markdown", "id": "d0b0208c", "metadata": { "lines_to_next_cell": 2, "papermill": { "duration": 0.00238, "end_time": "2026-08-10T18:23:40.258152+00:00", "exception": false, "start_time": "2026-08-10T18:23:40.255772+00:00", "status": "completed" } }, "source": [ "## C. Feature construction, one subsection per family\n", "\n", "### C.1 Momentum, volatility and their differences\n", "\n", "The trailing return, volatility and risk-adjusted return block comes from\n", "`case_studies/utils`, shared with the other panel case studies, so `ret_126d` means one\n", "construction and one denominator guard wherever it appears. What stays here is what is\n", "specific to this universe: the skip-recent windows, the differences between horizons, and the\n", "ratio of a short volatility window to a long one. The skip-recent pair counts both its start and\n", "the stretch it drops in the month of sessions `config/setup.yaml` declares, so the month dropped\n", "and the window it is dropped from cannot drift apart." ] }, { "cell_type": "code", "execution_count": 7, "id": "c9c41ca7", "metadata": { "execution": { "iopub.execute_input": "2026-08-10T18:23:40.263479Z", "iopub.status.busy": "2026-08-10T18:23:40.263373Z", "iopub.status.idle": "2026-08-10T18:23:40.267242Z", "shell.execute_reply": "2026-08-10T18:23:40.266703Z" }, "lines_to_next_cell": 2, "papermill": { "duration": 0.007216, "end_time": "2026-08-10T18:23:40.267575+00:00", "exception": false, "start_time": "2026-08-10T18:23:40.260359+00:00", "status": "completed" } }, "outputs": [], "source": [ "def momentum_features(df: pl.DataFrame) -> pl.DataFrame:\n", " \"\"\"The shared trailing block, plus the differences this case study builds on it.\"\"\"\n", " df = momentum_volatility_block(\n", " df,\n", " entity=\"symbol\",\n", " return_windows=WINDOWS[\"momentum\"],\n", " volatility_windows=WINDOWS[\"volatility\"],\n", " )\n", " month = WINDOWS[\"skip_recent\"]\n", " held = pl.col(\"close\").shift(month).over(\"symbol\")\n", " return df.with_columns(\n", " (held / pl.col(\"close\").shift(12 * month).over(\"symbol\").clip(lower_bound=EPS) - 1).alias(\n", " \"skip_recent_12_1\"\n", " ),\n", " (held / pl.col(\"close\").shift(6 * month).over(\"symbol\").clip(lower_bound=EPS) - 1).alias(\n", " \"skip_recent_6_1\"\n", " ),\n", " (pl.col(\"ret_21d\") - pl.col(\"ret_63d\")).alias(\"mom_accel_short\"),\n", " (pl.col(\"ret_63d\") - pl.col(\"ret_126d\")).alias(\"mom_accel_medium\"),\n", " (pl.col(\"ret_126d\") - pl.col(\"ret_252d\")).alias(\"mom_accel_long\"),\n", " (pl.col(\"vol_21d\") / pl.col(\"vol_63d\").clip(lower_bound=EPS)).alias(\"vol_ratio_short\"),\n", " (pl.col(\"vol_63d\") / pl.col(\"vol_126d\").clip(lower_bound=EPS)).alias(\"vol_ratio_medium\"),\n", " )" ] }, { "cell_type": "markdown", "id": "3991f23c", "metadata": { "lines_to_next_cell": 2, "papermill": { "duration": 0.002128, "end_time": "2026-08-10T18:23:40.271860+00:00", "exception": false, "start_time": "2026-08-10T18:23:40.269732+00:00", "status": "completed" } }, "source": [ "### C.2 Oscillators, trend ratios and range\n", "\n", "Everything in this block reads one fund's price against its own recent path. The\n", "constructions fall into four groups, and a reader who has not met them needs to know what\n", "each measures before the matrix can be read at all.\n", "\n", "**Where the price sits inside its recent range.** The relative strength index weighs the\n", "sessions that closed up against those that closed down over its window and reports the result\n", "between 0 and 100, so 50 is a window with as much up movement as down. The stochastic reports\n", "on the same scale where the close sits between the window's lowest low and its highest high.\n", "The commodity channel index measures the same idea without bounds: how far the session's\n", "average of high, low and close sits from that average over the window, in units of the\n", "window's own mean deviation. Bollinger %B is the fourth, and the notebook assembles it from\n", "the bands rather than taking it ready-made: 0 is two standard deviations below the window's\n", "mean close, 1 is two above, and the value is where the close falls between them.\n", "\n", "**How strongly the price is trending.** The average directional index summarizes how much of\n", "the session-to-session movement went one way rather than back and forth, without saying which\n", "way, so it separates a trend from a range without taking a side. Aroon counts how recently\n", "the window's high and its low occurred; subtracting the two counts gives one number that is\n", "positive when the high is the more recent. The moving-average convergence-divergence line is\n", "the gap between a fast and a slow exponential average of the close, which widens as a move\n", "accelerates and crosses zero when it turns.\n", "\n", "**How far the price moved and how much ground it covered doing so.** The normalized average\n", "true range is the typical session's high-to-low span, counting any overnight gap, as a\n", "percentage of the price - normalizing by price is what makes it comparable across funds\n", "quoted at different levels. The choppiness index divides those spans summed session by session\n", "by the span of the whole window: a market that ends near where it started after covering a lot\n", "of ground reads high, one that travelled in a straight line reads low. The Hurst exponent asks\n", "whether the size of a move grows with the length of the interval faster or slower than\n", "independent increments would give. Above one half, a move tends to be followed by another in\n", "the same direction; below one half, by a reversal; one half is what a random walk gives. It is\n", "estimated by comparing the range of the cumulative move against its dispersion at many\n", "interval lengths at once, which is why it needs a window several times longer than the others\n", "here.\n", "\n", "**Where the price sits against its own averages.** The close divided by its 50- and\n", "200-session simple averages and by its 26-session exponential average. Above one, the fund is\n", "trading above that average. Dividing rather than subtracting is what makes the three\n", "comparable across funds, for the same reason the range is normalized by price.\n", "\n", "All of them come from `ml4t.engineer.features` rather than being written here. The smoothing\n", "convention inside an oscillator is where two implementations of one name diverge - Wilder's\n", "recursive average and a simple moving average of the same gains give different numbers - and\n", "a shared implementation is what keeps `rsi_14` meaning one thing across the nine case studies." ] }, { "cell_type": "code", "execution_count": 8, "id": "54cec43a", "metadata": { "execution": { "iopub.execute_input": "2026-08-10T18:23:40.276922Z", "iopub.status.busy": "2026-08-10T18:23:40.276756Z", "iopub.status.idle": "2026-08-10T18:23:40.281252Z", "shell.execute_reply": "2026-08-10T18:23:40.280790Z" }, "lines_to_next_cell": 2, "papermill": { "duration": 0.007684, "end_time": "2026-08-10T18:23:40.281584+00:00", "exception": false, "start_time": "2026-08-10T18:23:40.273900+00:00", "status": "completed" } }, "outputs": [], "source": [ "def oscillator_features(df: pl.DataFrame) -> pl.DataFrame:\n", " \"\"\"Bounded oscillators, moving-average ratios, normalized range and regime exponents.\"\"\"\n", " adx_p, stoch_p, aroon_p = OSCILLATORS[\"adx\"], OSCILLATORS[\"stochastic\"], OSCILLATORS[\"aroon\"]\n", " natr_p, chop_p, hurst_p = OSCILLATORS[\"natr\"], OSCILLATORS[\"choppiness\"], OSCILLATORS[\"hurst\"]\n", " ema_p, bb_p = OSCILLATORS[\"ema\"], OSCILLATORS[\"bollinger\"]\n", " df = df.with_columns(\n", " *[rsi(\"close\", period=p).over(\"symbol\").alias(f\"rsi_{p}\") for p in OSCILLATORS[\"rsi\"]],\n", " macd(\"close\", fast_period=OSCILLATORS[\"macd_fast\"], slow_period=OSCILLATORS[\"macd_slow\"])\n", " .over(\"symbol\")\n", " .alias(\"macd_line\"),\n", " adx(\"high\", \"low\", \"close\", period=adx_p).over(\"symbol\").alias(f\"adx_{adx_p}\"),\n", " *[\n", " cci(\"high\", \"low\", \"close\", period=p).over(\"symbol\").alias(f\"cci_{p}\")\n", " for p in OSCILLATORS[\"cci\"]\n", " ],\n", " stochastic(\"high\", \"low\", \"close\", fastk_period=stoch_p).over(\"symbol\").alias(\"stoch_k\"),\n", " aroon(\"high\", \"low\", timeperiod=aroon_p).over(\"symbol\").alias(\"_aroon\"),\n", " natr(\"high\", \"low\", \"close\", period=natr_p).over(\"symbol\").alias(f\"natr_{natr_p}\"),\n", " choppiness_index(\"high\", \"low\", \"close\", period=chop_p)\n", " .over(\"symbol\")\n", " .alias(f\"chop_{chop_p}\"),\n", " hurst_exponent(\"close\", period=hurst_p).over(\"symbol\").alias(f\"hurst_{hurst_p}\"),\n", " *[\n", " (pl.col(\"close\") / sma(\"close\", period=p).over(\"symbol\")).alias(f\"sma_ratio_{p}\")\n", " for p in OSCILLATORS[\"sma\"]\n", " ],\n", " (pl.col(\"close\") / ema(\"close\", period=ema_p).over(\"symbol\")).alias(f\"ema_ratio_{ema_p}\"),\n", " bollinger_bands(\"close\", period=bb_p).over(\"symbol\").alias(\"_bb\"),\n", " )\n", " band = pl.col(\"_bb\").struct.field(\"upper\") - pl.col(\"_bb\").struct.field(\"lower\")\n", " return df.with_columns(\n", " (pl.col(\"_aroon\").struct.field(\"up\") - pl.col(\"_aroon\").struct.field(\"down\")).alias(\n", " \"aroon_diff\"\n", " ),\n", " pl.when(band > 0)\n", " .then((pl.col(\"close\") - pl.col(\"_bb\").struct.field(\"lower\")) / band)\n", " .alias(f\"bb_pctb_{bb_p}\"),\n", " ).drop([\"_aroon\", \"_bb\"])" ] }, { "cell_type": "markdown", "id": "5a0a6b05", "metadata": { "lines_to_next_cell": 2, "papermill": { "duration": 0.002086, "end_time": "2026-08-10T18:23:40.285863+00:00", "exception": false, "start_time": "2026-08-10T18:23:40.283777+00:00", "status": "completed" } }, "source": [ "### C.3 Drawdown, volume and distance from extremes\n", "\n", "`max_dd_63d` is the share by which the price currently sits below its highest close of the\n", "trailing quarter, so it is zero at a new high and negative otherwise - the *current*\n", "drawdown, not the worst decline inside the window, which is a different statistic.\n", "\n", "On-balance volume is a running total that adds the session's share volume when the fund\n", "closed up and subtracts it when it closed down, so it rises while buying pressure leads and\n", "falls while selling pressure does. The level of that total says nothing on its own, because it\n", "depends on how long the fund has traded and how much it trades; what is carried is its\n", "distance from its own trailing mean in trailing standard deviations, which is comparable\n", "across funds and across time.\n", "\n", "The last two read the 52-week extremes directly. The close over its highest close of the past\n", "year is one at a new high and falls away below it; the close over its lowest is one at a new\n", "low and rises above it. The share of the past quarter's sessions that closed up sits beside\n", "them as the same question asked without reference to any extreme.\n", "\n", "Relative volume - the session's volume over its own trailing mean - is built with the same\n", "per-fund block, and C.7 says why the extreme values it produces are cut back within the date\n", "rather than here." ] }, { "cell_type": "code", "execution_count": 9, "id": "06baf78c", "metadata": { "execution": { "iopub.execute_input": "2026-08-10T18:23:40.292154Z", "iopub.status.busy": "2026-08-10T18:23:40.292071Z", "iopub.status.idle": "2026-08-10T18:23:40.294659Z", "shell.execute_reply": "2026-08-10T18:23:40.294291Z" }, "lines_to_next_cell": 2, "papermill": { "duration": 0.006942, "end_time": "2026-08-10T18:23:40.294923+00:00", "exception": false, "start_time": "2026-08-10T18:23:40.287981+00:00", "status": "completed" } }, "outputs": [], "source": [ "def drawdown_and_extremes(df: pl.DataFrame) -> pl.DataFrame:\n", " \"\"\"Drawdown, on-balance volume and position in the 52-week range.\"\"\"\n", " obv_w, share_w, range_w = STATE[\"obv_zscore\"], STATE[\"positive_share\"], STATE[\"extremes\"]\n", " df = drawdown_block(df, entity=\"symbol\", windows=WINDOWS[\"drawdown\"])\n", " df = df.with_columns(obv(\"close\", \"volume\").over(\"symbol\").alias(\"_obv\"))\n", " return df.with_columns(\n", " rolling_zscore(\"_obv\", obv_w, \"symbol\").alias(f\"obv_zscore_{obv_w}d\"),\n", " (pl.col(\"log_return\") > 0)\n", " .cast(pl.Float64)\n", " .rolling_mean(share_w)\n", " .over(\"symbol\")\n", " .alias(f\"pct_positive_{share_w}d\"),\n", " (\n", " pl.col(\"close\")\n", " / pl.col(\"close\").rolling_max(range_w).over(\"symbol\").clip(lower_bound=EPS)\n", " ).alias(\"dist_52w_high\"),\n", " (\n", " pl.col(\"close\")\n", " / pl.col(\"close\").rolling_min(range_w).over(\"symbol\").clip(lower_bound=EPS)\n", " ).alias(\"dist_52w_low\"),\n", " ).drop(\"_obv\")" ] }, { "cell_type": "markdown", "id": "a7c51b79", "metadata": { "lines_to_next_cell": 2, "papermill": { "duration": 0.002744, "end_time": "2026-08-10T18:23:40.299973+00:00", "exception": false, "start_time": "2026-08-10T18:23:40.297229+00:00", "status": "completed" } }, "source": [ "### C.4 Cross-asset state and macro state\n", "\n", "The SPY-TLT correlation and the curve slope are one number per date, shared by every ETF. A\n", "rolling window reads row order rather than the timestamp column, so the pair is re-sorted after\n", "the join that builds it, and the curve is joined backward in time so a market holiday carries\n", "the previous session's spread forward and never a later one's." ] }, { "cell_type": "code", "execution_count": 10, "id": "a5d7a618", "metadata": { "execution": { "iopub.execute_input": "2026-08-10T18:23:40.304746Z", "iopub.status.busy": "2026-08-10T18:23:40.304677Z", "iopub.status.idle": "2026-08-10T18:23:40.306995Z", "shell.execute_reply": "2026-08-10T18:23:40.306572Z" }, "lines_to_next_cell": 2, "papermill": { "duration": 0.005309, "end_time": "2026-08-10T18:23:40.307320+00:00", "exception": false, "start_time": "2026-08-10T18:23:40.302011+00:00", "status": "completed" } }, "outputs": [], "source": [ "def equity_bond_correlation(df: pl.DataFrame) -> pl.DataFrame:\n", " \"\"\"The trailing SPY-TLT return correlation, one row per date.\"\"\"\n", " corr_w = STATE[\"correlation\"]\n", " return (\n", " df.filter(pl.col(\"symbol\") == \"SPY\")\n", " .select(\"timestamp\", pl.col(\"log_return\").alias(\"_spy\"))\n", " .join(\n", " df.filter(pl.col(\"symbol\") == \"TLT\").select(\n", " \"timestamp\", pl.col(\"log_return\").alias(\"_tlt\")\n", " ),\n", " on=\"timestamp\",\n", " how=\"inner\",\n", " )\n", " .sort(\"timestamp\")\n", " .select(\n", " \"timestamp\",\n", " pl.rolling_corr(pl.col(\"_spy\"), pl.col(\"_tlt\"), window_size=corr_w).alias(\n", " f\"corr_spy_tlt_{corr_w}d\"\n", " ),\n", " )\n", " )" ] }, { "cell_type": "markdown", "id": "45828e7b", "metadata": { "lines_to_next_cell": 2, "papermill": { "duration": 0.002298, "end_time": "2026-08-10T18:23:40.311810+00:00", "exception": false, "start_time": "2026-08-10T18:23:40.309512+00:00", "status": "completed" } }, "source": [ "The Treasury spread is the one input that is not on the tape at the decision: the value\n", "describing a day is published after that day has ended, so a decision taken at its close could\n", "not have read it. `config/setup.yaml` declares the macro policy - `alfred_initial_release_close_lagged`\n", "- and with it how long a published value waits before a decision may use it, and each value is\n", "re-stamped with the date it becomes readable rather than the date it describes. F4 draws the\n", "resulting gap; it is the only family in the register that has one.\n", "\n", "The stamped series is then reduced to the trading calendar before the z-score's window runs.\n", "The register counts every family's lookback in trading sessions, and the Treasury frame carries\n", "a row for every calendar day, so a window run over it before that reduction would be counting\n", "weekends." ] }, { "cell_type": "code", "execution_count": 11, "id": "aa221dbc", "metadata": { "execution": { "iopub.execute_input": "2026-08-10T18:23:40.318035Z", "iopub.status.busy": "2026-08-10T18:23:40.317967Z", "iopub.status.idle": "2026-08-10T18:23:40.320420Z", "shell.execute_reply": "2026-08-10T18:23:40.320027Z" }, "lines_to_next_cell": 2, "papermill": { "duration": 0.006432, "end_time": "2026-08-10T18:23:40.320673+00:00", "exception": false, "start_time": "2026-08-10T18:23:40.314241+00:00", "status": "completed" } }, "outputs": [], "source": [ "def yield_curve_state(df: pl.DataFrame) -> pl.DataFrame:\n", " \"\"\"Curve level, regime flag and trailing z-score, on the session calendar.\"\"\"\n", " zscore_w = STATE[\"curve_zscore\"]\n", " available = yield_curve.select(\n", " pl.col(\"timestamp\").dt.offset_by(f\"{MACRO_LAG_DAYS}d\"), \"slope\"\n", " ).sort(\"timestamp\")\n", " return (\n", " df.select(\"timestamp\")\n", " .unique()\n", " .sort(\"timestamp\")\n", " .join_asof(available, on=\"timestamp\", strategy=\"backward\")\n", " .select(\n", " \"timestamp\",\n", " pl.when(pl.col(\"slope\") > REGIME_THRESHOLD).then(1).otherwise(0).alias(\"regime\"),\n", " pl.col(\"slope\").alias(\"yield_curve_slope\"),\n", " (\n", " (pl.col(\"slope\") - pl.col(\"slope\").rolling_mean(zscore_w))\n", " / pl.col(\"slope\").rolling_std(zscore_w).clip(lower_bound=EPS)\n", " ).alias(\"yield_curve_zscore\"),\n", " )\n", " )" ] }, { "cell_type": "markdown", "id": "e5135ed6", "metadata": { "lines_to_next_cell": 2, "papermill": { "duration": 0.00209, "end_time": "2026-08-10T18:23:40.325064+00:00", "exception": false, "start_time": "2026-08-10T18:23:40.322974+00:00", "status": "completed" } }, "source": [ "Both state frames carry one row per date, so a left join broadcasts each to every ETF trading\n", "that day without changing the panel's row count." ] }, { "cell_type": "code", "execution_count": 12, "id": "a619e60d", "metadata": { "execution": { "iopub.execute_input": "2026-08-10T18:23:40.330250Z", "iopub.status.busy": "2026-08-10T18:23:40.330160Z", "iopub.status.idle": "2026-08-10T18:23:40.332643Z", "shell.execute_reply": "2026-08-10T18:23:40.332197Z" }, "lines_to_next_cell": 2, "papermill": { "duration": 0.005693, "end_time": "2026-08-10T18:23:40.332948+00:00", "exception": false, "start_time": "2026-08-10T18:23:40.327255+00:00", "status": "completed" } }, "outputs": [], "source": [ "def regime_and_state(df: pl.DataFrame) -> pl.DataFrame:\n", " \"\"\"Equity-bond correlation and yield-curve state, broadcast to every row.\"\"\"\n", " return (\n", " df.join(equity_bond_correlation(df), on=\"timestamp\", how=\"left\")\n", " .join(yield_curve_state(df), on=\"timestamp\", how=\"left\")\n", " .sort([\"symbol\", \"timestamp\"])\n", " )" ] }, { "cell_type": "markdown", "id": "a9185654", "metadata": { "lines_to_next_cell": 2, "papermill": { "duration": 0.002052, "end_time": "2026-08-10T18:23:40.337159+00:00", "exception": false, "start_time": "2026-08-10T18:23:40.335107+00:00", "status": "completed" } }, "source": [ "### C.5 The tradability gate\n", "\n", "`01_feasibility_analysis` decided, one fund-year at a time, which ETFs traded at enough volume\n", "to be worth ranking. Applying that here is a semi-join on the row's own calendar year, which\n", "filters and cannot duplicate a row." ] }, { "cell_type": "code", "execution_count": 13, "id": "aa990368", "metadata": { "execution": { "iopub.execute_input": "2026-08-10T18:23:40.342180Z", "iopub.status.busy": "2026-08-10T18:23:40.342088Z", "iopub.status.idle": "2026-08-10T18:23:40.344898Z", "shell.execute_reply": "2026-08-10T18:23:40.344282Z" }, "lines_to_next_cell": 2, "papermill": { "duration": 0.006098, "end_time": "2026-08-10T18:23:40.345303+00:00", "exception": false, "start_time": "2026-08-10T18:23:40.339205+00:00", "status": "completed" } }, "outputs": [], "source": [ "def gate_to_eligible(df: pl.DataFrame) -> pl.DataFrame:\n", " \"\"\"Drop the rows the annual tradability gate excludes, before anything ranks them.\"\"\"\n", " return (\n", " df.with_columns(pl.col(\"timestamp\").dt.year().alias(\"_year\"))\n", " .join(\n", " eligibility.select(\"symbol\", pl.col(\"eligible_year\").alias(\"_year\")),\n", " on=[\"symbol\", \"_year\"],\n", " how=\"semi\",\n", " )\n", " .drop(\"_year\")\n", " )" ] }, { "cell_type": "markdown", "id": "6e923f17", "metadata": { "lines_to_next_cell": 2, "papermill": { "duration": 0.002086, "end_time": "2026-08-10T18:23:40.349716+00:00", "exception": false, "start_time": "2026-08-10T18:23:40.347630+00:00", "status": "completed" } }, "source": [ "### C.6 What is computable before the gate\n", "\n", "None of the four blocks above needs the eligible cross-section: three read one fund's own\n", "history, and the fourth reads two named funds and the Treasury series and broadcasts one value\n", "per date. So they all run on every bar the ETF traded, before the gate removes any. The\n", "relative-volume ratio runs in this group for the same reason:\n", "its trailing mean has to see every bar, or a fund admitted to the universe this year divides by\n", "the mean of its first few eligible days, and one that re-enters after a gap averages across the\n", "gap." ] }, { "cell_type": "code", "execution_count": 14, "id": "f8a0991f", "metadata": { "execution": { "iopub.execute_input": "2026-08-10T18:23:40.354706Z", "iopub.status.busy": "2026-08-10T18:23:40.354608Z", "iopub.status.idle": "2026-08-10T18:23:40.357386Z", "shell.execute_reply": "2026-08-10T18:23:40.356941Z" }, "lines_to_next_cell": 2, "papermill": { "duration": 0.006166, "end_time": "2026-08-10T18:23:40.357908+00:00", "exception": false, "start_time": "2026-08-10T18:23:40.351742+00:00", "status": "completed" } }, "outputs": [], "source": [ "def per_entity_features(df: pl.DataFrame) -> pl.DataFrame:\n", " \"\"\"Everything computable before the gate: per-fund history plus the broadcast state.\"\"\"\n", " return (\n", " df.pipe(momentum_features)\n", " .pipe(oscillator_features)\n", " .pipe(drawdown_and_extremes)\n", " .pipe(regime_and_state)\n", " .pipe(trailing_volume_ratio, entity=\"symbol\", windows=WINDOWS[\"volume\"])\n", " )" ] }, { "cell_type": "markdown", "id": "47cc61f0", "metadata": { "lines_to_next_cell": 2, "papermill": { "duration": 0.002129, "end_time": "2026-08-10T18:23:40.362322+00:00", "exception": false, "start_time": "2026-08-10T18:23:40.360193+00:00", "status": "completed" } }, "source": [ "### C.7 The two steps taken within a date\n", "\n", "A percentile and a clip are properties of the cross-section they are taken over, so both run\n", "after the gate: ranking an ETF against one the strategy cannot trade moves the number written\n", "for every ETF that it can. Which features are carried as their percentile within the date as\n", "well as their level is `config/setup.yaml`'s `ranked` list, so the same ones are ranked in\n", "every stage that reads the register.\n", "\n", "The clip cuts relative volume back to the 1st and 99th percentile of its own date. An index\n", "rebalance puts one fund's volume orders of magnitude above its own trailing mean for a single\n", "session, and one such row otherwise sets the scale every model sees for the whole column.\n", "Cutting within the date rather than over the column is what keeps the bound from being fitted\n", "on the sample, which is the property D.3 checks." ] }, { "cell_type": "code", "execution_count": 15, "id": "2a37d999", "metadata": { "execution": { "iopub.execute_input": "2026-08-10T18:23:40.367623Z", "iopub.status.busy": "2026-08-10T18:23:40.367531Z", "iopub.status.idle": "2026-08-10T18:23:40.370339Z", "shell.execute_reply": "2026-08-10T18:23:40.369897Z" }, "lines_to_next_cell": 2, "papermill": { "duration": 0.006125, "end_time": "2026-08-10T18:23:40.370676+00:00", "exception": false, "start_time": "2026-08-10T18:23:40.364551+00:00", "status": "completed" } }, "outputs": [], "source": [ "def clip_and_rank(df: pl.DataFrame) -> pl.DataFrame:\n", " \"\"\"The two within-date steps, over the eligible cross-section only.\"\"\"\n", " clipped = clip_within_date(\n", " df, columns=[f\"vol_ratio_{w}d\" for w in WINDOWS[\"volume\"]], time=\"timestamp\"\n", " )\n", " return clipped.with_columns(\n", " cross_sectional_percentile(col, \"timestamp\").alias(f\"{col}_rank\") for col in RANKED\n", " )" ] }, { "cell_type": "markdown", "id": "af0e4978", "metadata": { "lines_to_next_cell": 2, "papermill": { "duration": 0.002213, "end_time": "2026-08-10T18:23:40.375015+00:00", "exception": false, "start_time": "2026-08-10T18:23:40.372802+00:00", "status": "completed" } }, "source": [ "### C.8 The construction as one function\n", "\n", "The three groups run in that order - per entity, then the gate, then within the date - and\n", "wrapping them in one function is what lets D.3 re-run the identical construction on a shorter\n", "panel and compare the results value by value." ] }, { "cell_type": "code", "execution_count": 16, "id": "d762e6f5", "metadata": { "execution": { "iopub.execute_input": "2026-08-10T18:23:40.379988Z", "iopub.status.busy": "2026-08-10T18:23:40.379899Z", "iopub.status.idle": "2026-08-10T18:23:40.382274Z", "shell.execute_reply": "2026-08-10T18:23:40.381789Z" }, "lines_to_next_cell": 2, "papermill": { "duration": 0.005425, "end_time": "2026-08-10T18:23:40.382542+00:00", "exception": false, "start_time": "2026-08-10T18:23:40.377117+00:00", "status": "completed" } }, "outputs": [], "source": [ "def build_features(df: pl.DataFrame) -> pl.DataFrame:\n", " \"\"\"The whole construction, as one function Section D can re-run on a shorter panel.\"\"\"\n", " return df.pipe(per_entity_features).pipe(gate_to_eligible).pipe(clip_and_rank)" ] }, { "cell_type": "markdown", "id": "51a92920", "metadata": { "lines_to_next_cell": 2, "papermill": { "duration": 0.002213, "end_time": "2026-08-10T18:23:40.386875+00:00", "exception": false, "start_time": "2026-08-10T18:23:40.384662+00:00", "status": "completed" } }, "source": [ "The panel is built once here in two stages, because D.2 audits the warmup on the ungated frame:\n", "the gate drops the early rows the warmup stretch is made of." ] }, { "cell_type": "code", "execution_count": 17, "id": "a0c726e4", "metadata": { "execution": { "iopub.execute_input": "2026-08-10T18:23:40.392227Z", "iopub.status.busy": "2026-08-10T18:23:40.392128Z", "iopub.status.idle": "2026-08-10T18:24:09.706105Z", "shell.execute_reply": "2026-08-10T18:24:09.705625Z" }, "papermill": { "duration": 29.317405, "end_time": "2026-08-10T18:24:09.706563+00:00", "exception": false, "start_time": "2026-08-10T18:23:40.389158+00:00", "status": "completed" } }, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "404,500 eligible bars carrying 57 features\n" ] } ], "source": [ "EXCLUDED = {\"symbol\", \"timestamp\", \"open\", \"high\", \"low\", \"close\", \"volume\", \"log_return\"}\n", "per_entity = per_entity_features(prices)\n", "built = per_entity.pipe(gate_to_eligible).pipe(clip_and_rank)\n", "feature_cols = [c for c in built.columns if c not in EXCLUDED]\n", "print(f\"{len(built):,} eligible bars carrying {len(feature_cols)} features\")" ] }, { "cell_type": "markdown", "id": "5531579b", "metadata": { "papermill": { "duration": 0.004152, "end_time": "2026-08-10T18:24:09.715139+00:00", "exception": false, "start_time": "2026-08-10T18:24:09.710987+00:00", "status": "completed" } }, "source": [ "## D. The timing contract\n", "\n", "### D.1 What each construction reads\n", "\n", "Three kinds of operation appear above. A **rolling** window ends at its own row and reads a\n", "fixed number of bars backward. A **cross-sectional** statistic - the three percentiles, and\n", "the volume clip - is taken with `.over(\"timestamp\")`, so it reads every ETF on that date and\n", "no other. An **as-of join** carries the most recent macro value at or before the row's\n", "timestamp. D.2 and D.3 establish that none of the three reaches forward.\n", "\n", "### D.2 Warmup\n", "\n", "A trailing window cannot produce a value until it has enough bars to fill, so every family has\n", "a leading stretch of nulls as long as its lookback. The audit checks that length rather than\n", "describing it: a column carrying a value before its window could have filled is reading bars\n", "that do not exist, and that is the failure it raises on. It runs on the panel before the\n", "eligibility gate, because the gate drops the early rows the warmup stretch is made of. The\n", "lengths it checks are the declared windows themselves, so each column is audited against the\n", "number its own construction ran on." ] }, { "cell_type": "code", "execution_count": 18, "id": "66cb907e", "metadata": { "execution": { "iopub.execute_input": "2026-08-10T18:24:09.721209Z", "iopub.status.busy": "2026-08-10T18:24:09.721138Z", "iopub.status.idle": "2026-08-10T18:24:10.019964Z", "shell.execute_reply": "2026-08-10T18:24:10.019425Z" }, "papermill": { "duration": 0.302677, "end_time": "2026-08-10T18:24:10.020679+00:00", "exception": false, "start_time": "2026-08-10T18:24:09.718002+00:00", "status": "completed" } }, "outputs": [ { "data": { "text/html": [ "
\n", "shape: (9, 4)
columndeclared warmup (bars)first populated barpopulated
stri64i64bool
"ret_252d"252253true
"skip_recent_12_1"252253true
"sharpe_252d"252253true
"vol_252d"252253true
"dist_52w_high"252252true
"sma_ratio_200"200200true
"max_dd_126d"126126true
"hurst_100"100100true
"obv_zscore_63d"6363true
" ], "text/plain": [ "shape: (9, 4)\n", "┌──────────────────┬────────────────────────┬─────────────────────┬───────────┐\n", "│ column ┆ declared warmup (bars) ┆ first populated bar ┆ populated │\n", "│ --- ┆ --- ┆ --- ┆ --- │\n", "│ str ┆ i64 ┆ i64 ┆ bool │\n", "╞══════════════════╪════════════════════════╪═════════════════════╪═══════════╡\n", "│ ret_252d ┆ 252 ┆ 253 ┆ true │\n", "│ skip_recent_12_1 ┆ 252 ┆ 253 ┆ true │\n", "│ sharpe_252d ┆ 252 ┆ 253 ┆ true │\n", "│ vol_252d ┆ 252 ┆ 253 ┆ true │\n", "│ dist_52w_high ┆ 252 ┆ 252 ┆ true │\n", "│ sma_ratio_200 ┆ 200 ┆ 200 ┆ true │\n", "│ max_dd_126d ┆ 126 ┆ 126 ┆ true │\n", "│ hurst_100 ┆ 100 ┆ 100 ┆ true │\n", "│ obv_zscore_63d ┆ 63 ┆ 63 ┆ true │\n", "└──────────────────┴────────────────────────┴─────────────────────┴───────────┘" ] }, "execution_count": 18, "metadata": {}, "output_type": "execute_result" } ], "source": [ "longest_return = max(WINDOWS[\"momentum\"])\n", "longest_vol = max(WINDOWS[\"volatility\"])\n", "longest_drawdown = max(WINDOWS[\"drawdown\"])\n", "trend_slow, hurst_p, obv_w = max(OSCILLATORS[\"sma\"]), OSCILLATORS[\"hurst\"], STATE[\"obv_zscore\"]\n", "warmup_audit(\n", " per_entity,\n", " {\n", " f\"ret_{longest_return}d\": longest_return,\n", " \"skip_recent_12_1\": 12 * WINDOWS[\"skip_recent\"],\n", " f\"sharpe_{longest_return}d\": longest_return,\n", " f\"vol_{longest_vol}d\": longest_vol,\n", " \"dist_52w_high\": STATE[\"extremes\"],\n", " f\"sma_ratio_{trend_slow}\": trend_slow,\n", " f\"max_dd_{longest_drawdown}d\": longest_drawdown,\n", " f\"hurst_{hurst_p}\": hurst_p,\n", " f\"obv_zscore_{obv_w}d\": obv_w,\n", " },\n", " entity=\"symbol\",\n", ")" ] }, { "cell_type": "markdown", "id": "03c6b3f7", "metadata": { "papermill": { "duration": 0.004143, "end_time": "2026-08-10T18:24:10.029488+00:00", "exception": false, "start_time": "2026-08-10T18:24:10.025345+00:00", "status": "completed" } }, "source": [ "### D.3 Withholding the holdout changes nothing\n", "\n", "Trailing and within-date statistics share a property worth checking directly: recomputed on a\n", "panel that stops before the holdout, they reproduce the same values on the rows the two panels\n", "share. A parameter fitted over a whole column - a winsorization bound, a scaler, an encoder -\n", "does not, because truncating the column moves the parameter and with it every row it was\n", "applied to. Building the panel twice and comparing tests the whole construction at once -\n", "every emitted column, not a sample - and does not depend on anyone having flagged the\n", "transform that fits. A value on one side against a null on the other counts as a difference,\n", "because that is the form of the failure a null-skipping comparison hides." ] }, { "cell_type": "code", "execution_count": 19, "id": "b4a0f9fd", "metadata": { "execution": { "iopub.execute_input": "2026-08-10T18:24:10.038304Z", "iopub.status.busy": "2026-08-10T18:24:10.038190Z", "iopub.status.idle": "2026-08-10T18:24:35.678002Z", "shell.execute_reply": "2026-08-10T18:24:35.677420Z" }, "papermill": { "duration": 25.644831, "end_time": "2026-08-10T18:24:35.678359+00:00", "exception": false, "start_time": "2026-08-10T18:24:10.033528+00:00", "status": "completed" } }, "outputs": [ { "data": { "text/html": [ "
\n", "shape: (3, 4)
columnrows comparednull only on one sidemax abs difference
stri64i64f64
"ret_126d"35756400.0
"yield_curve_zscore"35756400.0
"ret_126d_rank"35756400.0
" ], "text/plain": [ "shape: (3, 4)\n", "┌────────────────────┬───────────────┬───────────────────────┬────────────────────┐\n", "│ column ┆ rows compared ┆ null only on one side ┆ max abs difference │\n", "│ --- ┆ --- ┆ --- ┆ --- │\n", "│ str ┆ i64 ┆ i64 ┆ f64 │\n", "╞════════════════════╪═══════════════╪═══════════════════════╪════════════════════╡\n", "│ ret_126d ┆ 357564 ┆ 0 ┆ 0.0 │\n", "│ yield_curve_zscore ┆ 357564 ┆ 0 ┆ 0.0 │\n", "│ ret_126d_rank ┆ 357564 ┆ 0 ┆ 0.0 │\n", "└────────────────────┴───────────────┴───────────────────────┴────────────────────┘" ] }, "execution_count": 19, "metadata": {}, "output_type": "execute_result" } ], "source": [ "seal = assert_values_agree(\n", " built.filter(pl.col(\"timestamp\") < HOLDOUT_START),\n", " build_features(prices.filter(pl.col(\"timestamp\") < HOLDOUT_START)),\n", " columns=feature_cols,\n", " keys=[\"timestamp\", \"symbol\"],\n", ")\n", "seal.filter(pl.col(\"column\").is_in([\"ret_126d\", \"ret_126d_rank\", \"yield_curve_zscore\"]))" ] }, { "cell_type": "markdown", "id": "0c194843", "metadata": { "papermill": { "duration": 0.002654, "end_time": "2026-08-10T18:24:35.684212+00:00", "exception": false, "start_time": "2026-08-10T18:24:35.681558+00:00", "status": "completed" } }, "source": [ "## E. Matrix assembly and coverage\n", "\n", "The panel key is `symbol` + `timestamp`. Raw OHLC, volume and the intermediate log return are\n", "excluded: they are the inputs the features are made of, and a model handed the contemporaneous\n", "log return beside a label derived from the same prices would be reading its own answer. One\n", "null policy is applied once - a row is kept when the six-month risk-adjusted return has warmed\n", "up. Most of the warmup stretch is already gone by then, because the eligibility gate admits a\n", "fund only from its second year, so what stays thin is the 252-session families, and F1 shows\n", "where." ] }, { "cell_type": "code", "execution_count": 20, "id": "70c3f2f9", "metadata": { "execution": { "iopub.execute_input": "2026-08-10T18:24:35.691123Z", "iopub.status.busy": "2026-08-10T18:24:35.691041Z", "iopub.status.idle": "2026-08-10T18:24:35.770428Z", "shell.execute_reply": "2026-08-10T18:24:35.769888Z" }, "papermill": { "duration": 0.083437, "end_time": "2026-08-10T18:24:35.770856+00:00", "exception": false, "start_time": "2026-08-10T18:24:35.687419+00:00", "status": "completed" } }, "outputs": [ { "data": { "text/html": [ "
\n", "shape: (10, 4)
familycolumnsrolerepresentation
stri64strstr
"momentum"13"signal""simple return, and differences…
"risk-adjusted momentum"8"signal""window return over its own dis…
"volatility"6"state""annualized standard deviation,…
"oscillator and trend"12"signal""bounded oscillator, ratio to a…
"range and drawdown"5"state""normalized range, Hurst expone…
"volume"3"state""ratio to trailing mean, z-scor…
"extremes and consistency"3"signal""ratio to a rolling extreme, sh…
"cross-sectional position"3"signal""percentile within the decision…
"cross-asset regime"1"state""rolling correlation"
"yield curve"3"state""level, z-score, regime indicat…
" ], "text/plain": [ "shape: (10, 4)\n", "┌──────────────────────────┬─────────┬────────┬─────────────────────────────────┐\n", "│ family ┆ columns ┆ role ┆ representation │\n", "│ --- ┆ --- ┆ --- ┆ --- │\n", "│ str ┆ i64 ┆ str ┆ str │\n", "╞══════════════════════════╪═════════╪════════╪═════════════════════════════════╡\n", "│ momentum ┆ 13 ┆ signal ┆ simple return, and differences… │\n", "│ risk-adjusted momentum ┆ 8 ┆ signal ┆ window return over its own dis… │\n", "│ volatility ┆ 6 ┆ state ┆ annualized standard deviation,… │\n", "│ oscillator and trend ┆ 12 ┆ signal ┆ bounded oscillator, ratio to a… │\n", "│ range and drawdown ┆ 5 ┆ state ┆ normalized range, Hurst expone… │\n", "│ volume ┆ 3 ┆ state ┆ ratio to trailing mean, z-scor… │\n", "│ extremes and consistency ┆ 3 ┆ signal ┆ ratio to a rolling extreme, sh… │\n", "│ cross-sectional position ┆ 3 ┆ signal ┆ percentile within the decision… │\n", "│ cross-asset regime ┆ 1 ┆ state ┆ rolling correlation │\n", "│ yield curve ┆ 3 ┆ state ┆ level, z-score, regime indicat… │\n", "└──────────────────────────┴─────────┴────────┴─────────────────────────────────┘" ] }, "execution_count": 20, "metadata": {}, "output_type": "execute_result" } ], "source": [ "features = (\n", " built.select([\"timestamp\", \"symbol\", *feature_cols])\n", " .drop_nulls(subset=[\"sharpe_126d\"])\n", " .sort([\"timestamp\", \"symbol\"])\n", ")\n", "assert features.select([\"timestamp\", \"symbol\"]).is_duplicated().sum() == 0, \"duplicate panel key\"\n", "assignment = assign_families(feature_cols, FAMILIES)\n", "register_frame(FAMILIES, feature_cols).select([\"family\", \"columns\", \"role\", \"representation\"])" ] }, { "cell_type": "markdown", "id": "16582f18", "metadata": { "papermill": { "duration": 0.002703, "end_time": "2026-08-10T18:24:35.776371+00:00", "exception": false, "start_time": "2026-08-10T18:24:35.773668+00:00", "status": "completed" } }, "source": [ "### F1. Coverage through time\n", "\n", "The axis is scaled to the data, not pinned to zero: this matrix is dense everywhere, and on a\n", "zero-based axis every family would draw as one flat line at the top. What is left to see is\n", "where the residual percent sits, and it sits in the families with the longest windows - an ETF\n", "admitted by the eligibility gate partway through a year has not yet filled a 252-session\n", "lookback." ] }, { "cell_type": "code", "execution_count": 21, "id": "f654789c", "metadata": { "execution": { "iopub.execute_input": "2026-08-10T18:24:35.786170Z", "iopub.status.busy": "2026-08-10T18:24:35.786019Z", "iopub.status.idle": "2026-08-10T18:24:35.994303Z", "shell.execute_reply": "2026-08-10T18:24:35.993856Z" }, "papermill": { "duration": 0.214174, "end_time": "2026-08-10T18:24:35.995096+00:00", "exception": false, "start_time": "2026-08-10T18:24:35.780922+00:00", "status": "completed" } }, "outputs": [ { "data": { "image/png": "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", "text/plain": [ "
" ] }, "metadata": { "image/png": { "alt": "Line chart of non-null share by feature family on a y-axis that starts at about 0.985 and ends just above one. Six of the ten families sit exactly at one throughout. The four that do not - extremes and consistency, momentum, volatility, risk-adjusted momentum, all built on 252-session windows - dip in three or four months before 2012 and nowhere else, the deepest to about 0.987, and no family is ever materially incomplete." } }, "output_type": "display_data" } ], "source": [ "plot_coverage_through_time(\n", " family_coverage(features, assignment, every=\"1mo\"),\n", " warmup_boundary=features[\"timestamp\"].min(),\n", " title=\"Only the longest-window families are ever thin, and only early\",\n", " subtitle=\"Monthly non-null share per feature family, on an axis scaled to the data\",\n", " alt=(\n", " \"Line chart of non-null share by feature family on a y-axis that starts at about 0.985 \"\n", " \"and ends just above one. Six of the ten families sit exactly at one throughout. The \"\n", " \"four that do not - \"\n", " \"extremes and consistency, momentum, volatility, risk-adjusted momentum, all built on \"\n", " \"252-session windows - dip in three or four months before 2012 and nowhere else, the \"\n", " \"deepest to about 0.987, and no family is ever materially incomplete.\"\n", " ),\n", ")" ] }, { "cell_type": "markdown", "id": "53297d17", "metadata": { "papermill": { "duration": 0.004809, "end_time": "2026-08-10T18:24:36.005389+00:00", "exception": false, "start_time": "2026-08-10T18:24:36.000580+00:00", "status": "completed" } }, "source": [ "### F4. The timing contract" ] }, { "cell_type": "code", "execution_count": 22, "id": "734c73b8", "metadata": { "execution": { "iopub.execute_input": "2026-08-10T18:24:36.011626Z", "iopub.status.busy": "2026-08-10T18:24:36.011505Z", "iopub.status.idle": "2026-08-10T18:24:36.123514Z", "shell.execute_reply": "2026-08-10T18:24:36.122926Z" }, "papermill": { "duration": 0.115768, "end_time": "2026-08-10T18:24:36.123833+00:00", "exception": false, "start_time": "2026-08-10T18:24:36.008065+00:00", "status": "completed" } }, "outputs": [ { "data": { "image/png": "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", "text/plain": [ "
" ] }, "metadata": { "image/png": { "alt": "Horizontal bars, one per feature family, each extending leftward from the decision line by that family's lookback, from 63 sessions for volume and the cross-asset correlation to 252 for momentum and the yield curve. Every bar reaches the decision line except the yield curve, which stops one session short of it." } }, "output_type": "display_data" } ], "source": [ "plot_timing_contract(\n", " FAMILIES,\n", " bar_unit=\"trading sessions\",\n", " title=\"Only the yield curve waits for its input to publish\",\n", " subtitle=\"Register lookback per family; a gap at the right edge is an information lag\",\n", " alt=(\n", " \"Horizontal bars, one per feature family, each extending leftward from the decision \"\n", " \"line by that family's lookback, from 63 sessions for volume and the cross-asset \"\n", " \"correlation to 252 for momentum and the yield curve. Every bar reaches the decision \"\n", " \"line except the yield curve, which stops one session short of it.\"\n", " ),\n", ")" ] }, { "cell_type": "markdown", "id": "be504fde", "metadata": { "papermill": { "duration": 0.002879, "end_time": "2026-08-10T18:24:36.129561+00:00", "exception": false, "start_time": "2026-08-10T18:24:36.126682+00:00", "status": "completed" } }, "source": [ "## F. What the features look like\n", "\n", "Four properties decide whether this matrix can be used at all: the scale each feature arrives\n", "on, whether the cross-section disagrees enough to rank on, how much of the set is one ordering\n", "under several names, and how long a value lasts. `05_evaluation` is where the matrix is tested\n", "fold by fold for whether any of it predicts.\n", "\n", "### F2. Feature distributions" ] }, { "cell_type": "code", "execution_count": 23, "id": "c99c300b", "metadata": { "execution": { "iopub.execute_input": "2026-08-10T18:24:36.135913Z", "iopub.status.busy": "2026-08-10T18:24:36.135851Z", "iopub.status.idle": "2026-08-10T18:24:36.443132Z", "shell.execute_reply": "2026-08-10T18:24:36.442744Z" }, "papermill": { "duration": 0.311365, "end_time": "2026-08-10T18:24:36.443735+00:00", "exception": false, "start_time": "2026-08-10T18:24:36.132370+00:00", "status": "completed" } }, "outputs": [ { "data": { "image/png": "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", "text/plain": [ "
" ] }, "metadata": { "image/png": { "alt": "Six histograms in two rows. Along the top the trailing returns broaden from a narrow spike spanning about minus 0.22 to 0.21 at 21 sessions to a wide body running from about minus 0.55 to 0.88 at 252, every one of them sharply peaked with long tails either side. Below, the annualized risk-adjusted returns move the other way, from roughly minus eight to eleven at 21 sessions in to about minus two to four at 252, and all three are broad single-peaked bells rather than spikes." } }, "output_type": "display_data" } ], "source": [ "plot_feature_distributions(\n", " features,\n", " [\"ret_21d\", \"ret_126d\", \"ret_252d\", \"sharpe_21d\", \"sharpe_126d\", \"sharpe_252d\"],\n", " title=\"A longer window widens the return and narrows the ratio\",\n", " subtitle=\"Trailing return and risk-adjusted return, display tails clipped at 0.5%\",\n", " alt=(\n", " \"Six histograms in two rows. Along the top the trailing returns broaden from a narrow \"\n", " \"spike spanning about minus 0.22 to 0.21 at 21 sessions to a wide body running from \"\n", " \"about minus 0.55 to 0.88 at 252, every one of them sharply peaked with long tails \"\n", " \"either side. Below, the annualized risk-adjusted returns move the other way, from \"\n", " \"roughly minus eight to eleven at 21 sessions in to about minus two to four at 252, \"\n", " \"and all three are broad single-peaked bells rather than spikes.\"\n", " ),\n", ")" ] }, { "cell_type": "markdown", "id": "ddf97fcf", "metadata": { "papermill": { "duration": 0.003097, "end_time": "2026-08-10T18:24:36.450062+00:00", "exception": false, "start_time": "2026-08-10T18:24:36.446965+00:00", "status": "completed" } }, "source": [ "### F3. Cross-sectional dispersion through time\n", "\n", "A cross-sectional strategy needs the cross-section to disagree. On a date where the band\n", "narrows to nothing there is nothing to rank, whatever the average level of the feature." ] }, { "cell_type": "code", "execution_count": 24, "id": "841432b8", "metadata": { "execution": { "iopub.execute_input": "2026-08-10T18:24:36.457135Z", "iopub.status.busy": "2026-08-10T18:24:36.457031Z", "iopub.status.idle": "2026-08-10T18:24:36.557083Z", "shell.execute_reply": "2026-08-10T18:24:36.556631Z" }, "papermill": { "duration": 0.104844, "end_time": "2026-08-10T18:24:36.557942+00:00", "exception": false, "start_time": "2026-08-10T18:24:36.453098+00:00", "status": "completed" } }, "outputs": [ { "data": { "image/png": "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", "text/plain": [ "
" ] }, "metadata": { "image/png": { "alt": "Shaded band of the 10th to 90th percentile of six-month trailing return by month, with the median drawn through it. In most months the band is between about 0.15 and 0.30 wide. It is widest through 2008 and 2009, reaching about 0.62, when the median also falls to about minus 0.4, and it widens again to about 0.44 in 2020. The 2022 widening is smaller, about 0.34, and more than a tenth of all months are at least that wide." } }, "output_type": "display_data" } ], "source": [ "plot_cross_sectional_dispersion(\n", " features,\n", " \"ret_126d\",\n", " every=\"1mo\",\n", " title=\"The gap between leading and lagging ETFs widens in stress\",\n", " subtitle=\"10th-90th percentile of the six-month trailing return across the eligible universe\",\n", " alt=(\n", " \"Shaded band of the 10th to 90th percentile of six-month trailing return by month, with \"\n", " \"the median drawn through it. In most months the band is between about 0.15 and 0.30 \"\n", " \"wide. It is widest through 2008 and 2009, reaching about 0.62, when the median also \"\n", " \"falls to about minus 0.4, and it widens again to about 0.44 in 2020. The 2022 widening \"\n", " \"is smaller, about 0.34, and more than a tenth of all months are at least that wide.\"\n", " ),\n", ")" ] }, { "cell_type": "markdown", "id": "de369040", "metadata": { "papermill": { "duration": 0.007255, "end_time": "2026-08-10T18:24:36.572411+00:00", "exception": false, "start_time": "2026-08-10T18:24:36.565156+00:00", "status": "completed" } }, "source": [ "### F5. Redundancy structure\n", "\n", "Clustering on the distance $1 - |\\rho|$ groups features that carry the same ordering, whatever\n", "the sign. Above the cut two features are close enough that a linear model cannot separate\n", "their contributions. This states the clusters and stops there: `05_evaluation` measures each\n", "feature's predictive content fold by fold and reports which pairs are one piece of evidence\n", "counted twice, and the modelling notebooks decide what to keep." ] }, { "cell_type": "code", "execution_count": 25, "id": "f50beb1d", "metadata": { "execution": { "iopub.execute_input": "2026-08-10T18:24:36.585069Z", "iopub.status.busy": "2026-08-10T18:24:36.584912Z", "iopub.status.idle": "2026-08-10T18:24:37.492189Z", "shell.execute_reply": "2026-08-10T18:24:37.491566Z" }, "papermill": { "duration": 0.913995, "end_time": "2026-08-10T18:24:37.492747+00:00", "exception": false, "start_time": "2026-08-10T18:24:36.578752+00:00", "status": "completed" } }, "outputs": [ { "data": { "image/png": "iVBORw0KGgoAAAANSUhEUgAAAjYAAANMCAYAAABPeOcmAAAAOnRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjEwLjksIGh0dHBzOi8vbWF0cGxvdGxpYi5vcmcvJkbTWQAAAAlwSFlzAAAPYQAAD2EBqD+naQABAABJREFUeJzs3XdYFFcXwOHf7tKLICIgglhQsXdj7yYae2LX2GNvsRfsvfcay2c01kSjiUnsJfZu7AUVUcRKkc6y+/2BbETKgoALeN7nyROZnblzZmZ3Ody5c48iMjJCixBCCCFEFqA0dABCCCGEEGlFEhshhBBCZBmS2AghhBAiy5DERgghhBBZhsESG1uXUti6lGLGvBUA/HPqvG6Zt89TQ4WV5mKP6eftu9N1P31+GIetSykateyervvJyD7VuQ4Pj6DvkHG4l6ql22dSy9PKz9t3p0u7mcGH3xcJ+ZzPjxDiP6lObN7/MrHLU4Ynvn4f1Y61tRXly5SgfJkSmJqYpDasdJNRE7B8bi6UL1MCj0L5DR2KwcS+f+xzZE/X/azduJ3N2/fw6rU/xYsWpnyZEkkuz4wateyOrUsp+vwwztChiExkxrwV2LqUokSlhoYOJVkyW7zpwdvnqe532j+nzhs6nDRhlNoGNm/fo/u3RqNhy449DB/UM8XtlC5RhIO/b0ptOJ+d6OhoAEYM7sWIwb0MHI1hJef9o9VqUavVGBsbf/R+bt/xAqBC2ZIc2LNR7/LMJDIyChOTjz83QojMIzIyytAhpItU9dg8evyEU2cvAlCmVDEAtuz4Pd5612/epV6TjjgWqEDV+q04fe5SvHUS6gk5cvw0Db/pgnupWuTMVw5Xjyo0/KYLBw6fiLPt2+AQPKfMo3TVRuTMV458xWvwbYc+hIWFAzEJ14o1P1O57jc4FqiAW7FqdO41jEePn+jaeL/n6fjJc9Ro0AanAhWp0aAN5y/+C8Rk901a99BtU6ry18n+qzYoKJi+Q8bh6lGFIuXqMWfR6jiv+/sHMmzsdIpV+BL7vOUoWLo2PQeMxufpM9067/91seWX398db3me+PrFuxX1/vF8+F9sVh4WFs6UWUsoU7UxOfOVI2+x6rTvPpgbt+6l6LwAPH/xiu/7j6Zw2bo45C9PoTJ1aNK6B/sP/ZPoOXm/7X9Onaf6V61xLFCBKvVacuL0Bd16/964TdM23+vadi74BbUbtWfbr3/Eae/DW1Hvt3/wyEkq1WmBfd5ynDl/JdGY/tx/lAYtOpO7UCUcC1Sg+let+WnLTt3rJSo1ZOPWXQCcv/Sv7nokthz+u02o76/Ce16PaNt1IO6lauGQvzxFy9enZce+XLx8Ld66Zy9coXaj9uRy/yLetQA4fe4S33ToTZ4iVXHIX56KtZqzeMX/dIlw7LHYupRi3JT59Bs6njxFq/FNh97YupTi5JmY879lxx69PZQpvT5LVm2g54DRuBSunOBnITnfFymxadtv1GzYFqcCFXEu+AVfNe/M3n1HdK+//xfrz9t306Zzf3K5f0HJyg3jXHuI+Z6qUvdbHAtUoEGLzuw7dDxZt0BjP7u2LqU4cPgE5ao3wbngF3zffzQhoaHMWbSaAiVrUbhsXabPXR5n25R8N3xM+4FBbxk5fhbFv2hAznzlKFq+PmMmzSE0LEy3zvvfLz/+byslKjXEpXBl2nTuz/MXr4CYXr5ZC1YC4PPEN855SaxXIPY9mNCQhJ8276Tht11xKlCRL5t14uEjH/buO0K56k3IU6Qq3fqOIOhtcJLXfssvv1Pty9a4FK6MS+HKVKzVnJ4Dx+iNF6D3YE/KVmuCS+HK5MxXjuJfNGDEuJm6fTb8pgu2LqXoOWC0bn/R0dG6W9ELlq5NNC59bSd0zot/0QDHAhVo2bFvnDsj23fupU6j9uQvURP7vOVwKxbzOX7/e+P98/rbH/up06g9OfOVY8dvf1Kq8te69Zq07pElhjSkKrHZsuN3tFotjg72LJo9HoAHjx7H+SIKCwundad+XLh8DY1Gg1qtpk3nAclq/9ZdLy5cvoaVlSVFCruj1Wo5fe4y7boN4trNO0BMxtm4VXeWrvqJR95PcHJ0ILutDYePnyYiMhKA4Z4zGD1xNrfueJE/rysqpYrdew/wVfPOvHz1Ot5+W3XqR1hYOOroaP69fptu/UaiVqvJncuRwgX/u9VToljM7YZ8bi56j2XyzMUcPnYaExMTnj1/ybQ5yzhy/DQQMzajUcturNmwjecvX+OePw9v34awfdeffNn0O169fhOnLb/nL+j7wziMVCoccuZIcH/2ObLrbs2UL1MCu+y2utfMzUwBaNdtEPOWrOHR4yfkc3NFrVbz574jfNW8E3fvP0z2eQEYOmYaO377k5CQUIoUdsfE2JgTpy9w8cp1vecGoE3n/kRGRqFUKLl5+x5tOvfnmd8LAB77+HLi9AVMTEzwKFQAExMTLl+9Qa9BY9l36Hiy2m/ffRChYeHkdnZMdJ1tv/5B+26DOHP+CpaWFjjktOfajTsMHD6JuYt/BKBkscLksIu51WVtZUn5MiUoWaxwostTonu/kfx94BhqdTQehQqg0Wo5ePQkt+89iLdui3a9CA4OIUqtjnct/jl1niatv+fwsdMolUpcc+fi7v2HjJ+2gB9GTY3X1qr1m9m5ex8uzk6Ym5lRvkwJrK0sAchhl13vLeKUXp/JMxdz/NR5TE1N430WUvN9kZA5i1bTf+gErl67hb29HdbWVpy9cIUO3QfHS7wABo+czK27XhgbG/HYx5fBI6foPgvPX7yibZcB3LxzH6VCyRv/QLr1GZHimLr2GY5SqSQ0LJwdv/1JnUYdmL9kLdbWljx/8YrZC1dx6OhJIOXfDSltP/b7c9W6zbx69YZC7vl54x/I8h830bbLQLTauPO3nrt4hXFT5mNibERwSCj7Dv2D5+R5AHgUyo+zkwMAJibGqb41PHzcDF6+fI1Gq+Hcxat827Ev3fqOQKVS8TY4hJ179iWZPFy7eYe+P4zj+s07OOS0J4+LM77PnrN9595kxfvn/qMEBAaR182V3M5OPHn6jNXrtzBg2EQAun3XGoDf/zpMYNBbAE6dvcSr1/4olUpaf9M40dj0tf2+C5f/ZfzUBVhamBMVpebg0ZN06D5Yd20uXb3Bzdv3sctug0ehAoSHR3D42Gmat+ulSzrf13PgGHz9XuDmmhuFQkGJ976nChfMnzWGNERGRmg/5r+IiHBtkfJ1tRYOHtrhntO0kZER2go1m2otHDy0vQaO1q23ZsMWrYWDh9bCwUP7577D8ZZNmrFQGxkZoT109IRu2T2vB9rIyAjt/QcPtS9evtS19eLFS61j/nJaCwcP7fhp87SRkRHaDZt/0W03d9Eq3bpXr93QBgcHa+/e99JaOhbRWjh4aDds/kUbGRmhfePvry1YqmacdtZv2q5rZ8nK9drIyAjt4hXrdMuu37ydaJxJ/Re7bq2GbbTBwcHaZ35+WpvcxbUWDh7aMRNnaSMjI7TrNm7Trbdzz1/ayMgI7bmLV7RWTkW1Fg4e2onTF2gjIyO0k2Ys1K23cu1G3XUIDw/Tdu83Qmvh4KGt37RjvBhOnj6vzZGntNbCwUM7ZPQUbWRkhPbgkX90bS1cvlYbGRmhffjosdYhX1mthYOHtlvf4Sk6L+WrN9ZaOHhoN23dqdvvY58nutcT+u/9tlet26S7btmc454fnydPtU+e+uq2CwoK0hav+KXWwsFD26X30Hjnev2m7fHaj20rMjJCGxYWmmA8hcrU1lo4eGhrfNVK+/btW21ERLi2dae+WgsHD22OPKW1AYEB2sjIiETPdWLLx06arS1VuYG2QYtOSb5XYs/98ZNndMvu3vfSvc+Sey3qNemgtXDw0BYuW0f3+RkyeorWwsFDa+lYRHv3npc2MjJCW7hsHa2Fg4fW1aOy9sEj7zjnpn7TjloLBw9t934j9L7HU3p9kvosJPf7Qt/7KTIyQusf4K9737fu1FcbHh6mffv2rbb6ly115ycyMkJ7z+uBbrt2XQdoIyLCtZeuXIv3WRs/bZ7WwsFDa+VUVHvpyrU4y95/3yX03/uf3djPSJ1GbXXLjp88ow0PD9Ndk9R8N6Sk/f/9vENr4eChtXUpob11+642MjJCe/HKv7rtDhw+Hue9beVUVHvxyr/ayMgI3Wcjb7Gq8Y4z9tzG/vf+OT509IRueWw8Cf0e6D1oTMw5nvrfOZ4+d6k2MjJC26X3UN1nNbFzvmPXXq2Fg4e2ZKWvtOHhYbr395Hjp/TGG3se3v85No5szsW1QUFB2uDgYG2eIpXjvEcGjZiotXDw0H79bZckPzP62n7/nGdzLq69eu2GNjIyQrtq3aZ4n49bt+/qvpsiIyO0t+/c062zZsOWeOe1c88hcc5HYtcmM//30T02J05fwPtxTNd0m29jMtO27/6/e+8BXTfm7bsx4w4szM2oV7sqAC2afJXMpCuSvkPG416qFnZ5ypC3eHXeBocA4Of3EkDX3WZqakK/nt/pti1S2B0TE2Ou/HtTl9n2GeyJrUspXApX5umz5wBcuBS/mz/2eDwKFdAte/Eyfs9OSjRv8iUmJsbksMtOzhx2MW2+6y26fPUGEHOOGjeoA8SMOSpYIC8AV/69GactczMzunRoCYBCoUCpTPwyPnj4mLZdBxIWHk7zxl8yY+LwOPsEaNU85hZJbmdHKlcs+26fN+K1ldR5aVC/JhDTfVqmamPadO7P9p17cXLMmfSJeefbZjExFCnsTlEPdwBu3r6vO0bPyXPxKFePHG5lcXKvyINHjwHwe/4yWe336dFR92+VShXv9ZevXvPkXdd+k4Z1MTU1QaFQ8G2zBgCEhYfrxtCk1ITRgzh/bDd7tv2Y5HoN6sWcwyate1CxVnO++34Ih46ewskh/jlM6lrEXtsv61TD1iYb8N811mq1XLkW9/3U9Ou6uDg7AQmfG31Sen2S+iyk5vviQ7fveBEWHnM7+ttmDVAqlZiamtD063pAzO2HD3s8WrX4GoVCEecv1hcv3+jaAyjknpfiRQvFxNY45bHFflbyuDgDYGuTjUoVyuh61+DjvxtS2v6ldz2qkZFRlKvRFFuXUlT/srWurfOX4t7iLOpRkBJFY/7Cj33vpfa7MTEfHgf89xnJmyeml/xFAj3usSpVKI2tTTbuP/AmX4ka1G3cgaFjpiV7/0f/OUPlut/gVKAiti6ldL22arWaV2/8MTExpmOb5gBs2vobWq2WvX8fBqBdyyapavt9RT3cKVI45jsx9nsS4ObtmCEDAYFBtO82mLzFqpPdtTRlq/+374Q+fz27ttP9zviYz3tm8NGDhzfv+G/QcONWMeNOot91hQe9Deb3Pw/pvnwh5ssvpVp3GsCDR48xMjKiqEdBzExN+PfGbSIjo4jWaOKtr28fJYoVjtedHvtBf1/sL4P3L/qHXbIpZZPNWvdvlZHRuzY/rq0cObInmczEevX6Dd927MvrN/5UrVSeVYumfdR1iJXUeRk3cgBfVCjNoaOnuHXnPqfOXmLfoX84cfoC239a+tH7hJiu06P/nNH9wrG0sODOvQe8DQ4hOjr++yAhid2yy0hWLppKwy9rcuL0BW7fe8CBwyf4/a9D3Lpzn7nTxsRZNy3fozntU3duUnp9kvNZSM37NDViYzMy+u+rMbWf/Q9ls7YC/jt263c/w3/HnZpdfkz7JibGlCzmEa+t2PdZrDjXLgW/FN+/npr3vruD3r5NdJvY26Eqo//2Y21tGae9pM6To4M9Zw7vZNuvf3Dl2k1u3r7P/37+lZ+27GL/bxsoX7Zkottu37mXcVPmA+DkmJPcuRx57R/AI++YcZmxY9W6fteKxSs3cOnqDTZu2YWv3wusrSxp8nXdVLedHMEhoXzTsQ+BgW8xMzWlZHEPjI2MuPDuD/6EPn85c9olu/3M6qN6bIJDQtmz94Du56CgtwQFvSUk9L/BZrGJT2xWHxIaxuFjp4CYHh193vgH6P7qGzOsLyf2b2ft8lnxvvDKvXukNiIikuVr/nsq5s69B0RGRlG6RFHdNu1bNePg75s4+PsmDuzZyGTPIfTq1j5Fx25hbqb7d+h7x5sasQOvQ8PC+eNdxn/l2i3ueT0CoHTJonHWT86XfmhYGG06D+Chtw9FC7vz89oFmJr+l9TF7hNgx29/AfDU97lufFTpksVIiTPnL1O1UnlmTxnF79vXsHBWzIDq2MHl+uzasw+IuW6xPTWxPTcX3v3V2Ln9N5w+tJMdPy3D0tIiRfHpO2c57XPg8i7J/f2vQ0RERKLVavl1999ATC+ZR+ECSTWRqOQOHj519hKNG9Rlwcxx/PXrekb+EPOU26kzyTuHsWKv7f7DJwgIDALgl3fHoVAoKF1C//sp9n2enPd4WlyfWB/7fZFgW4ULYG4Wcxw79+xDo9EQERHJ738dAsDVxRn7HMn/ko/9q/melzd33o172vXHvo+KLblS+t3wse1HR2uYO32s7vvxjx1rGdC7Cy2bp+wx6Nj3TVhYWJyEMOd75/n+A28gptciMDDxxCa1nvm94NVrfwb17cr6FXM4e2QXhdzzodFoOH3+cpLxxvZUWVtZcvXUnxz642fq1Kgcbx9urrmpVyumZ3H0xNkANPm6Lhbm5onGldy2Y928ff+/99ue/95vRT0Kct/rke4cLp03iWN/bWXGxKTHfX34eY/zOy0sbX6nGdpH9djs3ntAl8ScPvSr7gMPsGLNz4yeOJt/Tp3nia8frZp/zYy5y3n2/CVtuw4kf15XfJ48S6xpney2NuTO5cjTZ8+ZMW85v+z6E1+/FxipVES8t963TRuwcu3PXL12C8/J81i9fgvGRkY89H7Cg2vHyOvmQuf23/C/n39l9MTZrFz3M5YWFjx5+oygt8Esmz9Z162cHPnyumJsbERUlJrm7XrhmjsXA3p1plnj+slu40MtmzVk+eqN3Lxzny69h+OePw+PvJ+i0WjI5ZiTnl3bprjNVWs36wbuhkdE0rJjX91rc6ePpUbVitSqXomj/5xh7KS5bPj5V3yfPSc4JBQrSwuG9E/ZqPhJMxZx6eoNcudyIls2K+7eixlwWaxI8s7tuKnzWbF2E499fFGr1ViYm9GzaztdG2cvXOGnLbs4c/4yz56/REHa/0U/bkR/eg0ay4XL1yhRqQGmpqb4PPEFYOjAHkl+WaWF3oPG4h8QRG5nR8zNzXS3ZZJ7DmONHtqHFu174/PEl9JVGpHDzhavhzF/JHzXtgV5kzHYvaB7Xg4ciekxqtGgDTlz2PHrzwlPjpeW1+djvy8SYmlhwZAB3Zk2Zxm//3WIkpW/JioqSjeg0nN4vxS116NzG5b/uJGQ0DDqNGqvG/SZntLjuyFe+z9u4satu9Rp1J7CBfMRFaXG5+kzIiIiuXr6z3i9Nkkp6J4PgFev/SlfoynZbW1Ys3Qmed1cqFiuFOcuXsVz8jx++2M/Fy9fQ6lUxunBSUt37j2gebte2OfIjpOjA2+Dg3XDJ4p5FEwy3uLvPnNvg0MoXeVrTExNCApK+Amsbp1asf/wP7rfie1aNk0yrpS0DWBqYkythu1wy+PM3fuPgJi7D3VrVSEw8C2WFuaEhIYxYNhE5i9dm+CA8qTY57DDLrstb/wD6DVwLAXy5aH1N41S/Ed/RvJRPTaxc9e453eLk9QAui642DltzM3N2PbTUsq+10Owac2CpINSKlEoFPy0eh5lSxVDpVQRrdHw45IZ2NnFHWFvYmLM79vX0L9XJ9zy5OaZ3wve+AdSq/oXuttO82d4Mn3icIp6FMTv+UuePH2Gq4sz/Xp+R7XK5VN07HbZbZk1eSQuzk68ePmaC5ev8fxl/JHnKWFmZsreX9bRo3MbHHPm4P6Dx1hZWdC6xdfs37MxRX9VxgqPiNT9+8Gjx1y4fE3339t3jxRuWbeIoQN64JYnN14PH6NSqfj6q9rs++0nCr37wCdXiyZfUaZkMd4GB3Pz9j1sbKz5tlkD1iybmaztd2xchqmpKeroaIoWdmfL/xbjnCvmCabl8ydTvUoFzExNCA0LZ8bEESn+ZZ8cbb5tzOZ1i6hUoTTBwSG8ePmKEsUKs3jOBIYN/P6j243tNYntgUpM+zbN8ChcgNdvArhz7wGOOe3p0uFb5kwdneR2H6pepQK/b/+R2jUqodFoePzEl0Lu+Zg0ZjALZnomq40BvTpTq3olLMzN+Pf6bS4nMJYjVlpen4/5vkjK8EE9WTJ3IqVKFOHVqzcEBQVTsVwpfl67MM6t8uRwdLBn6/+WUKRwAaLUamyyWbN4zsT/Yn/3tGFaSo/vhveZmpqw95e19OoWk6jdf+BNQGAQZUoWZdzIATik8DZlg3o16Nz+W+yyxyTTFy5f0/UCLF8wmcpfxIzh8332nLnTx5L73diu9JA3jwvfNmuAtZUVXg+8ef1u4sxFs8ZTp2aVJOP9rl0L+vX8jhx22QkOCaVa5fKMGdY3wf18Wac6ru/GAbm6OOv9nZKStiGm93zW5JGEhIZjbGxEnZqV+XntQhQKBba22fjfyrl4FMqPRqvBxMSYresXp+g8KRQKFs0eT/68eXgbHMLFK9fjTCWQGSkiIyPS9gbyR9q552+69R2JQqHg2b2zmKXDl0RKteg0mMG9O1KzSnnafz+Svt3bUq1SmRRtJxL38/bd9BsSM01AwJOrBo4meVJ6bbVaLQVK1iIsLJzTh39lzJQlyX4fpdaxUxdYuHITu35aCJCi97BI2JZf97J99wHdOZ27+Eemzo4ZQ3bu6G8p/oMgI9m7/ziX/r3FuGHJn+jzY7bJir7t0IdDx04xfFBPxqawJzAxfX4Yx5Yde6haqTx7f0n8sXYRX6pnHgaYOm81e/cfZ8uPs8ibJ3eKtx8xbqbuPnrlimUzRFLzoc0/zjJ0CCIT+vLb73njH8CEUQPJm8fFoO+j5O77c0nMP+Y4Z89fwfOXr2nXbRB+z1/qnlpq/U2jDJ/UZPTrqlarWbjyZ/YfOQko+KpOFQb17ohRAoOU6zSLe6s8MkpN3jzObFo54xNFG2Pu4h85dfYih4+dxsLcjO+7pO7WoEgbqU5sQkLDOHz8LNmsrdjz9zEG9kz5fbm/Dhzj7dtgalb7gkWzxsd7XavVotFoUamkGHlGkZmvSWKxp8cxWVlZsXv72gz7y0SkTNkyJdj792GOHDuNQqGgeNHCtG3ZmN7d2mfqz0RGsH7zbv69cYfNq2MS8CGec9iwZQ/dO7aIt+7h3XF7MDr2Hk29mpU+SZzvO3L8DKfPXaJAvjxMnzg8Uzx9+TlIdWJz6NgZzMxM6d2lFSv/t4O+3VpjZGTElp1/cfLMZZbO/u8x1YNHz/Djxl/ZtnYOoWHhLF+7jRNnLuGaJx8tv2nG0H6dsHr3NEWLToNp/nUd/jl9kXsPHrN28WTOX7rGzj8O8cY/kOy22WjTogGtmn2pa//Fy9dMnfcjN+7cx9XZiVrVKrD7ryO6buP39xkZGcUX5UvG2WdS3v9rp0WnwXzbuB5HT57nofdTCrnnZeKIPjg6xH9Tv/EP5AfP2VQoU5x+3duiUCjY8uufiR5HWh/DG/9A5i3bwKV/b2FqYkKDulXp0elb3V9BKTmWhK6Jez7XJI8nqfY7tG7GL38cYXDvjomeqw+FhISyYv0OTp69RFBwCG4uuZgxbjCODjmSPNbEYk9oubNTzmSf46SOfczUxTx/8ZrxM5ahUir5qk5Vzlz8N85fzWl5fV68fM20+T9y/fZ/750Pr9/7+97y659s27WPt8EhZMtmRdd2zThz8Vq8mEcO6pbkcSYnzqSuW2o+l0m1W/mrjmxYPo1CBdwA2Lrzb46fvsDyOZ4JXpuRg7ol45xW4k1QmO7z2KLTYKyz2dJ76JRkfU8dPHqGrbv+Ys2iSQCMnryIa7fu8seWZQAsXvUzUWo1Q/t1TtG1j72eKXkvJnS8CVGr1azdtIs9fx8lPDyCGlXKMeaH7zE2TpMOf50/9h1jUO+Oupl/O7drxtIfNyeY2Lzvxm0vHnk/pdGXNdI0nuRIz1tEKxZMYcWCKenWflaW6nfm7/uO8VWdKtSrVZmFKzdx4sxlalWrwFe1q7B87Vaev3it+yD+degEDetWA2DavNWoVCo2rpyOkZGKGQvWMG/ZBiaM6KNr+88Dx5k9cQguzk5oNBp8njxj6awxOOS049LVWwwZN4dC7nkpVSxmoOKEmctxdXFi9qQhvHj5miGec+LEmpx9Jtffh08ye+IP5LCzZfTkRaz+6Zd495l9nvoxxHMO3zSpR7tv/nts0snBPtHjSOtjGD9zGTmy2/LrhgUEBgUz1HMOZmamdGnXLEXHEuvDa6LveFJ7rj40Zd5qwsMjWL1wIjmy23DvwWPdY+z6jjWh2BNaPnHW8mSf46SOfbrnwHjJRItOg9Pt+kyYuZxcTjnZu2UZfi/iv3fe9/jJM1Zt+IX/LZ1K3jzOvPEP5I1/IE0b1k7wloW+a6wvzqSuW2o+l0m1m5SErk1CknNOU/I9VbZUESbOXkFIaBgW5mZcvXEHM1NTHj1+St48ublw9SbdOjRP1jn9UErfi8k1dd5qLMzN2bF+HoFBwfQaMon9R04lmUjMWbKe/UdOJ/r63MlDKVX8v6n8g96G8OLVGwrmz6NbVqiAG34vXuue1EzM7/uOUqlCKXJ+ZOkGkfWkqs/0ofdTrt+6z9f1qmNhbkbNquX5fd9RAOyy21ChTHH2HYmpSfImIJDzl6/ToF5V/AOCOHryPMP6d8HayhJzMzO+79SSg8fOxJlQqEWjeri5OqNSKTE2NqJ29Yo4OuRAoVBQrnRRKpUryeWrtwB4/uI1V67foW+3tpiZmpDHJRfNG9XRtZXcfSbXN43r4ezkgKmJCV/VqcLte3FrK92++5B+I6bR47tv4/2iTuw40voYXrx6w8UrNxnUqwMW5mbkcrSnc7tm/HkgbmFKfcfyvg+vSVLHkxbn6n1v/AM5dvICowZ3J+e7SQoLu+fF1sY6WceaUOwfLg8OCU3ROdZ37ElJy+sT+97p/307zMxMyZvHmRbvvXc+pFQq0Wq1PPR+QnhEJHbZbXB/75fKxxxnYnEmdd1S87lMqt20kNxzmpLvKbvsNuTJ7cTV63e46+WNk4M9Vb8ozcWrNwkMCubBoyeUfW9umpR8NlPzXkzMvzfucuf+I4b266x7j5YvXYyHjxMuiBpr+ICuHNi5OtH/3k9qAF3B4thJ+QBdMpPUXEph4eEcPHqGpg1qfeQRiqwoVT02v/99lIL581DwXVdvw3rVGTJ2Ni9evcHB3o6G9aqx/uff6NSmKQeOnKZEkYI4Odhz844XGo2Wbzv/EKc9pULJa/8AHOxjHmF0+qDLdd/hk2z+9U+e+b1Cq9USHhFBLqeY6eZjp7h+/0vNKae97t/Pnr9M1j6TK4edje7fZmamhL77YMba8/dR8rjkom7NL+Jtm9hxpPUxvHz1BhMTY+yy/xdr7lwOvHgVd54Dfcfyvg+vSVLHk9z2EzpX+w6fZNaide/2ac/mH2fx7PkrTIyNcXKw50PJOdaEYv9wub5znNJjT0paXp/Y946d7X/rOjnGP0+xXJwdGTesF7/sOcDU+asp7uFOvx7tdLdtPuY4E4szqeuWms9lUu2mheSe05R8TwGULVWUi1dvkiO7LeVKFaV4UXf2HT6Fna0N7vlcyWb93y/3lHw2U/NeTMzRkxeoXrlcnHFDgUHBuOdLPAn+GObvJokLDgnVff8Fh4QCYGGR+PxRh4+fw8zMhCpflE7TeETm9tGJjVqt5u9DJwkND6dR25jH27RaLdEaDX/u/4cu7ZtRvXI5Zi1ax+17D/n70Em+aRJTo8UxZw6USgW/b16S5BNQCuV/Yyz8XrxiypxVzJ82grKlimCkUjFy4gLdbJH2dtmJjIwiIPCt7oPh9978MsndZ1oZ3Lsje/cfZ+zUxUwbO0A3RXtSx5HWx5DT3o7IyCje+Afqfnk+e/4yxUnc+96/JvqOJ7kSOldf1anKV3Wqxlkvl6M9kVFRcW5vxkrOsX4Ye0LLU3KOk3PsyiRmPE7L6xP73nkTEKj7Rfz8RdI1fOrVrES9mpUIj4jkx59+YdLsFfy8ama8mFN7jZO6bqn5XCbVLsTMKxMe/t90nq/fBMR5PalrA8k/pyn5ngIoV6oIP237Hbvs2WjV7CuKeRRg9uL1ZLexpmypj5tJOLXvxcRcv3WPWlX/G6sVGBTMpau3GNSrQ5LbzVq0jn2HTyb6+vypwyld4r8SDtmsLXGwt+OelzcuzjHzV9178BjHnDmSvA215++jfF2veoJPTonP10ffivrn9CVCQsP439Kp/LR8Gj8tn8bGFdPp2r45f+w/hlarxczUhNrVK7Jy/XYePn5KneoVAchhZ0uNyuWZu2wDAe+mg379JoCjJ88nur+wsAi0aLGzzYZSoeDUuSucfa+ApaNDDkoWK8TK9dsJj4jE56kfu/88onv9Y/aZGiYmxsya+ANRUWpGT1lEVJRa73Gk9TE42NtRrlRRlvy4mbDwcPxevOJ/W3bzdb3qaXac+q5LciR2rj5kl92GGpXLMXvJOl699kej0XDn/iMCg96m2bGm5Bwn59jtstvw1PdFgvtKy+sT+95ZvnYb4RGRePv48tufhxNd39vHl3MXrxEeEYmxkREW5ma62j8fxpzaa5zUddN3vtds/JW+w6emuF2Awu55+fvQSdTR0dz18ubvQyfibZ/YtYGUn9PknqsyJYtw78Fjrt28T6nihbC2ivmlvu/wKcqX/rjEJrXvxYRotVq8Hvrwz5lLhISE8vpNABNmLqNB3arkcYlfY+99Iwd14/DutYn+935SE6vRlzX439bdvH4TwOs3AWzYupsmSdxi8vbx5drNe0muIz5PH53Y/L7vGPVrVSJvHmdy2Nnq/mvd/Ctevvbn4tWYmUob1qvG2YvXqFG5HJbvdSl6DuuJtZUF3QaMp26LHvQeOoU79x4lur98brnp3LYZ/UdOp0Gr3hw8dobqlcrGWWfSqL48ffaCRm37Mm76UhrUrYrJe2MpUrrP1DI1MWHm+MFotTBq8gIiI6P0HkdaH8OkUX2JiIikxXeD6fXDZKpWLE3H1o3S7BiTc12SI6FzlZBxw3vhYJ+DrgPGU//bnsxevJ6Id7Msp9WxJvccJ+fYO7Vtyi97DlD/m57MWbI+XhtpeX0mjerLi5dvaNSmLxNmLqfxlzUTXTdKHc3qn36hUdu+NGjVmwtXbjJuaM8EY06La5zUdUvqfPu9eE3JJEqeJNXukL6duH7rHl9+05Pla7fSsH7chFHftYGUnVNI3nvC1saafHlyk88tt66WVfnSxQiPiEzwF35ypMV78UNPfJ9jbGxErarl+bbLELr0H0dh97wM6dvpo2LUp1uH5hQvUpB234+g3fcjKFm0EJ3b/VeeYNaidbrb0xDzO6hU8cK45k6/2YtF5pRhZh5ODxu27uHilZssnjnK0KF8tKxwDEJ8rA69RrF8ztg4VaVF+vhwFuFDx8/yy54DrJibeBkOmXlYZERZaiapO/ce8uixL1qtltv3HvLL7v3UqVHR0GGlSFY4BiHSys+rZkpSYyD3vLzJn1d/wVQhMpq0nWHJwPwD3zJ78WLeBMRMUNW0YW2afFXL0GGlSFY4BiFE5lOwgBtWVv8N1L334DFVK5ZO0TZCZARZ+laUEEIIIT4vWepWlBBCCCE+b5LYCCGEECLLkMRGCCGEEFmGJDZCCCGEyDIksRFCCCFEliGJjRBCCCGyjBQnNlqtFrVanaIih0IIIYQQn0KKE5vo6GiqN+pCdHR0esQjhBAZjjrsDa9ubkUd9sbQoQgh9JBbUUIIoYc63J83t7ejDvc3dChCCD0ksRFCCCFEliGJjRBC6GFs6YRzpVEYWzoZOhSRRUyZuwqvRz5p1t6Va7dp//1IGrXtF++1/UdO0bB1nzTbV0YniY0QQuihMrHEyrkiKhNLQ4ciRILc8+dh3ZLJONjbxVkeHa3h8D/ncMhpl8iWWU+Wqu4thBDpQR3uT5D3YbK51cHILLuhwxGZkFarZe6yDXg9eIxKpcLMzIRfdh/g6bMXmJuZMnPCYDQaLVPmruT5yzdYmJsyYURfQkJCGT9zGTnsbHn2/CUDvm9P+dLF4rVvZZlwlfUDR09Rp3pFtvz6V3ofYoYhiY1I1IGT1/H2fWXoMIQwOCteUFa1lf3/RhOMg6HDERlMj1a19K5z4swllAoFK+ePB2DS7JWUKFqQkYO64TltCV4PfXj81A8HezsmjuzLXwdPsGP3fr6uV43XbwJZNnssoWHhDJ8wj/IL4yc2CYmO1nDo+FlmTfjhs0ps5FaUSJS37ys0GpmvSAghUuvRY1/KlCyi+1mpVFDIPS8ADjlzEBQcwhNfP4oUyg9AkUL5eeLrB0D+vC6YmBhja2OdoqlW/j58gro1vkCp/Lx+1UuPjUiSUqlI1l8jQmRl4f5ePD6yleb1ymGWvYChwxGZUN48zpy/fIM61SsCoNFoUSjeW0GrxcXZiZt3HlC7ekVu3X2Ai3PMYPWH3k+IilITGhaOSqVK9j4fefty1+sRfx86ic9TP+Yv/4khfTul5WFlSJLYCCGEEOmsWqWynLnwL72GTMbo3RibD9WoUo6jJ87TZ+gUzM3NmDCiD6GhYTjY2zF+xjJ8n7+gX/d2Cbb/6PFT5i37icdPnzFg5Az6f9+Ofj3a6l7v2n/cZ5HUACgiIyNSdK9BrVZTvVEX/tn7P4yMJC/KytbsOAok7/6xEFlZZIgfr65vxL74d5jII9/iE3rm95IlP25m+rhBhg4l05DMRAgh9DCxdML5i+GGDkMIALbt+ptjpy7ofi6Q15Wh/TobMKKMJcv22MgTPamn0WhljI0QgFYThTo8ECMzGxRKY0OHI4RIQpYdKi1P9Agh0kpE4GMe/t2TiMDHhg5FCKFHxu1ySQPS25A6sWNshBBCiMwiy/bYCCGEEOLzI4mNEEIIIbIMSWyEEEIIAwoNC6fv8KnJXl8qeSdNEhshhNDD1DYf7s22YWqbz9ChCCGVvPXI0oOHReppNFoZRCyEEElIzkMqb/wDGTd9KdHR0dhlt2HKmAEsXLmR+w99cM/nCoBGo+GHsbP5oU8ntFoti1f/zPypw1HEqb0glbz1kR4bkSg3Z3uUSoX+FYXI4szxp6TyV8zxN3QoIpOytrJk0cxRrJw/npz22dmy808Cg4JZMdeTKhVLA6BUKhkxsBtzl/6Pecs3MGJA13hJTWJiK3nXq1kpHY8ic5AeG5Go+lWLGzoEITKEmCKYG/m2Xkkpgik+SmDQW+Ys+R9BwSG8eu1Pl/bN8CgYc2uzaOH8uvVy53LA1sYaCwszcjnlTHb7n2sl74TIGRBCCCHS2f4jp6n6RWlWzPWkUvmSBAS85a7XIwBu3X2oW+/ytdsYGakIDAzmnpd3stt/5O3LXwdPMHjMLF0l78+V9NgIIYQQ6ax8mWJMmr2CE2cuY2pqgkfBfFhamNNn6BSKFIrpsQkLD2fV+u3MnjSU8IgIJsxczuIZozA2jvurWip5Jy3L1oqSytRCiLQScytqOHlqz5FbUUJkcBk3MxFCiAzC2CInjmX6YGyR/DEPQqQFqeSdchk2sUltde7YytRCCJFaKtNs2OSrb+gwxGeoTYsGtGnRwNBhZCoZdvCwVOcWQmQU0RFBBD48QHREkKFDEULokWF7bCB11bllUjkhRFqJCn3J88srMLXNj8o0m6HDEUIkIcP22AghhBBCpJQkNkIIIYTIMiSxEUIIITKAZ34vGTNl0Udv/9ufh3X//mnbHnz9XiR727fBIYyfsZT+I6Yzc9FaAE6fv0rPHybRa8hkpsxdhUajibNNSquSfyoZeoyNEEJkBEojc8zti6E0Mjd0KFlG+LTpRJ89a+gw0oTlnt2fdH8ajSbB0gm7/zxC86/rANCpTdMUtblm4690aNWYwu55dcvKly5G5QqlAJgydxXXbt6jVPHCHx/4JyKJjRBC6GFi7YxrjSmGDiNLiT57FtRqyMATvaaluUs30LBeVYp5uHP+8nXOXPiX128CeP7yDRbmpkwY0TfJ7dds/JVnz1/hHxBE766t2bT9D16+9kej0TBpVF9u33vI4yfP6Dt8Ks0a1uHcpWu0b/k1eV1zM2Xuyjj7yWZtGa/9u/e9CQ8/iI+vH21aNKBmlfK6GY+12pgnlHM52gMwb9mGOFXJM5rP4x0lhBCpoNVq0GqiUShVKBRyBz/NGBl98t4OQ6lfqxIHjp6hmIc7h46dxTW3Eyqlkokj+/LXwRPs2L2fr+tVS7INh5x2jBvWC4AxP/TAzMyUoyfP89vew/Tu2poNW/awfI4nAOcuXQPg2KkLONjbxdlP944t4rV9444XA3q2J2+e3PQZNoVK5UtiamLC3v3H2bj9D1ycHbG1ycatuw90VclPn7+K1yOfND5TqSefUCGE0CMi4CH3d7chIuCh/pWFSEDJYoW4cfs+UVFqHj5+SrQmWlcjqkih/Dzx9dPbRtF360dHa1i6Zit9hk7hp617ePXaP9Ftnvj6JWs/DvZ2FC1cAAtzM9xccvHyVUybjb6swdY1s3F0yMGxUxd44vs8warkGYkkNkIIIUQ6UygUlChaiLWbdlKhTDFcnJ24eecBALfuPsDF2SkZbcT8yr73wJvgkBBWzBvHd62boNW9Hn+b5O6nYH43fJ76ER2t4YnvC+ztbImMjNK9bmlhjpmpCS7OjglWJc9I5FaUEEII8QnUr1WJ7wdNZNOqmbjkduToifP0GToFc3MzJozoQ2hoWLLacXPNhd/zVwwcNRM311y65WVLFWXEhPk0+qqGblmNKuXi7Schvbu2YubCtURERtKsYW3MzEzZ+cchDh47jVYLeXI7UfWLMiiVynhVyTOaDFvdO7XVuaW6txAirUh177QX0rQZ8OmfKBJZX5busdFotFJaQQiRala8oKwKfjt4kWAy3mDJTKltf9ptXWroKDKk9K7o/cPY2URERup+7tCyEVW/KJNm7Rtalk1s3JztU1UdXAghYoWQgzPRXYnCwtChiM9Aelf0XjBtRLq1nRFk2cSmftXihg5BCCFEImJvRQmR1uSpKCGE0CMyxA/fs3OIDNH/SK4QwrAksRFCCD00kSEEPz2NJjLE0KEIIfSQxEYIIYTIAAxZBHPY+Hn0GjKZXkMmc+f+IwAOHT9Lm+7D6Np/XILbnDhzmTUbf/3oeNOLJDZCCCFEJvJhle1Yu/88ovt3pzZNcXZySHabP/T5jlXzxzPmhx6s3vALAOVLF2XTypmpC9YAsuzgYSGEEBmcWp0lBhEnZy6ejF4EM3eumCTI2NgIpTJmCmObbNbx1gsOCcVz2hIArK0s40wQmFFIj40QQuhhZG6HfbEOGJnbGTqULEP1xRefTWVv+K8IJsChY2exs7XBwd6OFXM9qVezMjt279fbhkNOO+ZPHU6hAm6M+aEHK+Z60u7bhvy29zC1qlYgj0suls/x5Ks6VXTbxBbBTO5+lqzeQvuWjRJ9ffefR6hVrQILp4/UVfvOaD6fd5UQQnwkI7Ps2BX+1tBhZClmY8cYOoRPqmSxQixds0VXBNM5V844xSljq3En5cMimF4PHxMRGUl+N5dEt/mwCGZS+/nxp18pVqQAZUp4JNHec5p9XVvXnlT3FkKITCg6MoRg33NEy1NR4iNl9CKYe/cf58WrN3Rs1TjJGFycHbn7bnDxrXsZswimJDZCCKFHVIgfvmdmEiXz2IhUqF+rEpu2/0G9mpWpUaUcz1++ps/QKRw4epqWTesnu533i2Be+veWbnlsEcz3yzEkZz/R0RpmLlqLt48vfYdPZercVQBcunqTASNn8PjpMwaMnMHL1/40bViLQ8fPMmj0TF699k/F2Ug/WbYIphBCpBUpgilE5iFjbIQQQogMRIpgpo4kNkIIIUQGIkUwU0fG2AghhB4KlQkm1q4oVCaGDkUIoYf02AghhB6m2VzJW//jp7oXQnw60mMjhBBCfGIJ1V+auWitASJJ2NvgEA4eO5Po6wnVkfrnzCW6D5xAryGTWbDip0S3Xb52K03a9Wfx6s1pGnMsSWyEEEKP8ICH3N/TgfCAjDlvh8gaRg3q/lHbJVY7KjWCg0M5fPxsoq8nVEeqYP48rFownlXzx+MfEMStuw8S3LZNiwZMHJl0CYnUkFtRQgihj1aDRh0G2rT/BSI+D9dv3WfBio2YmZpQpuR/M/vuPXCcm7e9GNa/C90GjGf90ilMmbsKpVKJ3/NX2GSzYtKofqhU8fsh2n0/gmKFC2BpaUHr5l8xZ8l6IqOiKFQgL4N7dyQ8IpLp81fz8nUAKpWSpbPGcOvuA5b+uIXo6GiqVy5Hh1aNWLPxV54+e0FgUDBh4REsmDaCnX8c4vK12/QdPpXh/buSzy13nH0nVEfKyeG/EgtGRkYoE5oxEMhhZ4u3j+/Hnkq9JLERQgiR5XW7vov9r+6nS9tPag3Xu87Jc5fp3rEFVSqWRqPRcPLsFXb+cQhvH1+G9e+C4oMkoJhHAcYO+Z7la7fyz+mL1KpWIV6bL169YdX8CWSztmTs1MUM698FF2dHZi9ez627D7h28x4eBfMzefTXul6d5Wu3MWP8YLJZWzJs/Dwa1qsGxMxQPGFEC5at2cr5S9f4pnFdnj57zvRxg1J8Pm7e8cI/IIjCBfOleNu0IImNEEKILG//q/tEaqMxUagMsv+WTeqzfvNu9h0+xVd1qhAREcm2XX+xbvHkeEkNgMe7pKBIofz4PE14xmsXZ0ddpW7vJ8+YvuBHAEJDw/mifAkePfalSYOaACiVMT0+Xg99GDV5AQBv34by/OVrAAq5uwExhTaDgj++dMiLl69ZuHITMycM/ug2UksSGyGEEJ8FE4UqWb0r6cHS0oJh/TsTFaWmS39PTE1NGNznOzynL2XGuEGYmZnGWf/u/Ud4FMzHrXsPKZJIz4dS8d/tqTwuuRjwfXtyOdqj1WqJ1mh4/uI1V67dpkih/Gg0GpRKJe758zB93ECsLC2IjtagVCo4efYyCt5LrrSgMlIRncKxOyGhYYybsYyRg7phZ2uTom3TkiQ2Qgihh4m1C3lqz8HEOvEqykIk5be9hzl28jzqaA2N6tfgwNHTlCpWiPYtv2bcjKVMGzswzvq37z1i/9Hp2Fhb0bNzS73t9+veltmL1xEZGYVSqWTs0O9p2rA2U+euos+wqboxNn27t2HU5IVoNVqMjY2YOeGHBNuzt7MlIiKSMVMW0adbG1xzxy2eeenqTdZv3q2rIzV+RG9+//sovn4vmbdsAwA9vvuWsiWLxGt7266/+evgCQKD3vLq9Rsmj+6f3NOYLFIrSgghRJbncnQOkLzxMIY2Ze4q2rf8mgJ5XQ0dSqYkPTZCCKFHVOhL3tzdhV2hFhhb5DR0OOIzExwSyoiJ8+MsG9izg24cTnpbvm4b12/d0/1csUwJurRvlqxtDVGXShIbIYTQIzoiiMAHf2PjVlcSG5Huxg3rFednK0sLls/xNFA00Ldbm4/e1hB1qSSxEUIIkebS8/Hqj2HIJ6LEpyUzDwshhEhzsY9XC/GpSY+NEEKIdGHIx6s/FDt4WGR90mMjhBB6qExtsHVvgsrUcHNzCCGSR3pshBBCD2MLexxKdjV0GEKIZJDERggh9NCow4gIfIypTR6URuaGDkdkUcdOXaBE0YIJztrr6/eCybNXolAqsTA3ZdKoflhZWtB3+FQ00RqUKiVNvqpFw3rVWLtpJ6fPXwWgZdP6NKhbLU5bJ85c5va9B/T47ttPclyfmiQ2QgihR+RbX3yOjSZP7TmYZS9g6HDER4rURqfLWJuPGUcUW+LgfcdPXcTF2THBxMbK0pI5k4dibWXJrr2H2P3nETq0agTA/GkjsDA3063boG41unf8Rle+4as6VROsR5VVSWIjhBAiy/vS3t3gj59funqTzb/+hUqlpGLZ4hw4cppojZZWzepTvIg7Zy78y0PvJ5QtVZT+PdrF2Ta22CWAsZERSmVMoqJUKBnqOQcrKwuG9O1MLkd7cudyAMDISIXqXfIUHBKK57QlAFhbWeLmmutTHLJBSGIjhBAiy1tXvIWhQwAgJCSU5XM96TVkMstmj0WpVNJn2BTq1qhEpfIl9ZZSeBscws4/DukmvpvmOQCbbNZc+vcW85dvYM6kobp1t+76m9rVK6JQKNj95xFqVatA86/rsHzt1nQ/TkOSp6KEEEKIT8SjUD78A4PweerHoDEzGTBqOsHBoQQEBundVq1WM3HWcgb16oBNNisAbLJZA1C2ZBFevQ7QrXv24jWuXr9D57ZNAXji+1xXgqFIofxpfFQZi/TYCCGEHgqlCpVJNhTKjDNzbUab2fdDMtNvwhQKJbbZrHFzycWi6aMwNjZCrVZjZGSEkUqFJlqT6LYzF66lTo0vKFW8sG5ZSEgolpYWPPR+irVVzO2q+w99WP/zLuZPG6Ebx+Pi7Mjd+4/wKJiPW/ceYmKcdX/9Z90jE0KINGJqk5cCjf9n6DDiiJ3ZV5KHzEepVNKlfXMGjp6JUqHA1saaaZ4DqVyhFAtXbqJCmeLxikxeuXabQ8fP4fv8JXv3H6dmlfK0adGA/iNnYGpiDMCw/l0AWLhyI0FvQxg2fi4AsycOoWnDWnhOW8Kh42fJYWeLs1PWrXmmiIyM0KZkA7VaTfVGXfhn7/8wMkq/vGjNjqMA9GhVK932IYQQmVXs0z0ZZWbfD2X0+ETWJT02QgihR0TQY3xPz8S58ihMs+UxdDgii/P28WXW4nVxlk0c2RcHezsDRZS5SGIjhBB6aKOjiArxQxsdZehQxGfAzdWZ5XM8DR1GpiVPRQkhhBAiy5DERgghhBBZhtyKEkIIIT6xrv3HsX7plE+2v5DQMEZOXEB0dDQqlYqxQ3uSy9GeKXNX8dD7CWZmplSpWJqOrRqz+88j7Pn7KEqlgppVy9OxVeM4bXk98mHzL38yblivTxZ/SkhiI4QQehhb5SJ31XEYW2XdaehFxpdQfankMjJSMWFkH3LmyM6ZC//y8469DOvfGYCxQ3vGme24QtliNG1YC4DeQ6fwdf3qCdavyqgksRFCCD1UxhZYOpYxdBgiE7t+6z4LVmzEzNSEMiU90Gg1zF26gVt3H1Cranm+a9OEcxevsX7LbiIiIqhVrQKd2jRl7/7jnLnwL2HhEXzTuC6rNuygSMH8eD3y0fWmBAS+ZcaCNYSEhpHDzpbxw3ujUsVNgExNTMiZwwSIW2tKQczEf+ZmZgz4vh0FC7jh7OSg285IpUKpUKKOjmbCjGUEvQ3BydH+k523jyGJjRBC6KEOe0PAw/3Y5vsSI3N55Da50quadkaTnLl6Tp67TPeOLahSsTQajYa/Dp6gQ6tGONjb0anPGL5r04SSxQqxYq4nGo2GHoMm0Lp5AyCmt2Xu5JgaUHOXbqDdtw1xze1Ev+HT+Lp+dTZt30ur5l9SvnQxNm77nWOnLlCnesUE44iKUrNm007G/NADgAE922OTzZpHj32ZMncVaxdP0q179OR5cjs7YGtjzeF/zuHi7Eifbm3Y+cchbtzOuLNeS2IjhBB6qMP9eXN7O1a5Kkhik0wZoZp2RtKySX3Wb97NvsOn+KpOFaytLMn1rufD5N3MwbfvPWTtpl2oo6N59vwV/gEx9aPer+1kbm6Km6szAO758/DM7yWPHj/l5u37rPt5FxERUTSoWzXROGYuWsu3jevhmtsJ+K/WVN48zigUEB2tQaVScs/Lm192H9AlVE98/XS1pooWzi+JjRBCiM9LRqmmnVFYWlowrH9noqLUdOnviYmxcbx1Nu3Yy4iBXcmdy4Eu/TxBG1MYQKFQ6NYJC4vA56kfLs6O3H/oQy7HnLi55qJmlfKULuEBxFQISMjaTTtxdnKgXq1KumWxtabeBAQSpVajUil58eoNsxavY8b4wZiZmQLg4uzEXS9valevyK27D9LsvKQHSWyEEEKIdPbb3sMcO3kedbSGRvVrcODo6Xjr1K5WgVGTF1IgrysW5mYJtmNtbcHWXX9z595DalYpj112G7q0a8aMhWtZs3EnAP16tI1Xwfv5i9es//k3ShYvxMWrNyhepCB9u7Vh4uwVBL0NQaPRMOD79gD8+NOvBAS+ZcLMZQCMHNiNGlXKcfDoafqPnE6e3Bl7EL3UihJCCD3C/b14fGQ4eWrPwSx7AUOHA0gtps/Vp35MPDOSHhshxGfnwMnrePu+Svb6ZgTipijMuYPXCccnHSNLvjGaskx3vGToMEQGFBwSyoiJ8+MsG9izg26MTFYniY0Q4rPj7fsKjUare+RVn3BsuKP9Kp2jEkK/5PTWWFlafNa1piSxEUJ8lpRKRbJvdWuiI1GHvcbIPAdKlUn6BpZMn8Nj1EJ8DKkVJYQQekQG+fBofz8igzLGbSghROIksRFCCCFEliGJjRBCCCGyDElshBBCiE+sa/9xn3yfk2avpGHrPuzYvV+37Or1O3QfOIHvB09i+bptQMykfcMnzKPf8Gks+XFzvHZCw8LpO3zqJ4s7pSSxEUIIITIBjUaTqu37dm9D/x7t4izbtOMPxg/vxY8LJ3DzjhcvX/vz259HqF65HMvmjCU8PIIbt71Std9PTZ6KEkIIPcyyF6DQNzsNHUaW1e36rkxdVyo5kyQauro3QM4c2eMty5cnN8Ehoaijo9FoNJiZmvD02Qualq4FQGH3vFy5dptiHgWYt2wD9x/64J7PNdXnLD1Jj40QQgiD2v/qPpHaaEOHka5iq3svmzOWbh1aEBwcSodWjVi9YAJ/HzoJoKvuvWbRJI6eOE94RCTwX3XvKhVL8/ZtKO2+bciq+eM5eeYybwIC+Wnb77Rq/iVLZ4/BPZ8rx05dSHZcNatWYPSURbTtPpwSRQpibWVJPrfcXLhyA4Dzl2/wNjiEW3cfEBgUzIq5nlSpWDrNz09akh4bIYTQI/LtU/wuLsGp3ABMrHMbOpwsyUShytLlITJKde8PLVq1ieVzPHF2ysnoyYt46P2UJg1qMXfp/xgwcga5nOzJYWfLE9/ncap7Z2SS2AghhB4adTjhb+6iUYcbOhSRSWWE6t4JUSgUWFtZolQqsbKyICQ0DDNTEzyH9gRg+oIfqfpFaQKDgjl59jIAt+4+/Ojz8ClIYiOEEEKkM0NX9wZYvm4bJ05fIlqj4emzFwzu3ZFuHVowxHMORkYq3FxzUcyjAHe9vFm4ciNKhZIGdavi7OSAs5MDlhbm9Bk6JcG2MxKp7i2E+Oyk9PtFqnunr6x0LOlNqnvrJz02QgghRBYi1b2FEEIkydjSAafygzC2dDB0KHFEaqOzRDHMSG00JgqVocPIFKS6t37yuLcQQuihMrEmW56aqEysDR2Kzpf27pIMCJEA6bERQgg91BGBBD85hZVLFYxMbQwdDgDrircwdAhpJiv0OomMQ3pshBBCD3XoK15c/RF16CtDhyKE0EMSGyGEEEJkGZLYCCGEEBnAsVMXeBMQmOjr/UdMp/43PTlx5rJu2ZF/ztFtwDi6D5ygq9r94uVrBoycQZ9hU9my86947Xg98mHK3FVpfwAZhIyxEUIIYXCZ+Qmvj5l/R6PRoFTG7Vs4fuoiLs6O2NkmPI5rwsg+7P7zcJxlG7f/zpKZozE3N6NT37F826QeP237g05tm1ChTHGGT5hHvZqVEiyAmVVJYiOEEHoojc2xcCiN0tjc0KFkSV/au2fq6t7JdenqTTb/+hcqlZKKZYtz4MhpojVaWjWrT/Ei7py58C8PvZ9QtlRR+vdoF2/7hJKTPC65CAkNQ6VSYWpiglKp5Omz5xQq4AaAe748XLt5jxpVyjFhxjKC3obg9K5GVVYliY0QQuhhYuWMS7Xxhg4jy8pKT3jpExISyvK5nvQaMplls8eiVCrpM2wKdWtUolL5krRv+TUF8romu726NSvRfeAElEolXdo3AyBfntxcuHKTmlXLc+nfWzg52Ot6g/p0a8POPw5x43bWTSQlsRFCCD202mg06giURqYoZO4YkQoehfLhHxiEz1M/Bo2ZCUBwcCgBgUEf1d6KtdvYtGoG5mZmDBg1g7o1KtGpbVNmL17Hnr+O4JjTDjs7Gx56P4lTnVsSGyGE+IxFBDzKcLWiROakUCixzWaNm0suFk0fhbGxEWq1GiMjI4xUKjTRmhS1Z2RshIW5OcbGRqhUSiIjI7HPkZ3p4wYRHa3Bc9oSypYsQlSUmrte3tSuXpFbdx+k09FlDJLYCCGEEJ9QzG2j5gwcPROlQoGtjTXTPAdSuUIpFq7cRIUyxXW3ld43dd5qLv97i+OnLvLA24dObZrS7puG9BoyGZVKSYUyxbHPkZ3T56+yafsfKBQKOrRqhKWFOTWqlOPg0dP0HzmdPLlzGeCoPx2p7i2E+OxkhereQoiESY+NEEIIkYF4+/gya/G6OMsmjuyLg72dgSLKXCSxEUIIITIQN1fnz7o6d2pJYiOEEHqY2riRv9F6VMaWhg5FCKGHJDZCiE/uwMnrePsarqCkRqNFqVQke32F0ijDVPUWQiRNakUJIT45b99XaDQpem7BoCKD/Xh6ajqRwX6GDkUIoYf02AghDEKpVBjsqcfYp6KSSxMVQojfBXIUaZM+AQmRTvbuP86XtatgbGzE3v3HyeOSixJFCyZr25mL1uL18AlarYaenVpSsVwJwsLDmTR7Jf4BQVSvVJaOrRsDsGzNVq7dvEsup5yMHfJ9uj41rY/02AghhBCZnEaT8MR+ew8cJ0qtBqDRlzWSndQAfNe6CT8unMD8qSNY+b8dAOz56yhVKpRi1fzxXLhygxev3nDPy5uXr9+wcv543FycOfzPudQfUCpIj40QQohM48HmMQTePmnoMOIoM/mY3nWu37rPghUbMTM1oUxJD/45c4niRQpy7eY9mjaoxY3b97n/wIeuHZpTp3pFNu34g1PnrhASGka/bm2pWK5EvDb37j/OmQv/EhYewTeN63L+8nVu33tEREQkowZ3JyIikntejxkydjY1q5YnJDQMj4L5qVapDItX/cyNO14YGxkxdsj35HLKGa/93LkcADAxNkKhiBmTdu3mPfp/H1Ogs2LZEly/dZ+AgCAqlo2Jr1KFkvyxL6aXyFAksRFCCJFpBN4+iTY6CoXK2NChpMjJc5fp3rEFVSqWRqPR8NfBE3Rq3QRra0uatOvPjv/Nw9jIiJGTFlCnekVaNqlPx1aNeRMQyNipSxJMbACMjFTMnTwUgLIli2BmZsqd+4/4ecdeJo3qS8ECeZg7eRgW5mas2fgrALfuPuDla39WzR/PlWu3WfvzLjyH9kw09hXrt9GqWX0A3gaHYGkRU+XeytKcoLfBBAWHkCOHLQCWFjHLDEkSGyGE0MPIPAc5S3TByDyHoUMRgEJlnKxekoykZZP6rN+8m32HT/FVnSpYW1ni6BDzfnLN7YSdbcxTd5GRUQD8degE+w+fQqFU8PpNQKLtFimUX/fvn3/Zy/lL1wFQqRIv1vrE97luuyKF87Ny/fZE1/193zHU0Roa1K0GgJWlBSGhYVhbWRIcEoaToz3R0dGEhIQBEBIaRjZrK32nI13JGBshhNDDyMyW7AWbYmRma+hQRCZlaWnBsP6d8Rzak2Vrt+pu7QBx/h3rl90HWDp7LFPHDECrTfwJwthtA4Pecu7SdVbOH8/gPt+hJWYbI5VRvPE3Ls6OukKYt+48wCW3U4Jtn7t0naMnzvFDn+90y0oULcT5yzcAOH/5OsU9CsRZdubCv5QslvxxPOlBemyEEEKP6MhgQl/8i4VDSVQmhv1rVGROv+09zLGT51FHa2hUvwYHjp5Ocv2SxQrRe+hkinm4Y25uprd9aytLsllb0nf4VIp7uOuWV69cFs9pS6hVrYJuWZFC+clhZ6srnpnYbajZi9dhZWnBwFEzMDUxYcG0ETRtWItJs1bwx75jVP2iNA45c+CQMwd22bPRe8hkHB3s6dCyUTLPSvqQIphCiE/O0J9vKYKZeV0eXxNI3oBd8XmSHhshhBBpKj2fXMqMA4fTwsGjZ9i596DuZ1ubbEz3HJhm7U+duwrf5y91PzeoW42mDWqlWfufkiQ2Qggh0lRmfXIpI6tXqxL1alVKt/Y9h/VKt7Y/NUlshBBCpLn0enIp9laUEImRp6KEEEIPhcoUU5t8KFSmhg5FCKGH9NgIIYQeptlccKs7z9BhCCGSQXpshBBCiAzg2KkLvAkITPT1/iOmU/+bnpw4c1m3bNj4efQaMpleQyZz5/4jALRaLcvXbaP/yOn0HT6ViMjIOO14PfJhytxV6XIMGYH02AghhB7hAQ/wOToK11ozMbPNr38DIfTQaDQolXH7Fo6fuoiLs6NuFuIPTRjZh91/Ho6z7Ic+35E7lwPePr4sXr2ZeVOGceTEeXJkt2HprDHpFn9GJomNEELoo9Wi1aghiRlgxaejjY7KUIOIkztI+tLVm2z+9S9UKiUVyxbnwJHTRGu0tGpWn+JF3Dlz4V8eej+hbKmi9O/RLt72OXNkj7cstlClsbERSmXMLMT/nL5Edhtr+g6fSrlSRene8RvU0dFMmLGMoLchODnap+JoMz65FSWEECLTsPGomqkfIw8JCWXm+MHsO3yKJbPGsHLeOH7ZcwDHnPZUKl+SsUN7JpjU6LNk9Rbav5vx1z8gEDs7G5bP8eSh91Ou37qv6w1aMmt0nPpSWZH02AghhMg08refbugQUsWjUD78A4PweerHoDEzAQgODiUgMOij2/zxp18pVqQAZUp4ADGFKsuXLgZAudJFeeD9hIDAIDwK5gOgaOH83Lh9P5VHknFJYiOEECJNZwuWyfkSp1Aosc1mjZtLLhZNH4WxsRFqtRojIyOMVCo00Rr9jbxn7/7jvHj1hrFDvtctK1msEHe9vPEomI97Xo+pW+MLrCwtuOvlTe3qFXUFMLMqSWyEEEIPk2wuuNVbiLGlo6FDSTcyW/Cno1Qq6dK+OQNHz0SpUGBrY800z4FUrlCKhSs3UaFMcbq0bxZvu6nzVnP531scP3WRB94+dGjZmJmL1lKkUH76Dp+Ks2NOPIf1okmDmkyZu5q/Dp7AzSUX5UoXRR0dzcGjp+k/cjp5cucywFF/OpLYCCGEHkqVKabZ8hg6jHSXVrMFZ6SBvRlJ2VJFKVuqKACVypekUvmScV6vVa1CnCrcH0qoCvc/ezfEW2ZuZhavjpSRSsX0cYM+JuxMRxIbIYTQIyr0Ba9v7SBHkVYYWzgYOhyRxXn7+DJr8bo4yyaO7IuDvZ2BIspcJLERQgg9oiPeEuR9CNv8DSSxEenOzdWZ5XM8DR1GpiWPewshhBAiy5DERgghhBBZhiQ2QgghhMgyJLERQgg9VGa2ZC/0DSozW0OHIrKIrv3HffJ9Tpq9koat+7Bj937dspmL1vL94En0GDSBcxev6ZZv3fm3roimr9+LOO2EhoXTd/jUTxZ3SsngYSFEgg6cvI6376t0aVuj0erq2mQGxuY5yFm8o6HDEJ+5hApnpkTf7m0oX7oooWHhumXftW5C7lwOBL0NYfCYWVQsV4Lb9x7y/OWrTFtEUxIbIUSCvH1fZboEJL1oosIID/DCzLYASmNzQ4eTKWS0QpXpKTlz/1y/dZ8FKzZiZmpCmZIeaLQa5i7dwK27D6hVtTzftWnCuYvXWL9lNxEREdSqVoFObZqyd/9xzlz4l7DwCL5pXJdVG3ZQpGB+vB75ULNqeTq2akxA4FtmLFhDSGgYOexsGT+8NypV/AQoqSKaJsZGKBQxn/Xjpy4SERFF/xHTyeeWm8G9v0OlUjJv2QbuP/TBPZ9rKs9Y+pLERgiRKKVSQY9WtdK83TU7jqZ5m+kpMtiXJ/+MJ0/tOZhlL2DocDI8G4+qaVaeIas4ee4y3Tu2oErF0mg0Gv46eIIOrRrhYG9Hpz5j+K5NE0oWK8SKuZ5oNBp6DJpA6+YNADAyUjF38lAA5i7dQLtvG+Ka24l+w6fxdf3qbNq+l1bNv6R86WJs3PY7x05doE71iimKb8X6bbRqVh+ANwGBqFQqls4ew9I1Wzj8z1lcnB0JDApmxVxPTp+/itcjn7Q9QWlIEhshxGdJo9EmO8Gy4gVlVfDbwYsEk3G/0FNDW8ST8rfSZtxEZi9UmR5aNqnP+s272Xf4FF/VqYK1lSW5HO0BMDGJKWNx+95D1m7ahTo6mmfPX+EfEFMY8/1q3Obmpri5OgPgnj8Pz/xe8ujxU27evs+6n3cRERFFg7pVUxTb7/uOoY7W0KBuNQCsLS0pWjhmn+VLF+PqjTsAcYpoZmSS2AghPjtuzvbpNn5IiIRYWlowrH9noqLUdOnviYlx/Jpcm3bsZcTAruTO5UCXfp6g1QLobhEBhIVF4PPUDxdnR+4/9CGXY07cXHNRs0p5Sr+r7q1Wq5Md17lL1zl64hyzJg7RLStZrCA37z6gdvWK3PPyJreTAy7Ojpw8exmAW3cfftQ5+FQksRFCfHbqVy2eovXD/b14fGQrzeuVy7K3oj6X8TCG8tvewxw7eR51tIZG9Wtw4OjpeOvUrlaBUZMXUiCvKxbmZgm2Y21twdZdf3Pn3kNqVimPXXYburRrxoyFa1mzcScA/Xq0jdPLE2v5um2cOH2JaI2Gp89eMLh3R2YvXoeVpQUDR83A1MSEBdNGUKViGU6cuUzf4VOxsbZm0qi+mJgYY2lhTp+hUxJsOyNRREZGaFOygVqtpnqjLvyz938YGaVfXhTbRZwe9/eFEPql52cws32+IwK9eXpyCrmrjsPUxs3Q4aSL2MQmLYpgivTTtf841i+dYugwMjTpsRFCCD1MbdzI//UaQ4chRLIEh4QyYuL8OMsG9uygGyOT1UliI4QQQmQSyemtsbK0+KyLaMrMw0IIoUdEoDcP/uxBRKC3oUMRQughiY0QQuih1ahRh79Bq0n+0yZCCMOQxEYIIYQQWYYkNkIIIcQnllARzJmL1hogkoS9DQ7h4LEzib6u1WpZvm6brlBmRGRkguslVHgzvcngYSGEECIDGDWo+0dtl9rimAkJDg7l8PGz1KtZKcHXj5w4T47sNnoLZSZUeDO9SWIjhBB6mFg541J9MiZWzoYORWRSHxbBjLX3wHFu3vZiWP8udBswnvVLpzBl7iqUSiV+z19hk82KSaP6JVjUst33IyhWuACWlha0bv4Vc5asJzIqikIF8jK4d0fCIyKZPn81L18HoFIpWTprDLfuPmDpj1uIjo6meuVydGjViDUbf+XpsxcEBgUTFh7Bgmkj2PnHIS5fu03f4VMZ3r8r+dxyx9n3P6cvkd3Gmr7Dp1KuVFG6d/wmweNOqPBmepPERggh9FAam2ORM2WzFWdGn1NF7rSUnEkNPyyCefLsFXb+cQhvH1+G9e8Sp2wCQDGPAowd8j3L127ln9MXqVWtQrw2X7x6w6r5E8hmbcnYqYsZ1r8LLs6OzF68nlt3H3Dt5j08CuZn8uiv0Wg0ACxfu40Z4weTzdqSYePn0bBeTH0oF2cnJoxowbI1Wzl/6RrfNK7L02fPmT5uUILH4x8QSIF8Liyf44nntCVcv3Wf4kXcU3rq0oUkNkIIoUdU2GsCvP7CtkBDjM1zGDqcdCEVudPXh0UwIyIi2bbrL9YtnhwvqYH/Ck4WKZQfn6d+Cbbp4uxINmtLALyfPGP6gh8BCA0N54vyJXj02JcmDWIS1dhbVV4PfRg1eQEAb9+G8vzlawAKucfMqO2Q046g4BC9x2NlaUH50sUAKFe6KA+8n0hiI4QQmUV0eAD+d3dinbtylk1spCJ3+vqwCKapqQmD+3yH5/SlzBg3CDMz0zjr373/CI+C+bh17yFFEpkxWKn47/ZUHpdcDPi+Pbkc7dFqtURrNDx/8Zor125TpFB+3Tgc9/x5mD5uIFaWFkRHa1AqFZw8exkF7yVXWlAZqYh+18uTkJLFCnHXyxuPgvm45/WYujW+SN0JSkOS2AghhBDpLKEimKWKFaJ9y68ZN2Mp08YOjLP+7XuP2H90OjbWVvTs3FJv+/26t2X24nVERkahVCoZO/R7mjaszdS5q+gzbKpujE3f7m0YNXkhWo0WY2MjZk74IcH27O1siYiIZMyURfTp1gbX3E5xXm/SoCZT5q7mr4MncHPJRbnSRRNsJ6HCm+lNimAKIRIkRTD/E1Pdezh5as/JstW9RcYxZe4q2rf8mgJ5XQ0dSqYkPTZCCCFEBmboopbL123j+q17up8rlilBl/bNMlSM75PERggh9FCZWpPNrS4qU2tDhyI+A+OG9Yrzs6GLWvbt1kbvOoaO8X2S2AghhB7GFg44letn6DCEEMkgJRWEEEIPTXQEEUGP0URHGDoUIYQe0mMjxGfqwMnrePu+SvR1jUaLUhl/fo3PUWTQExk8LEQmIT02QnymvH1fodGk6KFIIYTI8KTHRojPmFKpSPSR69hHsoUQIjORxEYIIYT4hLwe+bD5lz8xNjZKsKL32+AQzl68lmhl7TUbf+XIifPYZLPCwz0fA3t14J8zl/jf5t0YGanwKJiXH/p0SnFcXfuPY/3SKSneLqORxEYIYRAajTbT9ApZ8oIySiW7Dl4kBB9DhyMymI+daDKhpAYgODiUw8fPJprYAPTp2oZqlcrofi6YPw+rFozHSKVi/Iyl3Lr7gCKF8n9UXJmdjLERQnxybs72mWpgcggOnND0JwQHQ4ciMil1dDRjpy5mwMgZbN35NxDTQwKw849DdBswnv4jpnP05Hl2/nGIy9du03f4VB56P02wvdU//UKfYVO5cOUGAE4O9hipVAAYGRmhVCj47c/D7Dt8ivCISGo07oLfi1dc+vcWP/70a5Kx3n/oQ68hk+n5wyQ2bN0DwMBRM1Gr1ez5+yj9R8bUFZs6dxW+fi9Sf3LSmPTYCCE+ufpVixs6BCE+qeOnLuLi7Eifbm3Y+cchbty+r3vt8PGzLJk5CktLCzQaDYUL5OXps+dMHzcowbZaNfuKHt99yxv/QAaMmsH/lk7F2Djm1/nNO174BwRRuGA+TE1N2P7bfnLmsKVcqaJcvX4HX7+XlC5ROMlYV67fzujB3XFzdWbwmFl8Wasyhd3zcuf+I27cuo+FmRlqtZpnz1/h7JTxkn3psRFCCD0igp7gfWgoEUFPDB2KyKSe+PrpygsULRz3FlGfbm1YsGIjU+auwuepn962bLJZAWCX3Ya8rs68ePUGgBcvX7Nw5SbGDY+ZudjN1ZlHPk+5fO0Onds25er1u1y/dZ8SRQom2f4b/wDy5smNQqGgsHtenjx7Tqnihbl87TZh4eGUKenBP6cvYZ8je4rPw6cgiY0QQuihjY4gIvAhWpmgT3wkF2cn7np5A3Dr7oM4rxXI54rnsF40a1ibjdv/QGWkIlqjSbStkJBQAMLDI/B+8owcdraEhIYxbsYyRg7qhp2tDQAKhYJs1lZcu3mX0iU8eO0fQFSUGjMz0yRjzW5rw6PHT9Fqtdy5/wiXXI6ULFaIw8fPktPejlLFC7Nh6x5KFU+658dQ5FaUEEIIkc5qVCnHwaOn6T9yOnly54rz2pwl63n2/CVRUWp6dWmFvZ0tERGRjJmyiD7d2uCa2ynO+kvWbMHr4RM0Gg2d2zbFzNSEdT/vwtfvJfOWbQCgx3ffUrZkEUoVK8TVG3cByJHdFovcZnpj7d2lFdMXrEGrhaoVS5PLKScAkVFqShcvTGH3fPg89aNMCY+0ODVpThEZGZGiGbrUajXVG3Xhn73/w8go/fKi2KclPna0uRAiafo+Y/IZ/E+4v5fMPCxEJiE9NkIIIUQGtXzdNq7fuqf7uWKZEnRp3yxVbQaHhDJi4vw4ywb27KAbA5TZSWIjhBB6GFs6kqviMIwtHQ0divjM9O3WJs3btLK0YPkczzRvN6OQxEYIIfRQmVhh7VLF0GEIIZJBnooSQgg91OEB+N/bgzo8wNChCCH0kMRGCCH0UIe95uW1/6EOe23oUEQWETvr8Kc0afZKGrbuw47d+3XLZi5ay/eDJ9Fj0ATOXbymW75159/0HzmdvsOnxptdODQsnL7Dp6ZpbGs2/sqJM5fTpC25FSWEEEJkAhqNBqXy4/sj+nZvQ/nSRQkNC9ct+651E3LnciDobQiDx8yiYrkS3L73kOcvX7F01hiDxvuxJLERQojPnO+Z2YT4XTB0GJlWwebb9a5z/dZ9FqzYiJmpCWVKeqDRapi7dAO37j6gVtXyfNemCecuXmP9lt1ERERQq1oFOrVpyt79xzlz4V/CwiP4pnFdVm3YQZGC+fF65EPNquXp2KoxAYFvmbFgDSGhYeSws2X88N6oVPETipwJzBScO1dMSQQTYyMUipj6bcdPXSQiIor+I6aTzy03g3t/h0qlZN6yDdx/6IN7PtdEj3PNxl959vwV/gFB9O7amk3b/+Dla380Gg2TRvXFycGezv3GUqJIoTjHHuvZ81fMWrSWkYO6k8vRXu95TYjcihJCiM9ciN8FtBq1ocPI0k6eu0z3ji1YNmcs3Tq0IDg4lA6tGrF6wQT+PnQSgJLFCrFiridrFk3i6InzhEdEAmBkpGLu5KFUqViat29DafdtQ1bNH8/JM5d5ExDIT9t+p1XzL1k6ewzu+Vw5dirlSeqK9dto1aw+AG8CAlEoFSydPQZTUxMO/3OWW3cfEBgUzIq5nlSpWDrJthxy2jF/6nAKFXBjzA89WDHXk3bfNuS3vYcBEjx2gCe+z5m1aC1jh/b86KQGpMdGCCH0UhpbYulUHqWxpaFDSTcKpVGyeh7Ex2nZpD7rN+9m3+FTfFWnCtZWlrpf3iYmxgDcvveQtZt2oY6O1vV6ABQp9F9tKXNzU9xcnQFwz5+HZ34vefT4KTdv32fdz7uIiIiiQd2qKYrt933HUEdraFC3GgDWlpa6elblSxfj6o07AInWuvpQ0XfxRkdrWLpmK14PHxMRGUl+N5eY9hM4doCff9nLoJ4dEuxZSglJbIQQQg8TKydyV0ndeAPxebO0tGBY/85ERanp0t8TE2PjeOts2rGXEQO7kjuXA136eYI2pjBA7C0igLCwCHye+uHi7Mj9hz7kcsyJm2sualYpT+l3JQ7U6uT3vp27dJ2jJ84xa+IQ3bKSxQpy8+4DalevyD0vb3I7OeDi7MjJszGDe2/dfZhkmwpFzM2gew+8CQ4JYcW8cRz55xwn3m3//vG8b1DPDvyx/xiODjkoUTTpQp1JkcRGCCH00GrUREeFoDK2RKGUr02Rcr/tPcyxk+dRR2toVL8GB46ejrdO7WoVGDV5IQXyumJhnnBNJ2trC7bu+ps79x5Ss0p57LLb0KVdM2YsXMuajTsB6NejbZxenljL123jxOlLRGs0PH32gsG9OzJ78TqsLC0YOGoGpiYmLJg2gioVy3DizGX6Dp+KjbU1k0b1xcTEGEsLc/oMnZJg2wlxc82F3/NXDBw1EzfXXHrXNzMzZdrYAYyesph+PdpS2D1vsvbzIakVJYQBHDh5HW/fVwaNQaPRolQqpFZUMmT1WlH3fmsNJG8QrDCsrv3HsX7pFEOHkaHJnx5CGIC37ytdYiGEEGnpU9WC2rbr7zgDlQvkdWVov85puo+PIYmNEAaSVG/JpxDbI5MUjUabrPWyOiteUFYFvx28SDA+hg4nzWk1PallutrQYYhkSE5vzaeqBdWmRQPatGiQ7vtJKXncWwiRIDdne+lREkJkOtJjI4RIUP2qxQ0dQoYRM8ZmK83rlcvSY2yEyAoksRFCCD1MbfNSoMkmlEamhg5FCKGHJDZCCKGHQqFCZWxh6DCEEMkgY2yEEEKPyGBfnpyYTGSwr6FDEVnYsVMXeBMQmOjr/UdMp/43PeNUwR42fh69hkym15DJ3Ln/CACtVsvyddt01bkjIiPjtOP1yIcpc1elaexT5q7C61HGGFgvPTZCCKGHJiqM0BdX0ESFGToUkUUkVPn6+KmLuDg7Ymdrk+A2E0b2Yfefh+Ms+6HPd+TO5YC3jy+LV29m3pRhHDlxnhzZbVJVnVubwKzHmYUkNkIIIdBq1DKI+CMld2LDS1dvsvnXv1CplFQsW5wDR04TrdHSqll9ihdx58yFf3no/YSypYrSv0e7eNsnVZ3b2NhI9xTjP6cvkd3Gmr7Dp1KuVFG6d/wGdXQ0E2YsI+htCE5JFJicMncVZmam+DzxY+LIPkyYuRx1dDTGRkbMGDcQS0sL2vYYQdHC+bn/4DEdWjXiqzr/1aa6c/8RqzfsYMKIvmSzNkxtNUlshBDiM2fpVJ4Qv5RXhBYpFxISyvK5nvQaMplls8eiVCrpM2wKdWtUolL5krRv+TUF8rqmuN0lq7fQvmUjAPwDAimQz4XlczzxnLaE67fu8+LVG1ycHenTrQ07/zjEjdv3E22rsHtehvfvAsCcSUMwMzNly86/OHjsLM2+rs0b/wCG9u0EwMDRs3SJzY3bXhw/dZHJo/phaWm4MWmS2AghxGfOudIIQ4fw2fAolA//wCB8nvoxaMxMAIKDQwkIDProNn/86VeKFSlAmXdFMK0sLShfuhgA5UoX5YH3EwICg+JU504qsYmtzh0aFs6sRet48eoNQW+DqVO9IgDOTg66xEWj0cSJY8a4QQZNakAGDwshhF5GFvY4lPoeI4vEu/CFSA6FQoltNmvcXHKxaPools/x5KcV08hhZ4uRSoUmWqO/kffs3X+cF6/e0LFVY92yksUKcdfLG4B7Xo/fVed20i27dfdB0jG+u6V19sK/ODvlZMVcTxrVr6F33M2EEb1Zvm4rPk/9UnQMaU0SGyGE0MPI1AbbAg0xMk14UKcQKaFUKunSvjkDR8+k3/BpTJi5HIDKFUqxcOUm/rd5d4LbTZ23mr8OnmD1hh38tG0P0dEaZi5ai7ePL32HT2XquyedmjSoyZkL/9Jn2FQ0Gg3lShelRpVyePv40n/kdO55PU5WnMWKuHP6/FWGjpvDA+8netfPbpuNiSP6MH3BGp49N1yRX6nuLYQBZIT3d0aIIbOIjnxLiN8lLJ3KojKxNnQ4QogkyBgbIYTQIyrkBX4XFpGn9hxJbES68/bxZdbidXGWTRzZFwd7uzTdz/J127h+657u54plStClfbM03YchSGIjhBBCZCBurs6fpDp3325t0n0fhiBjbIQQQgiRZUhiI4QQQogsQxIbIYTQQ2lkhpldIZRGZoYORQihh4yxEUIIPUysc5On1kxDhyGESAbpsRFCCCFEgrr2H5fubf+0bQ++fi/SrF3psRHiM6bRaHXz2YjEWfGCsqqtXIpuSzAOhg5HZDAyF1TqdGrTNE3bk8RGiM+Um7M93r6Gmx1UiM/Npas3+Wnb75iamuD77CWd2zblj/3HCHobwoJpw9mwZQ837nhhbGTE2CHfk8spp66S9p37j+jQshH/nL7E46fPGDGwG6WKFYq3j78OnuD3fUcJCQ2jbYuGNKxXDf+AIKbOW01oWDgO9nZMGtWXU+eusO7nXZiYGNPkq1o0qFuV+ct/4sGjJyiVSsYN64lDzhyJHsve/cf558wl1OpoAoPe0qJRXf46eAKtVsvC6SNQqVQJtvfTtj38c/oSBfK6otHGlI+YMncV7Vt+TWDgW06cvcLAnu3xeuTD5l/+ZNywXnTuN5biRQpy7eY9mjaoxY3b97n/wIeuHZrr6le9TxIbIT5T9asWN3QImUa4vxePj2yleb1ymGUvYOhwRCam0WiZNeEHfvvzMAeOnWbh9JFs2/U3v/99jJev/Vk1fzxXrt1m7c+78Bzakzf+AQzr34VXr/3pN3wav2yYz2OfZ2z+9a8EE5va1SrQsF41wiMi6TVkEg3rVeOnbXto/GUNaleviEajQaPRsGLddlbO88TS0gKNRsPJs1ewtrJk2Zyx3Lh9n5+2/cGw/p2TPBYbaytG/9CDFeu2cdfLmyWzRrNw5SauXL9DeHhkvPa6dWjO6fP/snrBBLx9fBniOTdZ5yw4OJROrZtgbW1Jk3b92fG/eRgbGTFy0gJJbIQQQghDcs/nCoC9XXbc8+UBIKe9HY99nlHkXVXtIoXzs3L9diCmkraFuRn2ObLjktsJUxMTctrb8TY4JMH2z1z8l+2/7UOrhSe+zwF49NiXzm1jZhRWKpW88Q/EIaedrgq3Uqnk0eOnHDt1gSvXb6PVgmNO/bMcu+ePid8+R3bMzUxjjiVHdt6+DeHpsxfx2vP1e4l7PlcUCgV58+TG3Nw0boPvFdfUvlfsydrKEkeHmN4j19xO2NnG1GyLjIxKMC5JbIQQQohP5b1f3u8XyTY2NtJV3b515wEuuZ3erZPw+lptwmUe/7d5NyvmeaJAwbddhgCQN48zV67dpla1Cmg0GmxtrHn56g2hYeFYmJuh0Whwc81F3Rpf0K1DCyCmLmTKDuu9pAQSbC8wKBivRz5otVoeP3lGWFhEnDayWVny8tUbAO4/8E6w7cQqi79PEhshhNDDJJsreb9chpF54mMOhEitHHa29BoyGZVKiefQnh/VRq2qFeg9dAqF3fNibWUJxAzOnTpvFdt+26cbY9OrSysGjJyBmakJjb+qSYO6Vblw5Sb9hk9DoVDwZZ0qNG1Q66OPpVqlsgm2V7FsCb4fPJHCBfNhbW0RZ5sC+VwJj4hg4KiZ5M/r8tH7lureQhiAvL+FECJ9SI+NEELoERXynFc3t2BftB3Glo6GDkcI4NNV5z549Aw79x7U/Wxrk43pngPTfD9pRRIbIYTQIzoymLc+x8nu3kQSG5FhfKrq3PVqVaJerUqfZF9pQWYeFkIIIUSWIT024pM4cPK6TAb3Ho1Gi1Kpf3S/EEKIlJEeG/FJePu+QqNJ0Th1IYTIstKzBlNiXrx8zfAJ8+g3fBo//vQrAL/vO0bvIZPpMWgCy9dti7eN1yMfpsxd9alDTRXpsRGfjFKpkKeA3pH6TJmLkVl27DxaY2SW3dChiM+YRqNBqfz4/oila7YwfEBXHOz/m3yvQZ2qNPmqJgB9h0/lxcvXSZZSyAwksRFCCD2MzO2wL9rW0GGITOz6rfssWLERM1MTypT0QKPVMHfpBm7dfUCtquX5rk0Tzl28xvotu4mIiKBWtQp0atOUvfuPc+bCv4SFR/BN47qs2rCDIgXz4/XIh5pVy9OxVWMCAt8yY8EaQkLDyGFny/jhvVGp4iZAarWaZ89fsWT1ZvwDgujZuSUlixXC2DgmDVBHR2NtZUk2ayvU0dFMmLGMoLchODnaG+J0pYokNkIIoUd0VCjhb+5gZlcYlbGF/g1EmunUcyh/Hzxm6DCS9OLBBb3rnDx3me4dW1ClYmk0Gg1/HTxBh1aNcLC3o1OfMXzXpgklixVixVxPNBoNPQZNoHXzBgAYGamYO3koAHOXbqDdtw1xze1Ev+HT+Lp+dTZt30ur5l9SvnQxNm77nWOnLsSroRQQ+Ja7Xt5MGdMfYyMjhk+Yx7olUwD4adseftt7hIrlimNmZsrhf87h4uxIn25t2PnHIW7cvp/GZyx9SWIjhBB6RAU/4+nJKeSpPQeVFMH8pP4+eIzIyChMTIwNHUqqtGxSn/Wbd7Pv8Cm+qlMFaytLcr3rDYk9ttv3HrJ20y7U0dE8e/4K/4AgAF0NKQBzc1PcXJ2BmFpNz/xe8ujxU27evs+6n3cRERFFg7pV4+3fysoSF2dHnBxi9qlSGaGOjsZIpaJTm6Z0aNmYUZMXcP3WfZ74+uFRMB8ARQvnl8RGCCGESEsmJsbJ6hXJyCwtLRjWvzNRUWq69PfExDh+orZpx15GDOxK7lwOdOnnqasE+X59pLCwCHye+uHi7Mj9hz7kcsyJm2sualYpT+kSHkDCdZ7MTE2wyWbF2+AQjIxUREVFYaRS6ZJGlUqJuZkpZqYmuDg7cdfLm9rVK+rqV2UmktgIIYQQ6ey3vYc5dvI86mgNjerX4MDR0/HWqV2tAqMmL6RAXlcszM0SbMfa2oKtu/7mzr2H1KxSHrvsNnRp14wZC9eyZuNOAPr1aBunlydW7y6tGT5+HlHqaHp89y0Qcxvq0r+3iI7WULZUEdzz5yGvW24OHj1N/5HTyZM7VxqehU9DakWJT0KuZ1xyPjKXcH8vHh8ZTp7aczCTW1GflEP+8kDyxrF8Drr2H8f6pVMMHUaGJj02Qgihh0JljLGlEwpV5h7nIT4PwSGhjJg4P86ygT076MbNZHWS2AghhB6m2fKQ76vln2RfmeEpoE8pKwwcTkvJ6a2xsrRg+RzPTxBNxiQzDwshRAYS+xSQEOLjSI+NEELoERH4iCf/TMSl+kRMbfKm+/6ywlNAaSV2jI0QySU9NkIIoYdWE010ZBBaTbShQxFC6CGJjRBCCCGyDLkVJYQQQmQAx05doETRgtjZ2iT4utcjH5as3kxUlJq6Nb7gmyb1WL/5N85dvEZEZCQN61WnVbMv42xz4sxlbt97oJu35nMgiY0QQgjxiSVUqfv4qYu4ODsmmtisWLedaWMHYGn5X72yjq0a07V9c9TR0XzXezTfNK4XrwDm50YSGyGE0MPE2hnXmjMwsXY2dCifpcjIqAw9iDi5A70vXb3J5l//QqVSUrFscQ4cOU20RkurZvUpXsSdMxf+5aH3E8qWKkr/Hu3ibPv02QvUajUTZq1ArVYzuPd35M3jrKvOHRUVRe5cjqhUSoJDQvGctgQAaytL3Fwz3+zBqfF5p3VCCJEMSiNzzHMURmlkbuhQPjsN6tXMUvPYhISEMnP8YPYdPsWSWWNYOW8cv+w5gGNOeyqVL8nYoT3jJTUAb/wD8Xrow8SRfRjQsz2LV/+se23Bio207jqM4h7uAOz+8wi1qlVg4fSRukKbnxPpsRFCCD2iQl/hf/93srs3wdji8/tFYUg/rZ5n6BDSlEehfPgHBuHz1I9BY2YCEBwcSkBgUJLbWVtZULhgPqwsLbCytIiz/g99vqNv9zb0HjKFxl/V4Invc5p9XRuIqQzu9cgn/Q4oA5LERgiRZR04eR1v31epbseKF5RV/c7hO5YE45AGkSVu4pQZTBw3Ol33IQxHoVBim80aN5dcLJo+CmNjI9RqNUZGRhipVGiiNQlu55LbicCgt6jVat74B2JpETPOJnZmZhNjY8xMTTAxMcHF2ZG79x/hUTAft+49xMT48/pV/3kdrRDis+Lt+wqNRotSqTB0KELoKJVKurRvzsDRM1EqFNjaWDPNcyCVK5Ri4cpNVChTnC7tm8XZxkilokPLRvQbMR2NRsPg3t8BMbehvJ/4EhUVTYO6VchmbUnThrXwnLaEQ8fPksPOFmennIY4TIORxEYIkaUplYpUV1GPqe69leb1yqV7de+MPEhWpE7ZUkUpW6ooAJXKl6RS+ZJxXq9VrQK1qlVIdPuEXh85qFu89aytLFk0Y1QaRJw5SWIjhBBCZCDePr7MWrwuzrKJI/viYG9noIgyF0lshBBCD5VpNmzyN0Blms3QoYjPgJur82ddnTu1JLERQgg9jC1y4li6p6HDEEIkg8xjI4QQemjUEYT7e6FRRxg6FCGEHpLYCCGEHpFvn/D4yHAi3z4xdChCCD0ksRFCCCFEliGJjRBCCCES1LX/uHRZ932//Xn4o7ZLjAweFkKIDCajF30UcSW3CKaIT6PRsPvPIzT/uk6atSmJjRBC6KNQxhTAVKR/J3eDejX5++CxdN+P+PQuXb3JT9t+x9TUBN9nL+nctil/7D9G0NsQFkwbzoYte7hxxwtjIyPGDvmeXE45adtjBEUL5+fO/Ud0aNmIf05f4vHTZ4wY2I1SxQrF28dfB0/w+76jhISG0bZFQxrWq4Z/QBBT560mNCwcB3s7Jo3qy6lzV1j38y5MTIxp8lUtGtStyvzlP/Hg0ROUSiXjhvXEIWeORI/l+KmL/G/LbszNTKlb4wu+aVKPsPAIJs9Zyf0Hj+nQqhFf1anK/Yc+zFmyHq1WS9UvytC5bVPWbPyVZ89f4R8QRKnihXn85Bl9h0+lW4cWlC9dLNXnWRIbIQxEo9GyZsdRQ4eRpaVVOQUz23y4N/1Z/4ppIKsVfRRxaTRaZk34gd/+PMyBY6dZOH0k23b9ze9/H+Pla39WzR/PlWu3WfvzLjyH9uSNfwDD+nfh1Wt/+g2fxi8b5vPY5xmbf/0rwcSmdrUKNKxXjfCISHoNmUTDetX4adseGn9Zg9rVK6LRaNBoNKxYt52V8zyxtLRAo9Fw8uwVrK0sWTZnLDdu3+enbX8wrH/nRI/jyIlzjBvWi3xuudFoYupbvfEPYGjfTgAMHD2Lr+pUZeX67Ywe3B03V2cGj5nFl7UqA+CQ045xw3oBcPTE+TSdt0cSGyEMwM3ZPk2KMwohMhf3fK4A2Ntlxz1fHgBy2tvx2OcZRQrlB6BI4fysXL8dAGcnByzMzbDPkR2X3E6YmpiQ096Ot8EhCbZ/5uK/bP9tH1otPPF9DsCjx750bhtTe0qpVPLGPxCHnHZYWlrolj16/JRjpy5w5fpttFpwzJn0LMdd2zdn8y9/EhEZybdN6lO8iDvOTg66Nt9PdvLmyQ1AYfe8PHkWE1PRd8eaHiSxEcIA6lctbugQPgtp1SMWEeTDs7NzyfXFMEyzuaZJm+IzpVAk9E+MjY24dfcBALfuPMAlt9O7dRJeX6vVJtj8/zbvZsU8TxQo+LbLEADy5nHmyrXb1KpWAY1Gg62NNS9fvSE0LBwLczM0Gg1urrmoW+MLunVoAYBarU7yMBxy5mDU4O68fO3PpFkrWDp7TJxYY2W3teHR46e4uTpz5/4jWjSqy9Xrd1C8d1s3gc1SRRIbIYTQQxsdSeRbH7TRkYYORWRhOexs6TVkMiqVEs+hHzfTda2qFeg9dAqF3fNibWUJQKc2TZk6bxXbftunG2PTq0srBoycgZmpCY2/qkmDulW5cOUm/YZPQ6FQ8GWdKjRtUCvR/azbtJNrt+6jVqtp2fTLRNfr3aUV0xesQauFqhVLkyuBSuN5XHIxevIi2n7bMMHbaymliIyMSDjtS4RaraZ6oy78s/d/GBmlX14U+5dWaqvyioxBrqcwhLR638VU9x5Ontpz0r26txAidaTHRgghhMiElq/bxvVb93Q/VyxTgi7tm6X5fg4ePcPOvQd1P9vaZGO658A0309akcRGCCGEyIT6dmvzSfZTr1Yl6tWq9En2lRZk5mEhhNDD2NIJ50qjMLZ0MnQoQgg9pMdGCCH0UJlYYuVc0dBhCCGSQXpshBBCD3W4P2/u/Io63N/QoYgs7pnfS85evJYubc9ctBaImQH58ZNnANz18mbn7weT2ixZDhw9Tf8R0+k7fCrXbsaM+xk9eRF9hk2l+8AJXL52O942i1dv5tLVm6ne94ekx0YIIfRQh73h1Y2fsXAojZFZdkOHI7KwZ89jEpsvypVI87ZHDeoOwKV/b+FRMD95XHJRqIAbhQq4pardl6/9+ef0RZbMGh1nLpvJo/thbGzEM7+XTF+whiWzRqdqP8kliY0QQgjxCVy6epMNW/dgbGzMG/8AxgzpyYuXr+PUbDp59jLXbt7j9r0HzBg3GJtsVnHa6Dt8KgXyunL/wWMKubvxQ59OhISEMnH2CoJDwrC3s2X88N7cuf+IBSs2YmZqQpmSHvT47lu69h/Hinnj2HvgH46cOM+h42dp8lUNTpy9Qv1alfj972OMGNgVrVbL94MnsnrBBM5evMaGLbuJ1mhp1aw+X9auEu+4zlz4F2NjYwaOmol9DluGD+iKhbkZxsYxKUZIWDj587oAcM/LmxkL15Dd1ga1Wk21L0qn+XmWxEYIkaWlRU0uK15QVgW/HbxIMD5pE5jIMlIyT1J4RCQLp4/E28eXZWu34vf8dZyaTY457XDImYOBPdsn2ka1SmUZ2q8z46Yv5c69h1y4cpPKFUrzTeO6rPt5FweOnsbH14/uHVtQpWJpXXkDADNTExrVr45HwfxUq1RGdyuoSKH8zF++EXV0NDdve1HMwx2FQsH6zb+xbPZYlEolfYZNoW6NSqhUcUexvPEPJCDwLYtnjmLnHwf5Zc9+OrVpCkCfoVN4/NSP8cN7A7Bqwy9MGNEH19xO9BoyJdnnLSVkjI0QIstyc7ZPkyKYQqSVQgXcUCgU5M2TG6+HT+LVbEoOj4J5gZhkxOfpc574Pqdo4fz/LfP1o2WT+pw6d5UJM5dz5sK/yWq3fOmiXLxyk0PHz1C/VmX8A4PweerHoDEzGTBqOsHBoQQEBsXbztrSgnKliqJQKChfuhgPvZ/qXlsxbxxrF09i2dqtQEztKDdXZ5RKpe440pr02Aghsqy0qskVGeLHq+vefFO8BibyyLdIhXte3mi1Wh4/eUaBfC48f/E6Ts0mlZFRnB6WhNy5703FssW5fe8BZUsV4fnL19y87YVHwXzcuvsAV2cnLC0tGNa/M1FRarr096RKxdK67Y1UCe+jXq1KbP7lT574PueHPp1iaki55GLR9FEYGxuhVqsTrDhQslghfv5l77vje4yzkwNarZbo6GiMjIwwNzPDwswUiKkd5fPUDxdnR+7cf0TtahVScTYTJomNEELoYWLphPMXww0dhsgCLC0tGDZ+Hv4BgYz54Xuev3wdp2ZTjcplWbluG2OmLmb04O66ek/vO3P+Kut/3oV7/jx4FMyHi7MjE2Yu58CxM9jZZuO71k349feDHDt5HnW0hkb1a8TZvlzpoixfu5WLV29Qs0p53fICeV255+VNpfIlgZgepC7tmzNw9EyUCgW2NtZMS2DGYff8eXCwt6Pv8KmYGBszcWRfIqOi+GHsbAA00Rp6d20NQM/OLZkwcxnZbbMleGxpQWpFiU9CrqfIzLSaKNThgRiZ2aBQGhs6HJFJXbp6kxNnryQ5fkafvsOnMnfyMCzMzdIwsqxFemyEEEKPiMDHUgRTfHJT567C9/lL3c8N6lYzYDQxMkPdKElshBBCiE+gbKmilC1VNNnrew7rFW9Z0wa10jCilMsMdaPkqSghhBBCZBmS2AghhBAiy5DERgghhBBZhiQ2Qgihh6ltPtybbcPUNp+hQxFZRNf+4z7p/kJCw+g/Yjp9hk6h/4jpPHv+CoDtv+2jRafBjJmyKMHtduzez979xz9lqKkmiY0QQuihUChRqoxRKOQrUxiOvon7kmJkpGLCyD6smDeOjq0b8/OOmAn16tWqxJKZn6Y45aciT0UJIYQekW99eX55BY5l+mBi7WzocEQmdP3W/ThFKTVaDXOXbuDW3QfUqlqe79o04dzFa6zfspuIiAhqVatApzZN2bv/OGcu/EtYeATfNK7Lqg07KFIwP16PfKhZtTwdWzUmIPAtMxasISQ0jBzvimB+WM/J1MSEnDlMADA2MtKVGrGztSEiPDLOus9fvGbCzGWYm5thYmxMjSrlPs1JSiOS2AghhB4adRhhr26gUYcZOhQBhE+bTvTZs4YOQ8dyz26965w8dzlOUcq/Dp6gQ6tGONjb0anPGL5r04SSxQqxYq4nGo2GHoMm0Lp5AyCmt2Xu5KEAzF26gXbfNsQ1txP9hk/j6/rV2bR9L62af0n50sXYuO13jp26QJ3qFROMIypKzZpNOxnzQ49EY924/Q+6dmjBF+VKMG760o84I4YliY0QQohMJfrsWVCrIR1nv09rLZvUZ/3m3ew7fIqv6lTB2sqSXI72AJiYxMxmffveQ9Zu2oU6Oppnz1/hHxBTcLJIofy6dszNTXFzjek1dM+fh2d+L3n0+Ck3b99n3c+7iIiIokHdqonGMXPRWr5tXA/X3InXPHvi6xen0GZmk3neFUIIIUQsI6Nk9ZRkFB8WpTQxjl+aY9OOvYwY2JXcuRzo0s8TtDEVjxSK/yrUh4VF6IpI3n/oQy7HnLi55qJmlfKULuEBxJQ+SsjaTTtxdnLQO8Gei7NTnEKbX5Qr+bGHbRCS2AghhBDp7Le9h+MUpTxw9HS8dWpXq8CoyQspkNc10VpQ1tYWbN31N3fuPaRmlfLYZbehS7tmzFi4ljUbdwLQr0fbeD0tz1+8Zv3Pv1GyeCEuXr1B8SIF6dutDQeOnuaXPQfweerHgJEzWDRjJB1bN2LCzOVs/mUvlhbmaX8y0pkUwRSfhFxPkZlFRwQR7HsWK+cvUJlmM3Q4n72Qps2A5I1tyWq69h/H+qVTDB1GhiY9NkIIoYfKNBs2+eobOgwhkiU4JJQRE+fHWTawZwc8Cn4e8zBJYiOEEHpIj43IKJLTW2NlacHyOZ6fIJqMSRIbIYTQIyr0Jc8vr8DUNn+GT2wy2qPQ6SKTPRElPi2ZRlMIIbIQ3aPQQnymJOUVQoisJpM9Cp1SsYOHhUiI9NgIIYQQIsuQxEYIIfRQGpljbl8MpVHmm9NDiM+NJDZCCKGHibUzrjWmSAFMka6OnbrAm4DABF/z9XtB7yGT6TNsKkPHzSE4JBSA5Wu30qRdfxav3pzgdotXb+bS1ZvpFnNGJGNshBBCD61Wg1YTjUKpQqGQvwczBLU6w4y1+ZjxTBqNBqUy7nvp+KmLuDg7YmdrE299K0tL5kweirWVJbv2HmL3n0fo0KoRbVo04ItyJTl57srHhp/lSGIjhBB6RAQ85PGR4eSpPQez7AUMHc5nT/XFF5nykfZLV2+y+de/UKmUVCxbnANHThOt0dKqWX2KF3HnzIV/eej9hLKlitK/R7s422azttT929jICKUypn5UDjtbvH1846x7z8ubGQvXkN3WBrVaTbUvSqf7sWUkktgIIYTIVMzGjjF0CB8tJCSU5XM96TVkMstmj0WpVNJn2BTq1qhEpfIlad/yawrkdU10+7fBIez84xALpo1IdJ1VG35hwog+uOZ2oteQz6/8giQ2QgghxCfiUSgf/oFB+Dz1Y9CYmQAEB4cSEBikd1u1Ws3EWcsZ1KsDNtmsEl3vjX8Abq4x48E8CuZNk7gzE0lshBDiI2XIWX5lVt4MTaFQYpvNGjeXXCyaPgpjYyPUajVGRkYYqVRoojWJbjtz4Vrq1PiCUsULJ7mP7LY2+Dz1w8XZkTv3H1G7WoW0PowMTd79QgjxkXSz/EoiIVJAqVTSpX1zBo6eiVKhwNbGmmmeA6lcoRQLV26iQpnidGkfd2D0lWu3OXT8HL7PX7J3/3FqVilPmxYN2Lbrb/46eILAoLe8ev2GyaP707NzSybMXEZ222xYW1kmEkXWpYiMjNCmZAO1Wk31Rl34Z+//MErHD/OaHUcB6NGqVrrtQ3w6cj1FZqbVRKEOD8TIzAaF0li3PPapnIw0y29GjEmIT0n+zBBCCD0USmOMLewNHYb4THj7+DJr8bo4yyaO7IuDvZ2BIspcJLERQgg9IkP8eHV9I/bFv8PE0snQ4Ygszs3VmeVzPA0dRqYlM00JIYQemsgQgp+eRhMZYuhQhBB6SGIjhBBCiCxDEhshhBBCZBmS2AghhBBZWNf+4wDYu/84127eA2LmxOkzbCr3H/rE+XdWIIOHhRBCDyNzO+yLdcDIXJ5KEWkvoYKY6aHRlzV0/77r9Yh1S6bE+3dWIImNEELoYWSWHbvC3xo6jOTLQJWvPwfJmTPojX8g46YvJTo6GrvsNvTt3pZp81aTw86WggXcKO7hzor12wD4pnE9vq5fnfsPfZizZD1arZaqX5Shc9umrNn4K0+fvSAwKJiw8AgWTBuBmalJvP39dfAE23/bh2tuR0LDwgFYs/FXPArm58r123j7PKPv8KkUds+r+3fsTMiZXeY/AiGESGfRkSGEvbqBuX0xVCYZeybXzFr5OquztrJk0cxRGKlULFjxExev3OTla38WzxyNsbERPQZNYO7kYVhZmvP94EnUqfEFK9dvZ/Tg7ri5OjN4zCy+rFUZABdnJyaMaMGyNVs5f+ka1SuXi7Ov6GgNW3b+xZqFEwkNC+ebToPjvN6/RzsuXrmpe6T8yrU7WerxcklshBBCj6gQP3zPzCRP7Tn8n737DIyi3OIw/mxLb4R0EkLvvXdCUxRBURAERRABKSIdpPfeO0gRQZCigldEERWQpjQRpPdASEgI6XXL/RBZiUncQMruJuf35WZnp5wZuObwzjvzV9mUNnc5/8mak68LsuiYWOYt+5SYuHgiHj3GxdmJMqWKG0dI9Ho9bq7OAPj7eRPx6DGRj6MoUbwYAOXLlODegzAAypUJBMDL052YuIyvIIiKjsHLwx0bGw02Nhp8fTzz4xQthkweFkIIIfLY/l+O07h+DVbNH0+DOtXw8fJAqVAYv1colERFx6LVagm+H4pH0SIUcXPl9t37GAwGrly/jb+vd9q6/LMdmYQiubm68DAiktRULdExcTwIDc/r07MoMmIjhBBC5LE6NSszZe4qjpw4i20mc2I+6NWZ4RPmo1BApw5tsLO14YOenZm5aB0GAzSuVyPbIy8qlZIuHV+k79ApBAb44u1VNLdPx6JJCKbIF/LnKaxZ0uMb3P1lJMVbzMOuyD+3oiRwUgjLIyM2QghhgkJlg41zAApVxn9pC2FuQ8fNJTklxfi5e6d2NK5f04wVmZc0NkIIYYKtSwAl2iwxdxlCZGrRjFHmLsGiyORhIYQQQhQY0tgIIYQJSVG3uP5Nd5Kibpm7FCGECdLYCCGEKQY9em0iGPTmrkQIYYLMsRFCWJwfj17gTkiEucswcuIhtVSw+8Bp4ngqKLDrIN76Yrn5ChPChL37D/NCi0ZoNGr27j9McX9fqlYqm61tp81fw60797Czs6VRvRq83fkVEpOSmDJ3NY+jYmjaoBZvv/lKHp/Bs5PGRghhce6ERKDXG1AqFaZXFkJkGaS598fDtGhaD41GnS4EM7vGDe9L6RIBxs/f7DtIo7rV6fBSi7SYh5aN8PKwrHBYaWyEEBZJqVRYzHuP0t5j8wWvta6d6XtsRPYlzZhZ4LKssvMeowuXrrNo1WbsbG2oWa0Cv544Q5WKZTl/8Rod2gbx1+XrXL8ZTK/ur9GyaT227PyWY7//QXxCIgPf60q92lUz7HPv/sOcOPUniUnJvP5KK06evcDla7dJTk5hzJDeJCencO3GXYaNm0vzxnWIT0ikQtlSNGlQk6VrPuevKzfQqNWMG9Yn05f/KYDZi9djb2fHh33eomzpQM5fvMagPm8BUK9WVS5cuk7LpvVyfA1zkzQ2Qghhgo2zP8VbzMPG2d/cpVg93W+/gVYLefiCV0t09Pez9H67I43q1UCv17PvwBF6vNkeZ2dH2r81iJ2fLkCjVjN6yiJaNq1Hp/ZteLvzK0RGRTNu+rJMGxsAtVrF/KnDAahVrSJ2drZcuX6bz3fuZcqYAZQtXZz5U0fgYG/Hus1fAnDp6k3CHz1mzcKJ/HH+Mus//5rxw/tm2PeHfbvh6uLM7bshTJu/hvVLpxAbF4+jgz0ATo72xMTG5dEVe36F62+WEEI8B6XaNt1IjcghtbrQva25U/s2bNy6hx9+PsaLLRvh7ORojDoIKOaDu5srACkpqQDs++kI+38+hkKp4FFkVJb7rViulPHnz3ft5eSZCwCoVKost7kXEmbcrmL5UqzeuCPT9Vxd0kI5SxT3Q6FISw13cnQgPiERZydH4uIT8fH2yOYVyD/yVJQQQpiQmhBO2B9rSU0oXGGCIvc4OjowYtC7jB/elxXrv0CRLgAz41yyXXt+ZPnccUwf+yEGQ9bJR0+2jY6J5fczF1i9cCJD+r+D4e90TLVKjV6f/mk+fz9vLl29CcClKzfxL+aT6b7j4xMAiIyKJlWrRaVSUrVSOU6e/QuAk2cvUKWC5TX8MmIjhBAm6JJjiL75Pa6BrdA4ZC+IUIin7d77M4eOnkSr09OuTTN+PHj8P9evVrkcHwyfSuUKZbC3tzO5f2cnR1ycHRkwcjpVKpQxLm/asBbjZywjqEld47KK5UpR1N2NfsOmolIpM70NBTB57ipiYuPR6/V82KcbAB1eCmLKnFV8+8MhGtevgZen5QVsSgimyBfy5ymehaX9fZEQzNwj10zkNRmxEUIIISzcgYMn+GrvAeNnN1cXZo4fnGv7nz5/DSFh/9xqbduqCR3aBuXa/vOTNDZCCFGImP1x60L4RFRuaB3UgNZBDfJs/+NH9Muzfec3mTwshBAmqGxdcSvTHpWtq7lLyTHj49ZCFFDSNgshhAkaBw+8qvUydxm5x4yPW8tLDUVekxEbIYQwQa9NJPHRlbQgTCGERZPGRgghTEiJDSH40MekxIaYuxQhhAnS2AghhBBW6MiJs8aYhH8bMHI6CYlJHD95joNHT+ZzZeYlc2yEEELkL622QM21seR38jSsW93cJeQ7i2hsfjx6gTshEemW6fUGlMqMr5kWQghhvVT16xe4dO/sOnPuIp9t/x+2tjaEPAjn3a4d+Hb/IWJi41k0YyTjZyxHq9OhUauZNWEwjo4OfPvDIXZ/9zM2Nhp6dn2VShVKM37GMiDtbcOBAb7/ecy9+w+TkJhE51dfoOv7o6hUvhTXb96le+d2vNiyMfcfPGTeso2kpKZSrnQJhnzwdn5cijxlEY3NnZAIaWSEEBZLoVShsnFBocw6WFBkj924seYuwaz0egNzJg1l93c/8+Oh4yyeOZrtX3/P4eNnmDdlGHZ2tmz7ah8HDv1Gs0a12bPvF1bNn4BGk5b5tO3LfQQ1qctrL7dk5fovnunYkY+jGD6gBwCDP57Diy0bs3L9F4wY1BN/P2/mLt3Ipas30wVrWiOLaGwAlEpFutenP3mlurXJbPRJyAicsG62riUo/cqn5i5DFABlSgYA4OFehDIliwPg6eHOrTv3mLV4PQ8jIomJjaNl03qEhD6kfJmSaDRpv6qVSiX3QsJ49eUWQFrm043bwdk+tp+PF46ODgDGYMw79x4wc9EnACQkJFG/TlVpbER6MvokhBAiS+lSvf9ZbKPR4OfjyZQxA9i66zsSEhMp5uvNleu30Wq1qNVpIzb+ft5cvX6bCmVLcunaLWw02f81nlmKeHF/Xz7s0w1fbw8MBgO6fyWBWyNpbPLAv0efhPWOwAkBkBxzl5Djs/FrOAZbl+LmLkcUQCmpqRw/eY6rN25TxM0VH6+iuLk60+GlIPoOnYq9nS3vdu1Ah5eCGD9jGT8d/o2i7m74+eQsbX5g767MXbqBlJRUlEol44b3wcfLI5fOyjyksRFCCBMMulRS40Mx6FLNXYqwYrWqV6JW9UoANGlQkyYNagLQsmk9aFqP3m+/nmGb9i82p/2LzdMtWzJrjMljrZw3HoB2LzQzLtu4fFqGn4v5erFoxqhnPBPLJo2NEELkhLU9uiwhlAVOQUrmzg3yt1sIIZ5TYX50WViOgpTMnRuksRFCiOdkjY8uW9XokhDPQSIVhBDCBI2TL8UaT0Dj9N8vQxNCmJ80NkIIYYJK44Cjd01UGgdzlyIKiF6DJpjluA/CImj2Sk/j+29mL1lPnyFTeP+jSfx++nyG9Xfu2c/e/Yfzu8wckVtRQghhgjYxkqhb+3Er+QJqe3dzlyMKKb1ej1KZs/GILTu/pVqlcsbP77zZnmK+XsTExjNk7Bzq1a6a0zLNThobIUSO5MXbti3tJZfapMdEXt6Bk29daWzEc7lw6TqLVm3GztaGmtUqoDfomb98E5eu3iSocR3e6dKe30+fZ+O2PSQnJxPUpC49unRg7/7DnDj1J4lJybz+SivWbNpJxbJpbxxu3rgOb3d+hajoWGYtWkd8QiJF3d2YOPIDVKqMDVBI6EMUKPD2KmpcVszXCwAbjdr4Ar+wh4+YNHsF9vZ22Gg0NGtUO38uUi6RxkYIkSPytm2Rn5JmzLS4J9Gyk+599Pez9H67I43q1UCv17PvwBG6d26Hl4c7PfqP5Z0u7alWuRyr5o9Hr9fz/keTePO1tgCo1SrmTx0OwPzlm3jrjZcIKObDwJEzeLlNU7bs2Evn116gTo3KbN7+Pw4dO5X2bpx/2bz9W3p0ac+6LV9l+G7Vxu10frVN2no7vqVX947Ur12VCTOX5+TSmIU0NkKIHMvtt23Lm6pFVnS//WaV7+Lp1L4NG7fu4Yefj/Fiy0Y4Ozni6532hl8bGw0Al6/dYv2Wr9HqdDwIi+BxVAxAuuwme3tbAgP8AChTqjgPQsO5ffc+Fy9fZ8PnX5OcnErbVo0zHP9eSBgAvpm8qfh/PxxCq9PTtlWTv9cNpULZEhmObS2s62+GEEIIoVZna5TEkjg6OjBi0LukpmrpOWg8NhpNhnW27NzLqMG9KObrRc+B48FgANJnPCUmJhN8PxR/P2+u3wrG19uTwABfmjeqQ42qFQDQarUZ9n395l1u3bnHkLFzuHH7HvdCwlg6ewznLlzl4JHfmTN5mHFdfz8frly/Q71aVbh87Sb1a1fL7cuRp6SxEUIIE1Q2TjgHNENl42TuUoSV2r33Zw4dPYlWp6ddm2b8ePB4hnVaNKnLmKmLKV0iAAd7u0z34+zswBdff8+Va7do3qgO7kVc6fnWq8xavJ51m9NuMQ18v2uGkZagJnUJalIXgGnz19Ct08vY2tgwd+kGnBwdGDxmFrY2NiyaMYq332zHpNkr2bprL44O9rl8JfKeNDZCCGGCxtEb37pDzF2GsGJdX29L19fbGj936/Sy8ecnuU3tXmiWLtsJoN2/bh0pFUpGDuqZbpmrizOzJw7Jdi0TnnpT8a5PF2b43sfLgzULJ2Z7f5ZGGhshhDBBr0tBm/gItX1RlCobc5cjxH+Ki09g1OT0Dcvgvt2pULakmSrKX9LYCCGECSkxwdz9ZSTFW8zDrkhpc5eTc9YW3Pk0K5w4nJueTujOipOjgzHduzCSNw8LIUQhoqpfv1A3BqLgk7/dQghRiFhjcOfTrHakSeQbGbERQgghRIEhjY0QQgghCgxpbIQQwgS7IqUp9/pXBWPisLBYh46dIjIq+j/XOX/xGg1ffJuExCQARkxcQL9hU+k3bCpXrt/OsP7StVs5c+5iXpRrsWSOjRBCCJHPMkvqPnzsNP5+3ri7uWa53c49P6R7bHto/3co5uvFneAQlq7dyoJpI/KsZmshjY0QQpiQEnuf0NPL8Kn9ITbOxcxdjrCwx9WzG+9w5txFtn65D5VKSb1aVfjxl+Po9AY6v9qGKhXLcOLUn9y6c49a1Ssx6P23Mmx/7sIVSpcsTkRklHHZk3RujUZtDKK9duMOsxavo4ibK1qtlib1a+T4HK2J3IoSQggT9NokkiKvotcmmbuUQs/aH1ePj09g9sQh/PDzMZbNGcvqBRPY9c2PeHt60KBONcYN75tpUwOwffcPdOrQJtPvlq3dRrdO7QBYs2kXk0b1Z96UYSQkJufZuVgq6/3bIYQQotCx9sfVK5QryePoGILvh/LR2NkAxMUlEBUd85/bnfnzEmVLFc80u+mTz76kcsXS1Pw7BDPycZQxAfxJSndhIo2NEEIIkU8UCiVuLs4E+vuyZOYYNBo1Wq0WtVqNWqVCr9Nnut31m3c5dfYvzl24wvWbwUybt4ZZEz9i7/7DPIyIZNywPsZ1i7i5GhPAr1y/TYu/wy8LC2lshBBCiHykVCrp2e01Bn88G6VCgZurMzPGD6Zh3eosXr2FujWr0LNb+jlEb772Im++9iIAA0ZOZ8LIfuh0emYvWU/FcqUYMHI6ft6ejB/Rj77vdmLS7BUUcXPB2cnRHKdoVtLYCCGECRpHL3zqfITG0cvcpQgrVqt6JWpVrwRAgzrVaFCnWrrvg5rUJSgboytP50D9undThu/LlynBhmWmM6UKKmlshBDCBJWNMy7Fm5u7DFFI3AkOYc7SDemWTR49AC8PdzNVZF2ksRFCCBO0ydHE3TuGk38j1LZZv2NEiNwQGOBXqNO5c0oe9xZCCBO0CRE8PPcJ2oQIc5cihDBBGhshhBBCFBjS2AghhBCiwJDGRgghhBAFhkweFkIIE5Qaexy8aqDUZHzrqxDP40FoOMs+2crMCR+Zu5RnduN2MMvWbiU1VUurZvV5vX1rNm7dze+nz5OcksJLrZvS+dUX0m1z5MRZLl+7yfvvvJHn9UljI4SwSHq9gXU7D5q7jKc0gwdXgavmLkRYmPc7B5m7hBzLLG08K6s27GDGuA9xdHQwLnu78yv06vYaWp2Odz74mNdfaY1KZZ6bQtLYCCEsTqCfB3dCLOkJJD0qUtGhQe7gi+eh0+mZNn81YeGRONjb0vfdzoQ/esyYqYt5EBbOh326USrQnxkLP2HBtBEAfDh6FrMmfoTTUw0EwIIVm7hxO5iUlFS0Wh3Tx33IlLmrsdGoCfD3YcxHvfnr8g2Wrv0clUpJo3o1eLvzKyxd8zl/XbmBRq1m3LA++Pp48lafUVQuXxpHRwfefO1F5i3bSEpqKuVKl2DIB29nOI/7Dx6i1WqZNGcVWq2WIR+8Q4nifmg0ae1EamoqxXy9UamUxMUnMH7GMgCcnRwJDPDN46ucRhobIQQAPx698FzNhF5vQKlU5GotbRpXydX95VTS4xvc/WUkxVvMw65IaXOXI6zQoWOn8PJwZ/LoAew7cIQjJ87yKDKaFXPHkZCYxMhJC/hk8WT0ej3RMbEkJ6fg6GifoakBGD7wXQAmz1lJ21ZNOPPnJdq2asQb7dug16dlTS1Zs4VpHw/C26soer2eS1dvEv7oMWsWTuSP85dZ//nXjB/el4cRkaxZOAkXZ0fGTV/KiEE98ffzZu7SjVy6epOK5UqlO3bk42hu3Apm27q5hIU/Yunaz1k4fSQAi1Zt5ufDv/FG+7QE8j3f/UJQk7q89nJLVq7/Ii8vbzryTw8hBAB3QiLQ6w3mLkOIAuleSKixSahYrhT3QkIpVcIfGxsNbq7O6HQ6AFo0qcsvR07y06+/06pZ/Sz3t37L11SuUIYGdarRqll9QkLDmTR7Jd//dBQArVaLt1dRIC2b6l5I2D/HL1+Ke/dDAfD388bFOS1P6s69B8xc9AkDRk7n4pUbPIyIzHBcZycHypctiZOjA6VLBKRLJR/a/x12bVrIoWOniHj0mHshYVQoW9J4zvlFRmyEEEZKpeKZ5wtY1jwYISyTv58PF6/cpEXTely6ehN/Px+++/EwqalaEhKTUKlUAAQ1qcfEWcvR6fTMnTI00339dPg3Ih9H0fvtjgCoVCo+7NMNgG59RtO2VWM0Gg0PIyLx8nBHr9fj7+fN4WOnAbh05Sb+xXwAUCr+Gd8o7u/Lh3264evtgcFgQKfPmDTuX8yH6JhYtFotkY+jcXRIG1FKSUnFxkaDjUaDna0NNjY2+Pt5c/X6bSqULcmla7ew0eRPyyGNjRBCCJHHmjWqzcEjJ+k/fBr29nb0e7cTp85eYOKsFYSEPWRg77cAcHF2TGsOnG2xt7PLdF+rNmynqLsbA0ZOx9/Pm3o1q7Lrm/0A1K9TFaVSyeC+3Rk/Yxlqtco4x6aouxv9hk1FpVIyfnjfDPsd2Lsrc5duICUlFaVSybjhffDx8ki3jlqlonundgwcNRO9Xs+QD94B0m5D3bkXQmqqjratGuHi7EiHl4IYP2MZPx3+jaLubvj5eObmJc2SIiUl+ZnGnrVaLU3b9eTXvZ+iVudOX/TkX3xP/0sxs2XWwFrrzmtyXSzf8/4ZFYY/W5ljI4T1kBEbIYQwwdY1kFLtNqLSOJq7FFHIbP/6ew4dO2X8XLpEgHHycF6z1pRxaWyEEMIEhVItqd7CLLp0bEuXjm3NcmxrTRmXp6KEEMKElLhQ7h+bSUpcqLlLEUKYII2NEEKYoE+NJz70FPrUeHOXIoQwQRobIYQQIh/duB3MtPlrmL1kfabfx8bFc+DQiSy3X7f5S7r3G8OAkdNZuuZzAH49cYbegyfRb9hUFq367Lnq6jVoQoZln23/hpDQh8+0jbnJHBshhBDCDMZ81DvT5XFxCfx8+DdaN2+Q5bb9e3WhSYOaxs9lSxVnzaKJqFUqJs5anulbg59Hjy4dcryP/CaNjRBCCJHHtDodk2atICY2Hh/vtHfD9Bo0gY3Lp/HVtz/x7Q+HcLC3o9Orbfjr0g3Onr/MgJHTGTmoFyUDi2XY39rPdvH5rr30frsjdWpUTve+GbVajVKhYPd3P2NvZ0fzxnV44Y2+7Ngwn5DQcE7/cZE+PTKmbCcmJTN13mqu37xL987teLFlY6bNX0O3Ti8TGODHxJnLiY1LoHiAL0lJyUwY0Q+9Qc/85Zu4dPUmQY3r8E6X9nl3EbNJGhshhDBBbV8Uz6o9UdsXNXcphd57F75mf8R1c5eRzr2gkSbXOXzsNP5+3vR/rwtfffsTf13+5xx+Pvwby2aPwdHRAb1eT/nSJbj/IIyZEz7KdF+dX32R9995g8jH0Xw4ZhafLp9uDKG8eOUGj6NiKF+2JLa2NuzYvR/Pom7Url6JcxeuEBIaTo2q5TPdb+TjKIYP6AHA4I/n8GLLxunqDyjmQ//3urD7u585f/EakDa61L1zO7w83OnRf6xFNDYyx0YIIUxQ27lRpGwH1HZu5i6l0NsfcZ0Ug87cZTyzeyGhxtykSuXT3yLq/14XFq3azLT5awi+b/rJO1cXJwDci7hSIsDPmOn0MPwRi1dvYcLIfkDa49q3g+9z9vwV3u3agXMXrnLh0nWqViyb6X79fLxwdHQwNlhZ1f/kfyEttdvX2wOVSomNjSY7lyLPyYiNEEKYoEuJI+Hhnzh4VUNl42Tucgo9G4UqW6MklsTfz4erN+4Ys6KeVrpkAONH9OPPv66yece39H23U6Y5TU/Exyfg6OhAUlIyd+49oKi7G/EJiUyYtYLRH72Hu1vaO5cUCgUuzk6cv3iV3m93ZNtX+0hN1WJnZ5vpfhUKRbbqv3r9dra2MRdpbIQQwoTU+DAe/D6f4i3mSWMjnkuzRrU5cPA4g0bPpHgx33TfzVu2kQdh4aSmaunXszMe7m4kJ6cwdtoS+r/XhYC/AyufWLZuGzdu3UOv1/Nu1w7Y2dqw4fOvCQkNZ8GKTQC8/84b1KpWkeqVy3Hur6sAFC3ihkOxzPOnnqX+Yj5eqP8O7bREkhWVy6y17rwm18XySVZU1iQrynL4H5wHZG9ei8hdWq0WtVrN7u9+JjY23iLm02RGRmyEEEIIC7Vyw3YuXLpm/FyvZlV6dns1R/uMi09g1OSF6ZYN7ts93dyZzIyavIjExCQ0Gg3Txg7KUQ15SRobIYQQwkINeK9Lru/TydHhuTKgFk63jlEyaWyEEMIEhcoWW9eSKFSZT7oUGeXVY9kpBh02Csud3yHMTxobIYQwwdbFn8BWC8xdhlV58li2NCEiv0ljI4QQIk/kxWPZTyYPC5EVeUGfEEKYkBR1k2u73yQp6qbplYUQZiWNjRBCmGIwYNBrwfBMb8cQQpiBNDZCCCFEATF22hIehIanW3bkxFnWbf4y0/V37P6Bjj2GMHbaEuOyPd/9Qu/Bk+gzZDJbdn5rXH7m3EUGjZ7JwJEzOHj0ZIZ99Ro0IZfOImdkjo0QQgirkmLQWdRcG2t+WWDroAY0qleDleu/MC6rW6syHV4KAuCD4dN4uU1THOzt2frlPhZNH2UM3LRUll2dEEII8ZQXPMpYXLp3dp05d5HPtv8PW1sbQh6E827XDny7/xAxsfEsmjGS8TOWo9Xp0KjVzJowmKTkFMZNX8qS2WPY/8txQsMi6NPjjQz7/f3MBVas34avtyePIqOAtJfwjZ+xDEgLqgwM8M2wHYC7myvJSSnplvn5eBl/VqtUKBVKLly6hq2NhpGTFmBra8OoD3tR1N2NfQeOsGP3DwQU8yYhMSmXrlTOSGMjhBAm2Lj4E9h6MRpHb3OXUuhtqNLR3CXkiF5vYM6koez+7md+PHScxTNHs/3r7zl8/AzzpgzDzs6WbV/t48Ch33j15RZ06vACsxatIzziMUtmjc50n2s37WTprI+xt7OlS+8RQNrtpKAmdXnt5ZbpRmOexcGjJynm54WbqzORj6O59yCMdYuncPLsBdZt/ooRg3qy7at9rFs8mYTEJF7vMeR5L0uukjk2QghhglJli61LcZTygj6RQ2VKBgDg4V6EMiWLA+Dp4c7D8EfMWrye/iOm8+0Ph4iIfAxAUNO6nD53ifZtm2eZz6jX63F1ccLGRkPpv/d5LyTMGJFQsVypZ67z2o077NrzI8P69wDSRn2qVSqHRqOmTo3K3Lpzj6joGLw83LGx0eDm6oyvj+czHycvyIiNECLH9HqDMQyzILIlhuKK37lrqEcyLuYuxyqM1ddipvcZc5dheRSKzH7ERqPBz8eTKWMGsHXXdyQkJgKwefv/6PLai+ze+zPNGtXBwT5jOrdSqSQmNh47Wxtu3A4GwN/Pm6vXb1OhbEkuXbuFzTPMi3kYEcmcpRuYNXEIdnZpzXzFcqXYuus7DAYD127ewc/XCzdXFx5GRJKaqiUhMSnDpGVzkcZGCJEjgX4e3AmJMHcZeUpDEr7KizzQVZPGRuSJlNRUjp88x9Ubtyni5oqPV1Gu3bjD5au3mDN5KCUDi7Fs7VZGf/Rehm379OjEh2Nm4uvtiY9nUQA6vBTE+BnL+OnwbxR1d8Mvi9GUHw8eZ9c3PxJ8P5QPR89iyazRfPLZl0RFxzJp9goARg9+j8AAP5o3rsOAEdNBoWDcsD6oVEq6dHyRvkOnEBjgi7dX0by7QM9AkZKS/EwvZtBqtTRt15Nf936a5bDYs3ryL733Owf95zJrYK115zW5LpZP/oyylvT4Bnd/GUnxFvOwK1La3OVYhSdPLVnzE0PCOsmIjRBCCGElps9fQ0jYP7d82rZqQoe2QXm+rTWRxkYIIYSwEuNH9DPLttZEnooSQggTVHZuFCn3Oio7N3OXIoQwQUZshBDCBI19UTyrvG3uMoQQ2SAjNkIIYYI+NZGE8AvoUxPNXYoQwgRpbIQQwoSUuBDu/TqRlLgQc5cixH961hDMlJRUZi9Zz6BRMxk1aSGQ9nK+PkOm0H/EdEZMXEBiUsaoBEsJvMyM3IoSQgghCqmde/bTqG4Nmn1U27isZGAxPlk8CYB1m7/k0NFTtG3VxFwlPjNpbIQQQuQJS0vhzivZfVePJYZgnjj1J5FR0Xzx9T7aBDWkY7tW6d5Rl5ycSnH/tG0/2/4Nvx4/Q+kSAegN+me5RPlKbkUJIYTIdS94lMFGoTJ3GRbnSQjmGx1aG0MwX2zZyBiCuWr+eBrWq86BQ2lvDH4Sgvn9gSP06vZqpvt8EoI5dcxAIh6lZUw9CcFcPHM0vt4eWdbzMOIRlcqXZunsj9n/y3Eehj8C4PjJc/QYMI4zf17C38+HyMfRHD/5J2sXTaLr622JjU3I/YuTS2TERgghTFAo1ajt3FEo5T+Z2WXtKdx5JasQzFt37jFr8XoeRkQSExtHy6b1gLQQzCVrP2dg7y4mQzCBdCGYr77cAkjLeXqSIfVvTo6O1KlRCbVKRdVKZblz7wFenkVpWLc6DetWZ8uOb9n93U/UqlaJMiUDUCgUlCheDHt7yw2ElREbIYQwwdY1kFIvr8PWNdDcpQhrZyIEc9X88bRr0wyDIS3t6OkQzITEjJN44Z8QzJSU1AwhmACXrt3Kspxqlcty9cYdAK7fvIufjxcpKanG7x0d7bGztcXX24Mbt4MxGAzcCQ4hMTH5uU4/P8g/P4QQQggzM1cI5jtvtmfa/DV88tmX1KtVhWK+Xhw+dpptX+1DoQBXZ2cmjuqHvZ0d9WpVpc+QyZQvWxJnZ4c8vR45ISGYucxa685rcl0sn/wZZS05+g73j06jWOMJMmojhIWTERshhDDBoNeiTYrEoNeauxRRyEkIpmnS2AghhBBWQkIwTZPJw0IIIYQoMKSxEUIIIUSBIY2NEEKYYOPkh3/Tqdg4+Zm7FCGECTLHRgghTFBq7HHwrGLuMoR4Lru/+5nXXm4JpMUitG7eAD8fr2xtO2LiAmLj4gEYNqAH5cuUwGAwsGrjDi5euYFer2fRjFHY2tjkWf3PShobIYQwITXxEVE39uFW+iU09kXNXY4QmdLr9SiVGW/E7PnuF2Nj06NLh2fa59D+71DM14s7wSEsXbuVBdNG8MuRkxQt4sryOWNzpe7cJo2NEEKYoEuK4vHVr3Au1lAaGzO5uXUs0ZePmruMTNWcesjkOgaDgYUrP+Pm7XsolUomjOjL4I/nUKl8Ka5cv033Tu349fgZ7t5/wKjB71G9cjmWrNnC5Wu3SU5OYcyQ3pQrnfEdSus2f8mDsAgeR8XwQa832bLjW8IfPUav1zNlzAAuX7vF3XsPGDByOq++1JLfz5ynW6eXKRFQjGnzVxMWHomDvS2TRg3Axdkxw/6L+aaN7Gg0apTKtFcl/3r8DEVcnRkwcjq1q1ei99uv5/AK5i5pbIQQQli86MtHMehSUag05i7luRz97Q+cnRxZMW8cf12+zmfbvyXycRQjBvUk4tFjBo6cwa5NC7kb/ICtX+6jeuVy9Hu3M3Z2tly5fpvPd+5lypgBme7by9OdCX8/yj126PvY2dly8OhJdu/9mQ96vcmmbd+wct54AH4/cx6AQ8dO4eXhzuTRA9h34Ag79+yn99tZ53stW7uNbp3aAfA4KprSJf1ZOW8842cs48Kl61SpWCY3L1eOSGMjhBDCKihUmmyNjlii23fvc+jYKf64cBmDAbw93fHz8cLB3g6PokXwL+aDrY0Nnh7uxjktn+/ay8kzFwBQqbJOSq9UrhQAOp2e5eu+4MatuySnpFAq0D/Lbe6FhFLx7+0qlitlbHgy88lnX1K5YmlqVq0AgJOjA3VqVAagdo1K3Lxzz6IaG3kqSgghhMhjgQG+tGpWn5XzxrNq/njGD++LIotATIPBQHRMLL+fucDqhRMZ0v8dDGSdfqRQpP0qv3bzDnHx8axaMIF33mxv3OLpfT/h7+fDxSs3Abh09Sb+fj6Z7nvv/sM8jIjk7c6vGJdVq1zOGJx57cZdimVzInJ+kREbIYQwQWXrjEtgK1S2zuYuRVipJg1qceqPiwwcOQOFQsELLRv95/rOTo64ODsyYOR0qlTI3mhIYIAvoWERDB4zm8AAX+PyWtUrMWrSQtq92My4rFmj2hw8cpL+w6dhb2/HpFH9M+xPp9Mze8l6KpYrxYCR0/Hz9mT8iH60b9ucafPXsu/AEQL9faldo1I2r0L+kBDMXGatdec1uS6WT/6MhCU7O7E5kL2JuqJwkxEbIYQwQa9LJjU+DI2jN0qVrbnLEYXU9q+/59CxU8bPpUsEMHzgu7m2/6Hj5pKckmL83L1TOxrXr5lr+88v0tgIIYQJKTH3uPvLSIq3mIddkdLmLsfi5cWj2db8RFRu6dKxLV06ts2z/S+aMSrP9p2fZPKwEEKIXPXk0WwhzEFGbIQQQuS63H40+8kcGyFMkREbIYQQQhQY0tgIIYQpCgUKpTrzF4IIISyKNDZCCGGCnVspyr62Azu3UuYuRRRQx0+e4+DRk5l+t2P3D3TsMYSx05Zk+G7Okg2ZLi/MZI6NEEIIq2DQpVrkXJvnmUv07yTuhnWrZ7lu66AGNKpXg5Xrv0i3/EFoOBGRj9Hk0jvlCgq5GkIIYUJyzD1CTy7Cp+5QbF2yzt8Rece1QmOLTffOrjPnLrL1y32oVEoOHztN9crlUKqUTBjejzN/XiIhMYnOr76QYTt3N1eSk1IyLN+y81veeuNldu3Znx/lWw1pbIQQwgSDLpnk6FsYdMnmLqXQKtVtprlLyBXx8Qksm/Mx/UfMYNWCCSgUCvR6/TPv5/6DhwD4ennkdolWT+bYCCGEEPmkQrmSqNVqOnVozZS5q1i06jOSkjOOxpiyefv/6N65XR5UaP1kxEYIIYTIJwqFEp1OT6tm9XmxZWM+3baHg0cynzT8Xx6EhTN36UaSU1K5e+8B3+z7hQ4vtciDiq2PNDZCCCFEPkpITGTU5EUoFKBAweQxA/j99Pks1//x4HF2ffMjwfdD+XD0LJbMGs2SWWOAtAnEyz7ZKk3NU6SxEUIIEzSO3vjWG4HG0dvcpQgrVqt6JWpVrwTAqvnj033X7oVmWW7XJqghbYIaZvqdr48nMyd8lHtFFgDS2AghhAkqGyec/RuZuwxRCEyfv4aQsHDj57atmtChbZD5CrJC0tgIIYQJ2qQoYoMP4xzQDLWdm7nLEQXY+BH9zF2C1ZOnooQQwgRt4iPCz3+KNvGRuUsRQpggIzYi3+j1BtbtPGjuMkQW9HoDSqVkIQkhrJuM2Ih8EejnIb80hRCF2plzF1m6dmu6Zb0GTci1/T8IDee3p56uys6+fz1xht6DJ9Fv2FQWrfrMuHz719/Td+gURk5aQHx8Qq7VmB9kxEbkizaNq5i7BGGCjKYJYd0ehKU1NvVrV832NmVLFWfNoomoVSomzlrOpas38fX25NcTZ1izcCL7fznGrv8d4N2uHfKw8twljY0QQpig1Dji6FMHpcbR3KUIK3fj1l1GTFxA5OMoxg7ri96gZ/bi9dy4HUzzxnV4u/MrGbY5c+4im774Bo1GY9yuTMkAjv3+Bxs+/xobGw3tXwzi6G9nOX/xGpev3WTWhCEkJCYxcdZygu+H0fX1l3ixZcYn+3yeimRQq9UoFQouXb1JzaoVUSgUNKhTjanz1uTpNclt0tgIIYQJNk4+FGs01txlCDO6uXVsliGcz5LunZScwuKZo7kTHMKK9V8QG5vAW2+8REAxHwaOnMHLbZri7uZqcrs5k4ayasMOVi8Yj6OjA3q9Hm9Pd7w8izK4bzcAwiMiWbtoEvZ2trz/0SRaN2+ASpX5DJSLV27wOCqG8mVL8sPPx3B0sAfA0dGBmNj4bJ+fJZA5NkIIYYJBr0WbHI1BrzV3KcJMoi8fxaBLzfF+ypUORKFQUKJ4MR5FRmFvb0tggB9KpZIypYrzIDQ8W9tFRcfi5emOo6MDAEplxl/nvj6euLo4YWOjwcuzKNExsZnu+2H4Ixav3sKEkWmPmjs7ORCfkAikhXa6OFvXSKWM2AghhAnJ0Xe4+8tIireYh12R0uYuR5iJQqV5ptGZzFy7cQeDwcDdew8o6u7GjVv3CL4fir+fN9dvBePr7Zmt7dxcnQmPiCQhMQkHezv0ej0qtTpdUnhoWAQxsfHY2drwMPwRri7OGfYbn5DIhFkrGP3Re8aRoorlSrHty30AnDh1nmqVy+XonPObNDZCCCFEPnF0dGDExAU8jopm7NA+TFuwhi++/p4r127RvFEd3ItkvA2V2XZKpZJ+PTvz4ehZ2Nna8MqLzWnWsBarN2xn7PSlfDykN16e7ixa9Rm374bQrdPLmd6G2v7194SEhrNgxSYA3n/nDWpVq0ij+jXoO3QKzk6OTBkzIE+vSW6TxkYIIYTIB09nRT2xacWMbG0bGOBnnDvzROP6NWlcv2a6ZasW/POI97ZP5prc73vdO/Je944Zlr/1+ku89fpL2arN0khjI4QQQliIM39eYt3mL9Mt69XttVzZ94GDJ/hq7wHjZzdXF2aOH5wr+7Yk0tgIIYQo0P7riabsMuhSUag0uVRR1mpVq8jKeeMzLK9bM+fvAmsd1IDWQQ1yvB9LJ42NEEKYYOtWgtLtt6BU25q7FPEcnjzRlB+NiTA/aWyEEMIEhUKFSuNg7jJEDuT0iaazE5vnYjUiL8l7bIQQwoSUuBDuHZlKSlyIuUsRQpggIzZCCGGCPjWRhId/oE9NNHcp+SKnc1Lkto8wJxmxEUIIkU5uvWVXPJvZS9ZnujyrBO5Bo2bS5vW+HDlxNr9KtAoyYiOEECKDnMxJkfkoz2fMR70zXZ5ZAnfFcqWYNLo/e777OZ+rtHzS2AghhBDZYNClZtq0ZacBNBgMzF+xiRs376JSqRj4fldWbtiOXq+nQpmSDO7XnV6DJrBx+bQM22aWwA3gWbRIDs6m4JLGRgghTFA7eOBVvQ9qBw/TK4sCybVC4xzNOzpy4gxKhYLVCycCMGbqYga9/xYVypZMl+/0X55O4BZZk8ZGCCFMUNu64lbaOl8vL3JHqW4zc7T97bsh1KxW0fj5YfgjKvzdoGSWzP1vTxK4Z08akqM6CgOZPCyEECboUmKJuXsIXUqsuUsR2XRz61jOTmzO2YnNLWIidInifvxx/rLxs5dnUa5cuwVgcsQmswRukTUZsRFCCBNS4x8SemoJxVvMQ2XjbO5yRDZY2tuGmzSoxYlTf9Jv2FTUf8+xWbp2KwYMxjk2WckqgXv6grWc/fMSh4+d5uadYHp06ZBfp2PRpLERQghRID15sssSntJSKBSM/LBXumUr5o1L9zmzicOQdQL3+OF9c6/AAkQaGyGEEMJCFJYE7rwkjY0QQghhIQpLAndeksnDQghhglJth517OZRqO3OXIoQwQUZshBDCBBvnYhQPmm3uMoQQ2SAjNkIIIYQoMKSxEUIIE5Ie3+DqV6+T9PiGuUsRQpggt6KEEEIIMzt+8hzJKSkENa6b4bvZS9Zz994DAP66fIP/bV2Oi7NjfpdoNaSxeQ4/Hr3AnZCITL/T6w0olYp8rkgIIcR/ySrAMjc8Twq6Xq9PF6XQsG71LNd9kvr9IDScGQs/kabGBGlsnsOdkAhpYIQQwkrkNMAyt5w5d5GtX+5DpVJy+Nhpqlcuh1KlZMLwfpz58xIJiUl0fvWFLLf/6dffaNm0Xj5WbJ2ksXlOSqWC9zsHZVi+bufBfK9FCCEKq5tbx2batDwdp5DTAMvcFB+fwLI5H9N/xAxWLZiAQqHIdrr3oaOnmDN5aB5XaP2ksRFCCBNsXAIo8cIK1PZFzV2K+BdLy4QypUK5kqjVajp1aM2UuatwdXGiX883TW73IDQcW1sbCcHMBmlshBDCBKXKBhsnX3OXIbLwJBPqaZaQD5UZhUKJTqenVbP6vNiyMZ9u28PBIydNbvfzr7/Tsmn9fKjQ+snj3kIIYUJqfBgPTi4mNT7M3KWIAiAhMZEPx8xmwMjpnDxzgbq1qpjc5uDRU7RokvGJKZGRjNgIIYQJupQ4YoMPU6RMezSO3uYuR1ipWtUrUat6JQBWzR+f7rt2LzT7z20/WTwpz+oqaKSxEUIIISzE9PlrCAkLN35u26oJHdoGma8gKySNjRBCCGEhxo/oZ+4SrJ7MsRFCCCFEgSGNjRBCmKC2K4J7hTdR2xUxdylCCBPkVpQQQpigtnfHo1JXc5chhMgGGbERQggTdKkJxIedRZeaYO5SRAE2e8n6TJffvfeAHv3H0vyVXiQkJqX77vzFazR88e0MywszaWyEEMKE1LgH3D86jdS4B+YuRRRgT8Iu/83Lw52V88dTuWLpDN/t3PMDFcqWzOvSrIrcihJCCFEg5WWi99Oyk+5tMBiYv2ITN27eRaVSMfD9rqzcsB29Xk+FMiUZ3K87vQZNYOPyaRm2tbOzzXSf5y5coXTJ4kRERuX0FAoUaWyEEELkuvxqKrLKibKURO8njpw4g1KhYPXCiQCMmbqYQe+/RYWyJbMdgvlv23f/wLhhffjt9J+5WarVk8ZGCCFyKKuEaWuV01BJS2gqLCnRG+D23RBqVqto/Pww/JHxFpJS+eyzQs78eYmypYrj6GCfazUWFNLYCCGECQqVBo2jT5a/7K0tYTqv5WdTYalhl/9WorgfJ8/+Rcum9QDw8izKlWu3KP/3iM2zNjfXb97l1Nm/OHfhCtdvBjNt3hpmTfwoL0q3OtLYCCGECbYuxSn54sr/XCezhGlrZS3NgjVp0qAWJ079Sb9hU1H/Pcdm6dqtGDAY59hkJSY2nnHTl3L95l1GTlzA22++wpuvvcibr70IwICR05kwUt5Y/IQ0NkIIIUQeUygUjPywV7plK+aNS/c5s4nDAC7Ojiyb83GW+145b3yW3xVG0tgIIYQJydG3uffrZPybTsbWtYS5yxEF2IGDJ/hq7wHjZzdXF2aOH2zGiqyPNDZCCGGCQa9DlxKDQa8zdymigGsd1IDWQQ3MXYZVkxf0CSGEEKLAkMZGCCGEEAWGNDZCCCGEKDCksRFCCBNsnP0IaD4LG2c/c5ciCojd3/38zNv0GjTB5Dpnzl1k6dqtz1NSgSGNjRBCmKBU22NftDxKtbzlVeSOPd/9Yu4SCix5KkoIIUxITYjg8fX/UaRMezQOHuYuR/xLfuVSZSU7L2a8cOk6i1Ztxs7WhpjYeEJCHzJg5HRefaklpUsGMG/ZRgwGA43r1+Tdrh14HBXD9AVrSUhMwsvDnSljBqA36Jm/fBOXrt4kqHEd3unSPsvjJaekMHXual59qQX1alfNzdO1eNLYCCGECbrkaKKu/w+XgGbS2FgYS8ilyo6jv5+l99sdaVSvBnq9nt6DJxlfrDdi4gI+HtKbwAA/hoydwwtBDdmx5wdeeaEZLZrWM4ZkxsUl0L1zO7w83OnRf2yWjU1ycjITZ62gS8e21Hoqn6qwkMZGCCGE1bK0sMusdGrfho1b9/DDz8d4sWWjdN9FPo6iRPFiAJQvU4J7D8K4fTeEd7u+CvwTkuns5Iivd1pjbWOTdS7Z4eNnaNqgVqFsakDm2AghhBB5ztHRgRGD3mX88L6sWP8FCsU/3xVxc+X23fsYDAauXL+Nv683JYr78cf5ywDGERvF0xv9hzZBDbGx0bBj9w+5fh7WQEZshBBCiDy2e+/PHDp6Eq1OT7s2zYiMimbUpIW0e7EZH/TszMxF6zAYoHG9Gvj6eNKjSwemL1jD9t0/GOfYPIshH7zNnCUb+N8Ph2j/YuEKNZXGRgghTFDZuuBaqi0qWxdzlyKsVNfX29L19bZZfr920aR0n4u4ubBg2sh0y54OycwqMLNW9UrUql4JgNEfvfe85Vo1aWyEEMIEjYMn3jX6mrsMIdI58+cl1m3+Mt2yZbPHolIV7lkm0tgIIYQJem0yKbH3sHH2R6m2NXc5QgBQq1pF45NV4h+Fu60TQohsSIm9x91fRpISe8/cpQghTJDGRgghhBAFhjQ2QgghhCgwpLERQggh8ll2Ai1z25S5q3npzf7s3LPfuOz23RD6D59GnyFTOHn2QoZtdu7Zz979h/OzzByTycNCCGGKQpkWgKmQfwsK89Hr9ca3ED+PAb27UKdGJRISk4zLVn+6g7HD+uDu5sLQ8fOoW7NKbpRqVtLYCCGECXZuJSnT4XNzlyH+dnPrWIvKh3rWEMya1SpkGmj5++nzbNy2h+TkZIKa1KVHlw7s3X+YE6f+JDEpmddfacWaTTupWLYUN24H07xxHd7u/ApR0bHMWrSO+IREirq7MXHkB5k+8u1ZtEiGZRGPHhNQzAcAF2cnoqJjSU5OYdLsFdjb22Gj0dCsUe2cX6R8JP/8EEIIYVWiLx/FoEs1dxnP5EkI5op543ive0djoOXaRZP4/qe0Jq1a5XKsmj+edUumcPDISZKSUwBQq1XMnzqcRvVqEBubwFtvvMSahRM5euIskVHRfLb9f3R+7QWWzx1LmZIBHDp2Ktt16fUG489OjvbExMaxece39OrekUUzRv1nJpWlkhEbIYQwITkmmAe/zce3/ghsXQLMXY4AFCpNtkZKLMW/QzAzC7S8fO0W67d8jVan40FYBI+jYgCoWK6UcT/29rYEBvgBUKZUcR6EhnP77n0uXr7Ohs+/Jjk5lbatGme7LqXyn/ypuPhEXJyduBcSSoWyJTIc21pIYyOEECYYdCmkxAZj0KWYuxRhpZ6EYKamauk5aDw2mowjIVt27mXU4F4U8/Wi58DxYEgbTXk6/DIxMZng+6H4+3lz/VYwvt6eBAb40rxRHWpUrQCAVqvNdl1F3d24FxJGETcXYmLjcHN1xt/PhyvX71CvVhUuX7tJ/drVcnj2+UsaGyGEECKP/TsE88eDxzOs06JJXcZMXUzpEgE42Ntluh9nZwe++Pp7rly7RfNGdXAv4krPt15l1uL1rNv8FQAD3++a6UjLyg3bOXL8DDq9nvsPHjLkg7f5oOebTJ+/Bp3eQJ93Xgfg7TfbMWn2Srbu2oujg30uXoX8IY2NEMJIrzewbudBc5dhcZx4SC0V7D5wmjiCM3xvqDieOpemm6EyYS3+HYLZrdPLxp+fBFq2e6EZ7V5olm67dj6e6T4rFUpGDuqZbpmrizOzJw4xWcOA97ow4L0u6ZaVDCzG6oUT0y3z8fJgzb+WWRNpbIQQAAT6eXAnJMLcZQghciguPoFRkxemWza4b3cqlC1pporylzQ2QggA2jS2/vdX5BVdSjyJESUo7VEZlY1jhu/PTmxuhqrylkGXarHnZdClolBZ39M6ueHJ6M5/cXJ0KNThmPK4txBCmKCyccTJr16mTU1B5FqhcaFtHIT1kxEbIYQwQZv0mJg7P+MS2BK1XcaXnBU0pbrNNHcJ/8lSR5KEZZARGyGEMEGbGEnEX5+jTYw0dylCCBOksRFCCCFEgSGNjRBCCGEBDh07RWRUdJbfDxo1kzav9+XIibPGZecuXKHPkCn0GzaV67cyvopg6dqtnDl3MU/qtVTS2AghhBD5TK/XZ1h2+NhpY4xCZiaN7k+Xji+mW7b6050smDaCKWMGsHL9F7lepzWSycNCCGGC0sYRp2INURaSp6KsgSU9jp7dzKoz5y6y9ct9qFRK6tWqwo+/HEenN9D51TZUqViGE6f+5Nade9SqXolB77+VYft/p3MnJaegUipxcXbExdmRmNg4AK7duMOsxeso4uaKVqulSf0aOT5HayKNjRBCmGDj6INf/ZHmLkP8zbVCY6IvHzV3Gc8lPj6BlfPH02/YVFbMHYdSqaT/iGm0ataABnWq0a3Ty5Qukb2g1di4+HSRByqVitRULWs27WLSqP4EFPOh3zDT770paKSxEUIIEwz6VLRJ0ajtXFEo5f0u5mbpj6P/lwrlSvI4Oobg+6F8NHY2AHFxCURFZ30LKivOjg7EJyQaP+t0OjQaNZGPo4wJ4E9SugsTaWyEEMKE5Oi73P1lJMVbzMOuSGlzlyOsmEKhxM3FmUB/X5bMHINGo0ar1aJWq1GrVOh1GefeZMXOzhadTkdsXDwJCUm4ODsBUMTN1ZgAfuX6bVo0qZtXp2ORpLERQggh8pFSqaRnt9cY/PFslAoFbq7OzBg/mIZ1q7N49Rbq1qxCz26vZthu+oK1nP3zEoePnebmnWB6dOlA356dGTZ+PgoFxnDMvu92YtLsFRRxc8HZqfDNC5PGRgghhMgHtapXolb1SgA0qFONBnWqpfs+qEldgv5jdGX88L4ZltWsWoFPFk9Kt6x8mRJsWFb45tY8IY2NEEIIYUHuBIcwZ+mGdMsmjx6Al4e7mSqyLtLYCCGEEBYkMMCvUKdz55Q0NkIIYYKtW0nKvLodhVJl7lKEECZIYyOEECYoFEoUKnlRuxDWQP6fKoQQJqTEhhB8eAIpsSHmLkUIYYI0NkIIYYJem0hixF/otYmmVxYiG3oNmpDvx+zcazgDRk5nwMjp/H76fIbvx05bwoPQ8HyvK7fJrSghhBDCCuj1epTK5x+PcHJ0KBSTkqWxEUIIYRVubh1rkRlR2QnBvHDpOotWbcbO1oaa1SqgN+iZv3wTl67eJKhxHd7p0p7fT59n47Y9JCcnE9SkLj26dGDv/sOcOPUniUnJvP5KK9Zs2knFsqW4cTuY5o3r8HbnV4iKjmXWonXEJyRS1N2NiSM/QJXJnLCExCT6j5iOZ9EiDB/4Lq4uTvx+5gIr1m/D19uTR5FReXB18p/cihJCCGEVoi8fxaBLNXcZz+Xo72fp/XZHVswbx3vdOxIXl0D3zu1Yu2gS3/+U1qxVq1yOVfPHs27JFA4eOUlScgoAarWK+VOH06heDWJjE3jrjZdYs3AiR0+cJTIqms+2/4/Or73A8rljKVMygEPHTmVaw9pFE1k1fzwN6lRj3eYv05Zt2snSWR8zdcxAIh49zp+LkcdkxEYIIUzQOHjiXbM/GgdPc5dS6ClUmmyNkFiaTu3bsHHrHn74+RgvtmyEs5Mjvt4eANjYpAWrXr52i/Vbvkar0/EgLILHUWnBmBXLlTLux97e1hhwWaZUcR6EhnP77n0uXr7Ohs+/Jjk5lbatGmdag6uLMwAtm9bjf98fBNJub7m6pGVMlS5ZPPdP3AyksRFCCBNUti64lmxj7jKEFXN0dGDEoHdJTdXSc9B4bDQZU+K37NzLqMG9KObrRc+B48FgAEChUBjXSUxMNgZcXr8VjK+3J4EBvjRvVIcaVSsAoNVqM+w7NVWLwWDAxkbDHxeuUMzPG0jLrYqJjcfO1oYbt4Pz4tTznTQ2Qghhgi45hriQ33Dyq4/K1sXc5QgrtHvvzxw6ehKtTk+7Ns348eDxDOu0aFKXMVMXU7pEAA72dpnux9nZgS++/p4r127RvFEd3Iu40vOtV5m1eD3rNn8FwMD3u6Yb5QGIjYtn2Ph52NvZotFoGDesDwB9enTiwzEz8fX2xMezaC6ftXlIYyOEECakJoQTdnYVtm6lpLERz6Xr623p+npb4+dunV42/rxxeVpgZbsXmtHuhWbptmvnk/72p1KhNKZ4P+Hq4szsiUP+8/juRVz5dMX0DMvr165K/dpVs3MKVkMaGyGEEKIAiYtPYNTkhemWDe7bnQplS5qpovwljY0QQuQCgy6VsxObm7uMAs2gS0Whyjg3pTB5MrrzXwrL+2qyIo97CyFEDrlWaFzof+EKYSlkxEYIIUxQqu2x96iMUm2f6felus3M54oKJxkRE9khjY0QQphg4+xHQDPTtwCEEOYnt6KEEMIEg0GPXpeKwaA3dylCCBOksRFCCBOSo25xfU8XkqNumbsUUYAdOnaKyKjoLL9v9dr7xnTu67cyvkzPHInhlkhuRQkhhLAalvj02fNEPGSW1H342Gn8/bxxd3PNdJvi/r6F+mmn7JLGRgghhFVwrdDYItO9s+vMuYts/XIfKpWSerWq8OMvx9HpDXR+tQ1VKpbhxKk/uXXnHrWqV2LQ+29l2P7+gzD6D59GyUB/PvqgO7Y2Nuw7cIQdu38goJg3CYlJZjgryyONjRBCCKtQEJ4+i49PYOX88fQbNpUVc8ehVCrpP2IarZo1oEGdanTr9DKlSwRkuu3OjQtwdXFm/Zav+fKbA3Tp2JZtX+1j3eLJJCQm8XqPIfl7MhZKGhshhBAin1QoV5LH0TEE3w/lo7GzAYiLSyAqOsbktk+nc3+2/X9ERcfg5eGOjY0GGxsNvj6SPg/S2AghhEm2rsUp2XYtarvM5z4IkV0KhRI3F2cC/X1ZMnMMGo0arVaLWq1GrVKh12X+5F1iUhI2GhtUKiV/XLiMv583bq4uPIyIJDVVS0JiEg9Cw/P5bCyTNDZCCGGCQqlB4+Bh7jJEAaFUKunZ7TUGfzwbpUKBm6szM8YPpmHd6ixevYW6NavQs9ur6bYJvh/GjIVrcbCzw9nJkYmjPkClUtKl44v0HTqFwABfvL0KRjp3TkljI4QQJqTEhxJxYTMeVd7BxtHH3OUIK1WreiVqVa8EQIM61WhQp1q674Oa1CWoSd1Mty1XOpBNK2ZkWN6uTTPatWmWyRaFlzQ22fDj0QvcCYkwftbrDSiVCjNWJITIT/qUeOLuH8e93OvgaO5qREF3JziEOUs3pFs2efQAvDzczVSRdZHGJhvuhERIMyOEECJfBAb4yftqckAam2xSKhW83zkIgHU7D5q1FiGEEEJkTiIVhBBCCFFgSGMjhBAmqO3d8ajcHbW9zHEQwtLJrSghhDBBbVcE9/JvmLsMUcA9CA1n2SdbmTnhowzf/XH+MnOXbiQ6No69X6wA4K/LN1i8ejNqlQpPjyJMHPkBarWah+GPmLf8UxISkqhRtQJ9eqT/u7t07Vaa1K9hfEKroJHGRgghTNClxJMY8Rf2HpVR2chjUSL/lSlVnA3LptJ/xHTjMm9Pd5bNGYudrQ0rN2zn8PEztGxaj+XrtjHyw16F9ikqaWyEEMKE1PhQQk7MpniLeahsSpu7HPEcQk7MJT70VJ7su+xrO0yuM3/5Jl5q3ZjKFcpw8uwFTpz6k0eRUYSFR+Jgb8ukUQP+c3snR4cMyzyKFjH+rFGrUSoUaLVaHoRFsGztVh5HxdD33U5Uq1yOazfuMGvxOoq4uaLVamlSv8Yzn6e1kDk2QgghCrz40FMY9FqzHb9NUAN+PHgCgJ8O/Ya7myteHu6smj+e1s0bsnPP/ufe94OwCH4/c54mDWoSFR3L1Rt3GPh+V6aMGcDi1ZsBWLNpF5NG9WfelGEkJCbnyjlZKhmxEUIIUSgolOpsja7khWqVy7F83TZSU7XcunsfP19PKpYrBUDFcqX4/cz559pvfHwCU+euYvzwvqjVapycHPH388bHKy0CRKVSo9XpiHwcRWCAHwAVypbIlXOyVDJiI4QQQuQxhUJB1UrlWL/lK+rWrIy/nw8Xr9wE4NLVm/j7PXtUh1anY8KsFbz3dkdj02Jna4OrixOxcfEkJiWRmpqKWqWiiJsrwfdDMRgMXLl+OzdPzeLIiI0QQpigUNlg4xyAQmVj7lKEFWsT1IA+H01my5rZ+Bfz5uCRk/QfPg17ezsmjepPQkJiltvevnufBSs+4+79B3w4ehaD+rzFzdv3+OvyDTZu3c3Grbt5vV1rWgc14IOebzJy4gJStTrefyftiai+73Zi0uwVFHFzwdmpYE+Al8ZGCCFMsHUJoESbJeYuQ1i5iuVKcWTfZ8bPUz8emO57VxenTB/1BihRvBjL5nycbln5MiV4qXWTDOtWq1yO1QsnZlh3w7Jpz1u6VZHGRgghhMXL6VNNBr0WhdI6fuVt//p7Dh3751xLlwhg+MB3zViRdbGOP2UhhDCjpKhb3Ds8Hv9m07FzK2nucgqlJ081WUtzkhNdOralS8e25i7DahX8vyFCCJFTBj16bSIY9OaupFDLyVNN13a/mcvVCEslT0UJIYQQosCQxkYIIYQQBYY0NkIIIYQoMKSxEUIIE2yc/SneYh42zv7mLkUUYA9Cwxk7LevXCnzx1fcMGj2TASOnExL6EIC+Q6fQf8R0+g6dwq079zNsM3baEh6EhudZzZZIJg8LIYQJSrUtdkUk/NLaGfTaPJlEnB8xDZev3SIsPILlc8amW75i7jg0GjVnzl1k21ffMXZonzyvxdJJYyOEECakJoQTefVr3Mt1ROPgae5yxHNw9KmTZ+ne2ZHTdO/Dx06TnJzKoFEzKRlYjCEfvINKpUSjSfs1Hp+YRKnAAAB+P3OBFeu34evtyaPIqLw+NYsjjY0QQpigS44h+ub3uAa2ksbGSvk1GGXW4z9J965coQw/HfqNgGI+qJRKJo8ewL4DR9i5Zz8vZ/IW4Scio6JRqVQsnzuW5eu28fOvv9EmqCGPIqP4eNoSwh4+Yu6UYQCs3bSTpbM+xt7Oli69R+TXKVoMmWMjhBBC5LFqlcvx1+XrxnRvnV6XLt37Xkjof27v7OhIneqVAKhTozI379wDoKi7G2sXTWLmhI9YtWE7AHq9HlcXJ2xsNJQuWTwPz8oyyYiNEEKIfPO80QjW/tbhrNK9WzStl61072qVy3Lxatr6127coZiPF1qtFqVSiVKpxMnRHjs7WwCUSiUxsfHY2dpw43ZwfpyeRbHevyVCCCGsTmGKRvi3nKR7N6pXkyMnzjJg5HRcnZ2ZMmYAEZFRTJmzCqVSiUKhMOZJ9enRiQ/HzMTX2xMfz6L5dXoWo/D9zRJCiGeksnXFrUx7VLau5i6lQHieaISCEImQk3RvlUrJx0PfT7fMx8uDVQsmZFi3fu2q1K9dNRcqtk7S2AghhAkaBw+8qvUydxmikJB075yRxkYIIUzQaxNJjr6LrWtxlGp7c5cjCjhJ984ZeSpKCCFMSIkNIfjQx6TEhpi7FCGECdLYCCGEEKLAkMZGCCGEEAWGzLERQgghLMCD0HCWfbI10yej9nz3C998fxClUkHzxnV4u/MrPAgN573BEykZWAyAGeMGY2OjYfTkReh0OlQqFeOG98XX2yPdvsZOW8KHfbrh61Mw36ItjY0QQpigUKpQ2bigUKrMXYoopOrWqkyHl4IA+GD4NF5u0xSAmlUrpGuEklNSmDS6P55Fi3Di1J98vnMvIwYVrieqpLERQggTbF1LUPqVT81dRqGXV+ncOZWdd/LkNATTz8fL+LNapUKpSJtJ8ufFq3wwbCrVq5Tng15vYmtjg2dRGwA0ajVKpQIoXMGYMsdGCCGExXP0qWPVbyt+EoIJ8NOh33B3c8XLw51V88fTunlDdu7Zn639HDx6kmJ+Xri5OlPU3Y2dGxewasEEHkfHcPDISeN6qala1m35is6vvgD8E4w5dcxAIh49zv0TtCDW+7dECCHySXLMXUKOz8av4RhsXQpfqKAlMHc6d05Vq1yO5eu2GUMw/Xw904Vg/n7mvMl9XLtxh117fmT+1OEA2NhoAA0AQY3rcuHSdVo0rQfA7CXreeOV1gQUS8ugehKMCRT4YEwZsRFCCBMMulRS40Mx6FLNXYqwUlmFYALZCsF8GBHJnKUbmDS6vzHsMv6pbKk/LlzB388bgPVbvsLPx4vWQQ2M3z8JxkxJSS3wwZgyYiOEEELkg5yEYH7y2ZdERccyafYKAEYPfo+Q0HDWfLoTO1sbfH286PtuJ8IePmLj57upVqUcp8/9RZWKZRnwXpdCFYwpjY0QQligkBNziQ89ZXpFK1NYk70hZyGY44b1ybAsMMCPhnWrp1vm7VU03TGeKEzBmIXzb5cQQli4+NBThboJKMwkBDNn5P8xQghhgsbJl2KNJ6Bx8s3X4yqU6mw9SmxNLPFxbUsjIZg5I42NEEKYoNI44Ohd09xlCCGyQZ6KEkIIE7SJkURc/AJtYqS5SxFCmCCNjRBCmKBNekzk5R1okwr2i82EKAiksRFCCCEsyN79h0lN1WZr3R27f6BjjyGMnbYkw3dzlmzIdHlBJ42NEEIIYUH2/niYVG3Gxkav12dY1jqoActmf5xh+YPQcCIiC+cIo0weFkIIUSjl1ruCsvvk2plzF/ls+/+wtbUh5EE4k8f059sfDnH52m2Sk1MYM6Q3yckpXLtxl2Hj5tK8cR3iExJ5EBbB46gYPuj1JuVKB6bbp7ubK8lJKRmOtWXnt7z1xsvsymYGVUEijY0QQpigsnHCOaAZKhsnc5cicpE53hWk1elYPGkox0+e49sfDtHv3c7Y2dly5fptPt+5lyljBlC2dHHmTx2Bg70d6zZ/iZenOxNG9Mv2Me4/eAiAr5dHXp2GRZPGRgghTNA4euNbd4i5yxB5IL/fFVSuVNqIi5enO7Gx8Xy+ay8nz1wAQKVSZbpNpb/DMrNr8/b/0aNr+5wVasWksXnKj0cvcCckIsNyvd6AUqkwQ0VCCEug16WgTXyE2r4oSpWNucsR1kzxz++SqJg4gkPCWLNwIpev3WLp2s8BUKvU6ebTKBTPNh32QVg4c5duJDkllbv3HvDNvl/o8FKL3KnfCkhj85Q7IRHSxAghMkiJCebuLyMp3mIedkVKm7scUUA4OzmiUMCAkdOpUqGMcXnThrUYP2MZQU3qmtzHjwePs+ubHwm+H8qHo2exZNZolswaA6RNIF72ydZC1dSANDYZKJUK3u8clG7Zup0HzVKLEEKIgqNW9UrUql4JSMt/mjTqg0zX6/zqC3R+9YVs7bNNUEPaBDXM9DtfH88sQzULMmlshBBCCCswff4aQsLCjZ/btmpCh7ZB5ivIQkljI4QQIl8Z9FqLCMO0tvT08c/wZFRhJi/oE0IIkW8cfepYVTMhrI/87coDer1B5uUIUeAMhgPBQHC+HM2g70uQ7dp8OVZ+8mswytwlGFnCqJHIfTJik8sC/TzkqSohhBDCTGTEJpe1aVzF3CUIIXJZSux9Qk8vw6f2h9g4F8uXY8poghDPR0ZshBDCBL02iaTIq+i1SeYuRRRgD0LDs0zj3vPdL/QePIk+QyazZee3AMTFJ/DehxNp+WpvbtzO/BZpr0ET8qxeSyUjNkIIIQqt3HhCKz8iGerWqkyHl4IA+GD4NF5u0xQXJ0cWTBvB8nXb8vz41kQaGyGEEIWSo0+dXEn3zo75yzfxUuvGVK5QhpNnL3Di1J88iowiLDwSB3tbJo0a8J/b+/l4GX9Wq1QoFUrUajVF3FwyrPvZ9m/49fgZSpcIQG/QZ/i+oJPGRgghRKGUn09otQlqwI8HT1C5Qhl+OvQbAcV8UCmVTB49gH0HjrBzz35ebt3E5H4OHj1JMT8v3FydM/0+8nE0x0/+ydpFk7gTHMKw8fNz+1QsnsyxEUIIEzSOXvjU+QiNo5fplYXIRLXK5fjr8nVSU7XcunsfnV5Hxb9TuyuWK8W9kFCT+7h24w679vzIsP49slwnJDScMiUDUCgUlCheDHt721w7B2shjY0QQpigsnHGpXhzVDaZ/ytZCFMUCgVVK5Vj/ZavqFuzMv5+Ply8chOAS1dv4u/n85/bP4yIZM7SDUwa3R87u6ybFV9vD27cDsZgMHAnOITExORcPQ9rII2NEEKYoE2OJurGPrTJ0eYuRVixNkEN2LLjW1o3b0izRrUJC39E/+HT+PHgcTp1aPOf237y2ZdERccyafYKBoyczp3gEACGjZ/H76fPM3vxevbuP0xRdzfq1apKnyGT2bFnP87ODvlxahZF5tgIIYQJ2oQIHp77BDv3cqhtXc1djrBSFcuV4si+z4yfp348MN33ri5OWaZxjxvWJ9PlC6ePzLCs51uv0vOtV3NQqXWTxkYIIYSwINu//p5Dx/55Wqt0iQCGD3zXjBVZF2lshBBCCAvSpWNbunRsa+4yrJbMsRFCCCFEgSGNjRBCmKDU2OPgVQOlxt7cpQghTJBbUUIIYYKNkx/+TSaauwwhRDbIiI0QQphgMOjQpSZgMOjMXYoowP4rBBPgi6++Z9DomQwYOZ2Q0IcA9B06hf4jptN36BRu3bmfYZux05bwIDQ832q0BDJiI4QQJiRH3ebuLyMp3mIedkVKm7scUQhdvnaLsPAIls8Zm275irnj0GjUnDl3kW1ffcfYoZk/Fv5f9Ho9SmXBGeeQxkYIISxUbiRPi7yXnXTvnIZgHj52muTkVAaNmknJwGIM+eAdVColGk3ar/H4xCRKBQYA8PuZC6xYvw1fb08eRUZlur8HoeFMnbeaou5ulC0diJeHO//74SDxCYl07fgSL7VuwrrNX3L/wUOiY+JITEpm0Yx/srX0ej3zl2+iSsUyvNymaTavVP4oOC2aEEIUII4+dVAo5d+eBcWTEEyAnw79hrubK14e7qyaP57WzRuyc8/+/9w+MioahVLB8rljsbW14edffwPgUWQUfYdOYf6yT6lZrQIAazftZOmsj5k6ZiARjx5nuc/wR4+ZNKo/73btQIsmdVk5bzxrFk7ii6/3Gdfx9/Nh4fSRVKlQhpNnzgOg0+uZuWgdtapVtLimBmTERgghLFJ+Jk+LvFetcjmWr9tmDMH08/VMF4L5+99NQ1acHR2pVD5t/To1KnPurysAFHV3Y+2iSfx1+QarNmxn8czR6PV6XF2cAChdsniW+yxTqrhxxOfE6T/ZsfsHDAa4FxJmXKdcmUAAvDzdiYmLB+Di5RuULR1I66AGz3Mp8pyM2AghhBB5LKchmNUql+XqzTtAWsp3MR8vtFoter0eACdHe2M4plKpJCY2npSUVG7cDs5yn0qFwvjzp1v3sGDaCBZNH4md7T8hmwr+WQdD2v9UrVSW+rWrsnTN59m/APlIRmyEEMIEW9dASrXbiErjaO5ShBVrE9SAPh9NZsua2fgX8+bgkZP0Hz4Ne3s7Jo3qT0JCYpbbNqpXkyMnzjJg5HRcnZ2ZMmYAEZFRTJmzCqVSiUKhMMYu9OnRiQ/HzMTX2xMfz6LZqi2ocV0+GD6N8mVK4Oxk+u95l45t+XTbHtZu2kXfdztl7wLkE0VKSrLhWTbQarU0bdeTX/d+ilqdO33Rup0HAXi/c9B/LstrWR3z38vNUZsQQgghTJMRGyGEMCElLpTwPzfgWe09bJz++5aBEDmV2yGYd4JDmLN0Q7plk0cPwMvD/bn3acmksRFCCBP0qfHEh56iaMUu5i5FFAK5HYIZGODHynnjc21/ls6iGxu93mC87ZNfx1MqFaZXFEIIIYRFstinogL9PKTJEEIIIcQzsdgRmzaNq+T7MfNzdEgIIYQQuc9iR2yEEMJSqO2L4lm1J2r77D06K8TzyGnA5O7vfjb+/Nn2b4xBmc9j3eYvOXLi7DNvd+bcRe7ee/Dcx80N0tgIIYQJajs3ipTtgNrOzdylCGF8Kd+/7fnuF+PPPbp0wM/HK79KAtLqOvPnJe7eC83X4/6bxd6KsnT5PbFZCGE+apJwI5goAtBiZ+5yhIXJzjvNchqCuW7zlzwIi+BxVAwf9HqTLTu+JfzRY/R6PVPGDODytVvcvfeAASOn8+pLLfn9zHm6dXqZEgHFmDZ/dbrjuDinfwGfVqtlzNQlJCSmvSDwSdjlT4dP8NW3B4wBmHa2Nixd8zl/XbmBRq1m3LA++Pp48lafUVQuXxq1Rs1vp8/zy5GT/HT4NyaN+uD5LmgOSWPzHAL9PLgTEmHuMoQQ+cSOGCqp9nFG15U4aWzEc3gSglm5Qhl+OvQbAcV8UCmVTB49gH0HjrBzz35ebt3kP/fh5enOhBH9ABg79H3s7Gw5ePQku/f+zAe93mTTtm+Mj3U/yZ46dOwUXh7u6Y7T++2O6fYb+vARdrY2zJ86HIPBgOLvqAV/Px8mjerIinVfcPLMeTyKFiH80WPWLJzIH+cvs/7zrxk/vC8PIyJZs3ASLs6OrNv8JRXKlqJJg5q5fQmzTRqb52COic1CCPNJenyDu798wWuta2NXpLS5yxFWKKchmACV/l5fp9OzfN0X3Lh1l+SUFEoF+me5zb2QUJPH8ffzpmqlskyesxIfLw/69EiLSPh3AGZScso/+ypfitUbdxi3//cokDnJHBshhBAij+U0BDNtH2m/sq/dvENcfDyrFkzgnTfbP8mmRJHJG1Kyc5yUlFQ6v/oCk0cPICo6lj8vXk3b378CMP39vLl09e99XbmJf7G0fSkV/7QSapU6yzlA+UUaGyGEECIftAlqwJYd39K6eUOaNapNWPgj+g+fxo8Hj9OpQ5ts7ycwwJfQsAgGj5nNmT8vGZfXql6JUZMWpotjyM5xQh9GMGDEdAaMnE74o0jKlymR6XErlitFUXc3+g2bytrPdvFe99cyrFO7RiW2ffkdi1Z9lu3zyW0WG4JpDtkNwRRCFC7JMfcIPbkIn7pDsXXJethfCGF+MsdGCCFMsHXxJ7DVAnOXIQqJ3A7B/Leh4+aSnJJi/Ny9Uzsa1zffZN/cJo2NEEIIYUFyOwTz3548zl1QyRwbIYQwISnqJtd2v0lS1E1zlyKEMEEaGyGEMMVgwKDXguGZpiQKIcxAGhshhBBCFBjS2AghhBCiwJDGRgghhLAAlpTu/ax27tnP3v2HeRQZxSeffZlvx82MPBWVTRJ6KUThpUSLHd05cuAWeoLNXY6wMPn9jjO9Xo9SmXFcYs93v/Dayy2BtHRvcyjq7kafHm+Y5dhPSGOTDRJ6KUThpkdNAkXNXYawYpac7g3w7sBxVKlYlvMXr9GhbRB/Xb7O9ZvB9Or+Gi2b1uPS1Zss/2QbOp2Opg1r071zO8IePmLS7BXY29tho9HQrFFtHoSGs+yTrcyc8BEDRk5n/tQRONjbMXbaEj7s040zf17i1xNn0Gp1RMfE0rFdK/YdOILBYGDxzFG58uJfaWyyQUIvhSjcUhMe8ujSTopW7IzGwcvc5QgrZMnp3gBxcQn0eLM9zs6OtH9rEDs/XYBGrWb0lEW0bFqPleu3M2viEFycHRkxcQEvtW7C5h3f0qt7R+rXrsqEmcuzfS1cnZ34eOj7rNqwnas37rBszscsXr2FPy5coU6NytneT1aksRFCCBN0ybHE3PkJt1JtpbERz8WS070BnJ0c8fZKG5UMKOaDu5srkBaQCXDjVjBjpi4CIDY2gbDwR9wLCaVC2RLGff/b0yGaT78poUyp4gB4FC2CvZ0tAJ5FixAbG//fFyCbZPKwEEIIkccsOd37SX2Z/fxEmVLFmTt5GCvnjefTFdOpULYk/n4+XLl+B4DL1zK+vNLZ2ZGH4ZFodTpu3rln8ri59ZYoGbERQggh8kGboAb0+WgyW9bMxr+YNwePnKT/8GnY29sxaVR/EhISs7Wfp9O9AwN8jcufpHu3e7GZcVmzRrUzHOd5DOjdhTFTF2PQG9Bo1MyeNJS332zHpNkr2bprL44O9hm2ef2V1oybsZQyJYvjXsT1uY77PCTd2wLrEEJYlqTHN7j7y0iKt5iHXZHS5i5HCPEfZMRGCCFMUNm5UaTc66js3MxdiigEJN07Z6SxEUIIEzT2RfGs8ra5yxCFhKR754xMHhZCCBP0qYkkhF9An5q9ORBCCPORxkYIIUxIiQvh3q8TSYkLMXcpQggTpLERQgghRKZ6DZqQ6fK9+w+zc8/+fK4me6SxEUIIIUSBIZOHhRBCiHxw5txFPtv+P2xtbQh5EM67XTvw7f5DxMTGs2jGSDZt+4a/rtxAo1YzblgffH086fr+KCqVL8WV67fp3qkdvx4/w937Dxg1+D2qVy6X4Rj7Dhzhfz8cJD4hka4dX+Kl1k14HBXD9AVrSUhMwsvDnSljBnDs9z/Y8PnX2NhoaP9iEG1bNWbhys+4efseSqWSCSP64uWZvXy0bV9+x8+//o5SqWRY/3coX7Yk7w4cR9WK5bh09SZBjevwTpf2hD6MYOKsFTg5OuDoYE+DOtVo90Iz0wd4RtLY/IukeAsh/s2BCKoqHfnqwFkSJN1b/MuzvPtMrzcwZ9JQdn/3Mz8eOs7imaPZ/vX3/O/7Q4Q/esyahRP54/xl1n/+NeOH9yXycRQjBvUk4tFjBo6cwa5NC7kb/ICtX+7LtLFp0aQuL7VuQlJyCv2GTeGl1k34bPs3vPJCM1o0rYder0ev17Nqww5WLxiPo6MDer2eo7/9gbOTIyvmjeOvy9f5bPu3jBhk+hHzR5FRHD52mjULJxL28BGzFq9n6ewxxMUl0L1zO7w83OnRfyzvdGnPlh176f3269SvXZWJs1Y8yyV+JtLYPEVSvIUQmUnAg9/0vc1dhigAypQMAMDDvQhlSqZlJnl6uHM3+ME/mU7lS7F64w4A/Hy8cLC3w6NoEfyL+WBrY4OnhzuxcZnnKp04/Sc7dv+AwQD3QsIAuH03hHe7vgqAUqkk8nE0Xp7uODo6GJfdvnufQ8dO8ceFyxgM4O3pnq3zeRAWQZlSxVEqlfj6eBIXn1aXs5Mjvt4eANjYaADSZUtVKFsy+xftGUlj8xRJ8RZCCJGn0mUy/bNYo1Fz6erfmU5XbuJfzOfvdTJf32DIPDTg0617WLVgPAoUvNFzGAAlivvxx/nLBDWpi16vx83VmfCISBISk3Cwt0Ov1xMY4EurZvV5r3ta8rdWq83W6fh6e3Dtxh30ej1hDx/h5OiYoe4nnmRL1atVhas3blO3Zt78zpXGRgghTEiOvsP9o9Mo1ngCtq6B5i5HFFBF3d3oN2wqKpWS8cP7Ptc+ghrX5YPh0yhfpgTOTmlNRo8uHZi+YA3bd/9gnGPTr2dnPhw9CztbG155sTltWzXm1B8XGThyBgqFghdaNqJD26Bs1dy0YW36Dp2CQqFk+MAeWa779pvtmDhrBdu+/A5bWxvUatVznaMpkhUlhBAmSFaUEDmn1elQq9KamYmzVtCl44tUrlAm148jIzZCCCGEFVq5YTsXLl0zfq5Xsyo9u72a68c5cPAEX+09YPzs5urCzPGDn3k/oWERTJ+/Bp1eT5mSxfOkqQFpbIQQQgirNOC9LvlynNZBDWgd1CDH+/H382b1wom5UNF/kxf0CSGEEKLAkMZGCCFMsHHyw7/pVGyc/MxdihDCBLkVJYQQJig19jh4yusghLAGMmIjhBAmpCY+IvzCFlITH5m7FCGECdLYCCGECbqkKB5f/QpdUpS5SxEFxNUbd/jqfwey/H7a/DXcuJ0+vuPG7WCmzV+T16U9lzPnLrJ07VZzlwHIrSghhBAi35UrHUi50uZ72eOTNxdn9oZgayeNjRBCFHI9+g7n+wOHzF2G1Xp485TJdRat+oxWzRpQrXI5fjt9nvVbvqJKxbIM7tuN4yfPsWnbHnR6A51fbcMLLRoZt9PqdEyatYKY2Hh8/s5eykzw/VDmLN2AXq+nQpmSDO7XnV6DJrBx+TQA48/T5q/Bzs6W4Huh1KtdBR9PD1oHNeBeSBhrN+1k6seD+HTbHn47fR4MBoYP6mnMt3ramk07OXX2IjYaNf3/9dj5jweP88VX36NQwPvvvEGDOtUYMHI6pUsEcP3mXcqVCWRo/x4kp6Qwa9E6Ih5FYW9vy+RR/Y35VTkhjY0QQhRy3x84REpKqjGsUOS+l1s35ZvvD1Ktcjl++PkobVs14V5IGAaDgY1bd7Ni7jiUSiX9R0yjVbN/3hlz+Nhp/P286f9eF7769if+unw90/2vWP8Fg95/iwplS6LX6/+zlvJlSjByUE/CHj5i8eottA5qwIFDJ2jVvAE3bgdz994DVs0fT/ijx8xbupG5U4Zl2Mfvpy+wZtFE1CoVer2eP85fBkCn0/PZF/9j3dLJaFO1DBo9iwZ1qgHQpEEthg98lwkzl3Pl2i3+vHiN2jUq0/7F5hw4eILd3/1C987tnvMK/0MaGyGEMEFl64xLYCtUts7mLiXP2NhosjXyIJ5P+bIlubX2c+LiE3gYHkmJF3y5FxLG4+gYgu+H8tHY2QDExSUQFR1j3C4tETstCbtS+VJZNjYPwx8Z11MqM06ffTo0s9LfKeLeXkWJi08gPj6B306fp9sbL3P4+GnOX7zGgJHTs9wXwPvvvM6MBZ9ga6uhzztvGJdHRcfg7VUUWxsbbG3S8qC0Oh2AMdm7YrlSBN8P4/bd+1y6epN9B35Fq9VRo0p50xcyG6SxEUIIEzQOXvjUHmjuMoSVa1yvJnOXbqBpw1rGZW4uzgT6+7Jk5hg0GjVarTZdDqO/nw9Xb9yhRdN6xvTvzHh5FuXKtVuU/3vERqlUkpKSik6nJ/xRJLFx8cZ1Fcp/5tU0a1SbzTu+pZiPJzY2GkoE+FGzWgXGDu0DZJ3yXbNaRRrWrc4PPx9j93e/ULNqWlPi5upC6MMIklNS0KZqSU3VGvOhniR7X752k1rVKxIZFU2VimV5qXWT/zzWs5LGRgghTNDrkkmND0Pj6I1SZWvucoSVerFlI9Zu2sWQ/u9w+859IG1EpGe31xj88WyUCgVurs7MeCqHqVmj2hw4eJxBo2dSvJhvlvse2Lsrsxevx4DBOMfmhZaN6DNkMjWqVsDJKfO5K62a1ee1dz5izqShAJQpVZwAPx/6j5iOUqGgbq0q9HwrY/7UmCmLSElNa5xGftiLmJhYAFQqJe+82Z7+w6ejVCro17OzcZsTJ8+x8fOvKVOqOBXKlqRkYDFmL97A3v2HAXjrjZdoXL/mM17VjCTdWwghTCjo6d5epeoA2ZsEK57fo8go5i7dyJzJQ81dSr4bMHI686eOwMHeLs+PJSM2QgghRB4799dVln+yjQ/7vJXjfeVXqre5jpdTMmIjxHP68egF7oREmLsMkQ+ceEgt1Rec0XUlDi9zl5PrUlJSmTzhYxmxEQWCvHlYiOd0JyQCvf6Z/l0ghBAij8mtKCFyQKlUyEhjIZAUdZPgg7t4rWUd7NxKmbucXPdkjo0QBYE0NkIIYYKdWynKvrbD3GUIIbJBbkUJIYQQFspUWKbISBobIYQwITnmHnd+Gk5yzD1zlyIKCFOxB0+UKx3I6+1b53E1BYvcihJCCBMMumSSo29h0CWbu5Q8k5KSKnNtnlN2nyY7c+4iW7/ch0qlpF6tKvz4y/F0wZdXrt9m9uL1eBR1w2CAbm+8BMCR3/5gcN9uvDtwHFUqluX8xWt0aBvEX5evc/1mML26v0bLv99MvPyTbeh0Opo2rJ0ruUvWSEZshBCikGvburkEYOaT+PgEZk8cwg8/H2PZnLGsXjCBXd/8iE6nZ+2mXUwZM4C5k4eli0B4Ii4ugR5vtmf1ggms2rCdD/t2Y+X8cez6Zj8AK9dvZ9bEIaxeOJGz5y8T+Tg6v0/PIsiIjRBCFHKfrV1g7hIKjQrlSmYZfPk4Kpri/mmxCeVKB2bY1tnJEW+vogAEFPPB3c0VSBttA7hxK5gxUxcBEBubQFj4I9yLuOb5OVkaaWyEEEKIfKJQKLMMvizi5krw/VD8/by5duMOLZrU/de2ikx/fqJMqeLMnDAYJ0cHdDo9SmXGdQoDaWxyibyFtvDR6w2F9j8chY3G0RvfeiPQOHqbuxRRAGQVfNn33U5MnLUC9yKu2NnZolKr0T1D4vWA3l0YM3UxBr0BjUbN7ElDsbO1ycMzsUwSqZBL1u08KL/oCpknf97W/PdWCGE5nozc6PV6Bo6aybSPB+JRtIi5y7I6MmKTi+SXXOHypCEXBZ82KYrY4MM4BzRDbedm7nJEAfXX5Rus3riD5JQUmjasLU3Nc5LGRgghTNAmPiL8/KfYe1SWxkbkmepVyrNqwQRzl2H15HFvIYQQQhQY0tgIIYQQosCQxkYIIYQQBYY0NkIIYYJS44ijTx2UGkdzlyJErrt64w5/Xb4BwIPQcMZOWwLA3v2HSU3N+nHzM+cuMmj0TAaOnMHBoycBmL98E/1HTOe9Dyfw86+/Z9hm55797N1/OA/O4h8yeVgIIUywcfKhWKOx5i5DFCJ6vR6lUpnh57xw7cYdEhKTqFyhdLrle388TIum9dBoMrYKSckpbP1yH4umj0r3/Uf9uqPRqIlPSKT/8Gm0bFovz+rOijQ2QghhgkGvRZcaj0rjiEIp/9kUz85gMDB/xSZu3LyLSqWiT483WLVxBwaDgcb1a/Ju1w6s2/wlD8IieBwVwwe93mTRys8o6u5G2dKBvNu1Q7r9HT52mk+37cHezpZWzerzevvWdOk9kvJlAgm+H0bX11+iVbN6DBg5g7WLJgEwafZKer/d0Rjb8MRX3/5ETGwcR06cYcxHvQE4f/Ea127cZdi4uTRvXIe33ng53TYXLl3D1kbDyEkLsLW1YdSHvSjq7mZscpKTUyhRvBgAYQ8fMWn2Cuzt7bDRaGjWqHaeXOMn5P+hQuSAXm+Q99kUAk48pJbqC87ouhKHl7nLERYmO+8vO3LiDEqFgtULJwIwYuICPh7Sm8AAP4aMncMLQQ0B8PJ0Z8KIfjwIDSf80WOWzv440xGTX478zoQR/SgZWAy9Xg9AeEQkaxdNwt7Olvc/mkTr5g2oULYkf12+QcnAYjx6HJWhqQF4/ZVWJCQm0fnVF3gQGg5A1UplKVu6OPOnjsDB3i7DNpGPo7n3IIx1i6dw8uwF1m3+itEfvQfAhJnLOX3uIgPf7wrA5h3f0qt7R+rXrsqEmcuzcUVzxmIbG2uLKJC3Dhc+gX4eVvV3VAhhPrfvhlCzWkXj58jHUcYRjfJlSnDvQRgAlcqVMq5TplTxTJsagF7dXmPrru9ITknhjfZtqFKxDL4+nri6OAHg5VmU6JhYXm7TlO9+/JWK5UrRvFGdXDsfZydHqlUqh0ajpk6Nynz2xTfG76aNHURMbDzvfzSJl1o14V5IKBXKlgCg4lPnl1cstrG5ExIhzYKwaG0aVzF3CSKfJD2+wd1fvuC11rWxK1La9AZC/EuJ4n6cPPuXcc6Jm6sLt+/eJzDAjyvXb9OxXSvOXbiCQvHPXBplJkGXT3h5FmXMkN6EP3rMlDmrWD53LKFhEcTExmNna8PD8Ee4ujjjXsSV5Z9sI/h+KJNG9c90X2q1yjjqk265Sp3pckhrULbu+g6DwcC1m3fw800byUxJScXGRoOdrQ0O9nYolUr8/Xy4cv0O9WpV4fK1m9SvXS3b1+15WGxjA9YVUSC3I4QQQmSlSYNanDj1J/2GTUWtUtH7ndeZuWgdBgM0rlcDXx/PZ9rfhi1fcf7SdbRaLZ06vACk3cZatOozbt8NoVunl1Gp0pqkBnWr8cf5K7i5Ome6ryoVyzJ13mr+unyD/r3eNC5v2rAW42csI6hJXV57uWW6bdxcnWneuA4DRkwHhYJxw/oAabehYuPjSU3V0fOtVwF4+812TJq9kq279uLoYP9M5/k8LDYE09qCMa2tXiFE9qWN2IykeIt5MmIjLFavQRPYuHxahuVbd32Hj7eHWZ5QMgeLHrERQghLYOtWgtLtt6BU25q7FFEIHTh4gq/2HjB+dnN1Yeb4wdnadv2Wrzl/8Srzp4147n2d+fMS6zZ/mW7ZstljjSNClkYaGyGEMEGhUKHSOJi7DFFItQ5qQOugBibXy2y0pvfbHZ9rX0+rVa0iK+eNf6ZtzMky2y0hhLAgKXEh3DsylZS4EHOXIoQwQRobIYQwQZ+aSMLDP9CnJpq7FCGECdLYCCGEEKLAkMZGCCGEEAWGNDZCCCFEPrt64w5f/e9Alt9Pm7+GG7eD0y27cTuYafPX5HVpVk+eihJCCBPUDh54Ve+D2sHD3KWIAqJc6UDKlQ402/ENhrRX2Cn+4+3G+S23UsylsclFEogoREFmD2fOmrsIYYGy82LWRas+o1WzBlSrXI7fTp9n/ZavqFKxLIP7duP4yXNs2rYHnd5A51fb8EKLRsbttDodk2atICY2Hh/vrBvr4PuhzFm6Ab1eT4UyJRncr3u6F/Y9+Xna/DXY2dkSfC+UerWr4OPpQeugBtwLCWPtpp1M/XgQn27bw2+nz4PBwPBBPSlTMiDD8cZOX0pUdAxxcQkU8/Wme+d2LFq1GTtbG2pWq8D777zBsd//YMPnX2Njo6H9i0G80KIR0+avJiw8Egd7WyaNGkB8fAJT5602ppjXq1WF5Z9sQ6fT0bRhbbp3bvfMfx4W09j8uymwtpwoCUQUouBSk4S74jaRhhJoyZh0LIQpL7duyjffH6Ra5XL88PNR2rZqwr2QMAwGAxu37mbF3HEolUr6j5hGq2b/vGfm8LHT+Pt50/+9Lnz17U/8dfl6pvtfsf4LBr3/FhXKlswy3+mJ8mVKMHJQT8IePmLx6i20DmrAgUMnaNW8ATduB3P33gNWzR9P+KPHzFu6kblThmXYx8zxg0lN1TJ6yiLe6/4aPx/5nd5vd6RRvRro9Xr0ej2rNuxg9YLxODo6oNfrOXj0FF4e7kwePYB9B46wc89+Xm7dJF2K+YejZzFr4hBcnB0ZMXEBL7VugnsR12e61hbR2BSEpkACEYUouNIiFdZKpIJ4buXLluTW2s+Ji0/gYXgkJV7w5V5IGI+jYwi+H8pHY2cDEBeXQFR0jHG7tGTskgBUKl8qy8bmYfgj43qZ3c55cusJ/kkQ9/YqSlx8AvHxCfx2+jzd3niZw8dPc/7iNQaMnJ7lvp6Yt3wjr73ckrKlA3Ev4srGrXv44edjvNiyERXKlsTL0x1HRwfjfu6FhBrTvSuWK8XvZ84D6VPMb9wKZszURQDExiYQFv7IOhubzJoCuaUjhBCiIGlcryZzl26gacNaxmVuLs4E+vuyZOYYNBo1Wq02XQ6jv58PV2/coUXTely6ejPLfXt5FuXKtVuU/3vERqlUkpKSik6nJ/xRJLFx8cZ1FU/dDWnWqDabd3xLMR9PbGw0lAjwo2a1CowdmhZqqdVqMz3e5zv34u/nTbNGtQFwdHRgxKB3SU3V0nPQeDavmkl4RCQJiUk42Nuh1+vx9/Ph4pWbxnPx9/MB0qeYlylVnJkTBuPk6IBOp3+uOzcW0dgIIYQQBd2LLRuxdtMuhvR/h9t37gNpIxk9u73G4I9no1QocHN1ZsZT2U3NGtXmwMHjDBo9k+LFfLPc98DeXZm9eD0GDMY5Ni+0bESfIZOpUbUCTk6ZR4K0alaf1975iDmThgJpjUWAnw/9R0xHqVBQt1YVY0r309Zs2kmVimU4cepPqlcuj6uLM4eOnkSr09OuTTOUSiX9enbmw9GzsLO14ZUXm9OmRUMOHjlJ/+HTsLe3Y9Ko/iQkpH/p5YDeXRgzdTEGvQGNRs3sSUOxs7V5putsEenemZG0bCGEpZB0b5EbHkVGMXfpRuZMHmruUgo0GbERQggTlGo77NzLoVTLxGHxfM79dZXln2zjwz5v5XhfKzds58Kla8bP9WpWpWe3jKMquSW/j5dTMmIjhBBCiAJD3jwshBBCiAJDGhshhDAh6fENrn71OkmPb5i7FCGECdLYCCGEEKLAkMZGCCGEEAWGNDZCCCFEPjhz7iJL127N8X4OHTtFZFR0lt8PGjWTNq/35ciJf7LNzl24Qp8hU+g3bCrXb6Wlhj+KjOKjj2fTd+gUvv/pSI7rshQW/bi3hEoKISyBEw+ppYLdB04TR7C5yxEWJi+f3s0s8fpJfpS7W+ZRA5NG92fPdz+nW7b6050smDaChMRE5i7dyMLpI9m841u6d36FWtUr0n/4NFo0rYetzbO9DM8SWWxjUxDyo4QQBUM87vyu60EyTuYuRVi5m7eDGT1lESEPwpk8pj/T56/NkMC9bvOXPAiL4HFUDP3e7cSStZ+jQIGjgz1D+r/NiVN/cuvOPWpVr8Sg9zO+F8ezaJF0n5OSU1Aplbg4O+Li7EhMbBwAF6/cYHDfbiiVSiqULcnN2/eMWU7WzGIbGwmVFEIIUdBodToWTxrK8ZPn+PaHQ1mu5+XpzoQR/Th59gKVypdm0PtvGUdvGtSpRrdOL1O6REC2jhkbF4+jg73xs0qlIjVVi1arM44GOTk6EBMbn9UurIrMsRFCCBNS48N4cHIxqfFh5i5FWLlypQKBtMYl9l+NRGYJ3DWrVcTezpZJs1ey7ct9z3VMZ0cH4p/KZNLpdGg0atRqFXq9HoC4+ARcnB2fa/+WRhobIYQwQZcSR2zwYXQpceYuRVi7p5KsDaSFYMYnJBKfkEhI6MOnVkv79azV6uj99utMGTOA38+cJ/RhBGqVCr1On+1D2tnZotPpiI2LJ+zhI1yc026pVixXijN/XkKr03H52m1KBvrnzjmamcXeihJCCCEKujfat6b/8GlUKl8aD/ciGb6/dPUmazbuQKFU4uXhjpeHOw3rVmfx6i3UrVkl08ym6QvWcvbPSxw+dpqbd4Lp0aUDfXt2Ztj4+SgUMHJQTwDeefMVps5bw5pPd9GxXctnTtG2VBabFSWEEJZC0r2FsB7P3Jk8uQeo1WpzvRghhLBEWq0Ore7v/5X/9olMqFQqFE/dZsoPd4JDmLN0Q7plk0cPwMvDPV/rsDTPPGKTlJREi1ffz6t6hBBCCKsjdzEsxzP/KdjY2PDLnnW8O3A8n6+ZnRc1iae8/cHHbFk9y9xlFApyrfOHXOf8I9c6f7z9wceoVCpzlyH+9syNjVKpxM7ODqVSKd1pPlAoFHKd84lc6/wh1zn/yLXOHwqFIt9vQ4msyePeQgghhCgwnruxeaN9m9ysQ2RBrnP+kWudP+Q65x+51vlDrrNleebJw0IIIYQQlkpuRQkhhBCiwJDGRgghhBAFRpbT5Reu/IxfT5whNCyCTStnUK50YKbrffP9QTZv/x8Gg4Ha1Ssx8sOeMgv/GQXfD2XqvDVEx8Ti5GjP+OH9KFUifWaHXq9n2SdbOXHqT1QqFa7OTowZ0puAYj5mqto6ZedaA1y/FczClZuIfBwDwAc9OxPUpG5+l2u1snudIe2lnx+OnsWV67f58au1+Vyp9cvOtT71x1+sXL+dxKQkFChoVL8GA97rYkx2FqZl9++0/E40vyz/VrdoWo81Cybg4+2R5cYhoQ/5ZNMuVi+YwM6NC4iMimb3d7/kSaEF2ZwlG3jt5Rbs2DCft99sz/QFazKs8+uJM/z51zU2r5rJltWzqFOzMqs37jBDtdYtO9c6KZk3GQ8AAApoSURBVCmZ0ZMX0u/dznyxbi6fr5lN9SrlzVCt9crOdX7ii6/2UczXKx+rK1iyc62dnRyZNnYQ2z6Zy8YV0zh/8Rr7DhwxQ7XWKzvXWX4nWoYsG5uaVSvg5Vn0Pzf++dffadKgFkXd3VAoFHRs14ofDx7P9SILssioaC5du8mLrRoD0KJJXcLCIwm+H5puPQUKUlNTSU5JxWAwEJ+QiJdn4X5t9rPK7rXe/8sxKlcoY2xmVColRdxc8r1ea5Xd6wxw8/Y9Dh87zTtd2ud3mQVCdq91+TIljM2jrY0NZUsF8iAsPN/rtVbZvc7yO9Ey5GgcMuzho3QjOr7enoQ9fJTjogqTh+GReLi7of77rZUKhQJvz6KEhae/jk0a1KRmtYq80nUQr7w1iFNn/6JPj07mKNlqZfda37p7HxuNmuET5tOj/1imzF3N46gYc5RslbJ7nbVaLbMWr2f0R++hklsizyW71/ppjyKj+OXI7zSuXzO/yrR62b3O8jvRMsh/TazEpau3uHn7Ht9sXcr/ti6jTs3KzP1X+JnIHTqdnpNn/2L0R++xaeUMPD2KMG/ZRnOXVeCs3/I1QY3rUKJ4MXOXUmjExycwctIC3u7cjorlSpm7HCHyRI4aG2+vooSGRRg/PwgLx9vrv29fifS8PN2JiIxCq9MBaRMpw8If4f2v24D7Dhyhdo1KODs5olQqebl1U06fu2SOkq1Wdq+1t2dRalWvhJeHOwqFgrYtG3Ph8nVzlGyVsnudz/55iV3f7KdjjyH0Gz6V+IREOvYYIqNjzyC71xogPiGRIePm0bRhbd564+X8LtWqZfu/HfI70SLkqLFp0aQeR06c4VFkFAaDga/3/kTr5g1yq7ZCwd3NlfJlSvDDT0cB+OXISbw83DM87VTM15PTf1wkNVULwNHfzlI6i6dMROaye61bNa/Ppas3iY9PAODYyT8oW6p4vtdrrbJ7nVcvnMjXm5fw9WeLWbNgIo4O9nz92WKZz/QMsnutExKTGDpuLg3qVKNXt9fMUKl1y+51lt+JliHLNw/PXrKeY7//QWRkNC4uTjjY27Hr04XMXPQJTRvUomnD2gDs+e4XNu/4HwA1q1Vk9OBe8mjbM7oTHML0BWuJjonD0cGeccP7UqZkQLprnZKSyoIVmzj311XUahVFi7gyavB78jTJM8rOtYa0EbItO75FoVTgWbQIYz7qLf/yegbZvc5PPAgNp8eAcfK493PIzrX+dOse1m35ilKB/9z2a9m0Pj27vWrGyq1Ldv9Oy+9E85NIBSGEEEIUGDJ5WAghhBAFhjQ2QgghhCgwpLERQgghRIEhjY0QQgghCgxpbIQQQghRYEhjI4QQQogCQxobIYQQQhQY0tgIIYQQosCQxkYIIYQQBYY0NkL87dCxU3TsMQSAbn1Gc+TEWfMWZOH27j/MtPlrzL4PIYR4mjQ2QmRi6ydzaNKgpsn1OvYYwqFjp/Khome3c89+eg2aQLNXejJ68iJzlyOEEPlCkrmEKKA8ixahZ7dXOXnmL8IjIs1djhBC5AtpbESh9TD8ETMWfsKFy9cJ8PMhqEld43cdewxhyAdv07xRHQC2ffkd27/+gdi4eFxcnOj11qucOH2esIePmDhrBSqlkhdbNmb0R++x7cvv+Orbn4h8HE0RNxe6dGxL51dfSLfvN15pzcGjJ7l15z7lypRg8qj+xvTw+PgEVm3cydHfzhATF0+gvy+zJgzB26soCYlJrFy/nSMnzpCSkkr9OtUYPrAHTo4OGc7vyflcu3En3xobrVbL+i1f8833B0lKSqZZo9qMHdoHjUb+UyOEyB/yXxtRaE2avRJfH0/2bltB6MNHDBs/L9P17t57wJpNu/h0+XRKFPcj8nE0kY+j6fBSiwwNEICPlwfL54zFy9OdM+cuMWzCPMqVKUH1yuWM63z/81HmTh5KUXc3Pp66hLWf7WLCiH4ATFuwlqSkZNYunkzRIq5cu3kXW1sbAGYsWItKpWLz6pmo1SpmLVrHghWbmDSqfx5eqeybvmAtDvb27Ny4gOiYOPoNm8L+X47R7oVm5i5NCFFISGMjCqWwh4/448IVZkwYjJ2dLSWK+9GxXUu+2vtThnWVSiUGg4Fbd+7h4+2BexFX3Iu4ZrnvFk3rGX+uXaMSDWpX4+y5S+kam9dfaY2fjxcAL7ZsxGfb/wdA5ONoDh09xdebF+NZtAgA5cuUAOBxVAwHj55k347VODs5AtCnRye69R3N+OH9UKmef8rc8AnzOfb7H1l+/9WmRfj6eP7nPv786ypXrt9my+rZqFRKHOztqFOjMrfu3n/uuoQQ4llJYyMKpYjIx9jYaHB3+6dB8fH2yHRdfz9vJozox65vfmT6wrVUqVCGge+/RbnSgZmu/8PPR9n65Xc8CI3AYDCQlJycoSko6v7Pce3sbElITALgQVgENhoNPl4Za3kQFo5eb+CNd4emW65UKHn0OAovD/fsnXwmpo4ZQKpWl+X3Ls6OJvdx8Ogpmjasna7Bio6Jo0zJ4s9dlxBCPCtpbESh5OFehJSUVCKjoo3NTdjDR1mu37p5A1o3b0BScgqffLaLKXNX8fma2SgVinTrhT6MYNq8NSycMYpa1SuiVqkYPXkRBoMhW3X5enuQkppK2MNHxjk3T3h7FkWpVPC/rcuws7N9xjP+b46ZzNF5VhcuXSOo8T/zlKJj4jhz7hIf9euOTqdn3vKN3L33AK1Wx8DeXalepXyOjymEEP8mj3uLQsnbqyjVKpdj5frtJCWncCc4hN3f/ZzpuneCQ/j99HmSklPQqNU42NuhUqkAcC/iyv2Qh8Z1ExOTMWDA3c0FpULBsd//4Lcz57Ndl3sRV5o1rM3cZRuIePQYvV7Pleu3iY6Jpai7G80a1mH+ik1ERccC8CgyioNHT2a6L61OR3JKCjqdHr1BT3JKCqmp2mzX8iwMBgM3bgXz64kzxMcn8CgyikmzV9C2VWOK+/ty/dZdEhKSWDlvPGsXTaJqpbJ5UocQQsiIjSi0powZwMyF62jXZQABxXx45YXm7Pn+lwzrpWp1rP1sF7fu3kepUFKmVHEmDO8LQI+uHVi0cjMbt+7mhRYNGflhL97t+iqDRs9Er9fTpEEtmjao9Ux1TRjZjxXrvqDXhxNJSEykREAxZk0YDMD4EX1Zt/lL3vtwItGxsbi7udK6eYN0IyVPfLp1N+u3fG38HNT+PWpWq8DKeeOfqZ7suBcShkajJqhxHd7oOQyNRsPLrZvw/jtvAFC6ZAD+ft6Mnb6UujUr07Fdq1yvQQghABQpKcnZGyMXQoin7N1/mDN/XmLCiH78dPg3dn3zI6vmZ940JSUlY2dni06np0vvEez6dGGGfQjx//bt3gRhIAzA8DdARhAiSvpUNu5gbeUETpA22LqAowUF7UScIBYW1kHQYPH5PO0dx5Uv9wNjcGIDfK07nmM2nQyO7/aHuN7u0feP2KxXP9wZ8G+EDfCRal5GUbweHXenSywX9eDcttm+XQNgDK6iAIA0/IoCANIQNgBAGsIGAEhD2AAAaQgbACANYQMApPEEGjrV2ATt+WgAAAAASUVORK5CYII=", "text/plain": [ "
" ] }, "metadata": { "image/png": { "alt": "Dendrogram of every feature in the matrix. Neighbouring horizons join at very small distances - the five- and ten-day returns with their risk-adjusted twins, the six-month return with its risk-adjusted twin, its skip-recent version and the 200-day trend ratio - and the bounded oscillators join that shortest-horizon block rather than standing apart from it. The nine- and twelve-month returns form a block of their own that stays separate from the six-month one at the cut and joins it only further out; the short-horizon and long-horizon blocks join later still. Volatility and the macro features attach as separate branches." } }, "output_type": "display_data" } ], "source": [ "clusters = plot_redundancy_clusters(\n", " features,\n", " feature_cols,\n", " cut=0.7,\n", " title=\"Adjacent horizons pair off; short and long momentum stay apart\",\n", " subtitle=r\"Average linkage on rank-correlation distance, cut drawn at $|\\rho| = 0.7$\",\n", " alt=(\n", " \"Dendrogram of every feature in the matrix. Neighbouring horizons join at very small \"\n", " \"distances - the five- and ten-day returns with their risk-adjusted twins, the six-month \"\n", " \"return with its risk-adjusted twin, its skip-recent version and the 200-day trend ratio \"\n", " \"- and the bounded oscillators join that shortest-horizon block rather than standing \"\n", " \"apart from it. The nine- and twelve-month returns form a block of their own that stays \"\n", " \"separate from the six-month one at the cut and joins it only further out; the \"\n", " \"short-horizon and long-horizon blocks join later still. Volatility and the macro \"\n", " \"features attach as separate branches.\"\n", " ),\n", ")" ] }, { "cell_type": "markdown", "id": "8b375388", "metadata": { "papermill": { "duration": 0.003316, "end_time": "2026-08-10T18:24:37.502395+00:00", "exception": false, "start_time": "2026-08-10T18:24:37.499079+00:00", "status": "completed" } }, "source": [ "### F6. Persistence and rank stability\n", "\n", "The right-hand panel compares the ordering across consecutive **rebalances**, which\n", "`config/setup.yaml` declares as `monthly_month_end` - a varying number of sessions apart, so\n", "a fixed lag would correlate dates the strategy never puts side by side.\n", "\n", "The autocorrelation on the left is of the feature, not of the return, and it runs past one\n", "full decision cycle. A feature whose value has decayed inside a cycle cannot support that\n", "rebalance cadence, however well it predicts on the day it is computed. It is estimated per ETF\n", "on pairs of dates exactly one lag apart and summarized by the median over ETFs, with a\n", "bootstrap interval over ETFs: a correlation pooled over every ETF-date pair would read high\n", "whenever ETFs sit at different levels, whether or not any one of them persists." ] }, { "cell_type": "code", "execution_count": 26, "id": "148dd4d4", "metadata": { "execution": { "iopub.execute_input": "2026-08-10T18:24:37.510553Z", "iopub.status.busy": "2026-08-10T18:24:37.510451Z", "iopub.status.idle": "2026-08-10T18:24:39.307867Z", "shell.execute_reply": "2026-08-10T18:24:39.307376Z" }, "papermill": { "duration": 1.802443, "end_time": "2026-08-10T18:24:39.308595+00:00", "exception": false, "start_time": "2026-08-10T18:24:37.506152+00:00", "status": "completed" } }, "outputs": [ { "data": { "image/png": "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", "text/plain": [ "
" ] }, "metadata": { "image/png": { "alt": "Two panels. On the left, autocorrelation against lag: the six-month return, the three-month volatility and the six-month risk-adjusted return decay slowly and are all still above 0.6 at 42 sessions, while the one-month return reaches zero at about 21 sessions - the length of its own window - and the 14-day oscillator falls below 0.1. The bootstrap ribbon around each curve is only a few hundredths wide. On the right, the cross-sectional rank correlation between consecutive rebalances separates the features sharply: about 0.98 for the three-month volatility and 0.83 for the two six-month features, against 0.2 for the oscillator and almost nothing for the one-month return, whose window is about one rebalance long." } }, "output_type": "display_data" } ], "source": [ "DECISION_DATES = (\n", " features.group_by(pl.col(\"timestamp\").dt.truncate(\"1mo\"))\n", " .agg(pl.col(\"timestamp\").max().alias(\"decision\"))[\"decision\"]\n", " .sort()\n", " .to_list()\n", ")\n", "\n", "plot_persistence(\n", " features,\n", " [\"ret_21d\", \"ret_126d\", \"sharpe_126d\", \"vol_63d\", \"rsi_14\"],\n", " entity=\"symbol\",\n", " max_lag=2 * DECISION_CYCLE,\n", " decision_dates=DECISION_DATES,\n", " title=\"Six-month features hold their ordering a month out; one-month ones do not\",\n", " subtitle=(\n", " f\"Median over ETFs to {2 * DECISION_CYCLE} sessions; rank correlation across rebalances\"\n", " ),\n", " alt=(\n", " \"Two panels. On the left, autocorrelation against lag: the six-month return, the \"\n", " \"three-month volatility and the six-month risk-adjusted return decay slowly and are all \"\n", " \"still above 0.6 at 42 sessions, while the one-month return reaches zero at about 21 \"\n", " \"sessions - the length of its own window - and the 14-day oscillator falls below 0.1. The \"\n", " \"bootstrap ribbon around each curve is only a few hundredths wide. On the right, the \"\n", " \"cross-sectional rank correlation between consecutive rebalances separates the features \"\n", " \"sharply: about 0.98 for the three-month volatility and 0.83 for the two six-month \"\n", " \"features, against 0.2 for the oscillator and almost nothing for the one-month return, \"\n", " \"whose window is about one rebalance long.\"\n", " ),\n", ")" ] }, { "cell_type": "markdown", "id": "a66e145e", "metadata": { "papermill": { "duration": 0.003928, "end_time": "2026-08-10T18:24:39.316472+00:00", "exception": false, "start_time": "2026-08-10T18:24:39.312544+00:00", "status": "completed" } }, "source": [ "## G. Emit\n", "\n", "The matrix goes to a parquet file, and beside it a small JSON file records what was written.\n", "Its purpose is to let a later stage tell whether the matrix it is reading is the one it was\n", "tested against, without re-reading the whole file: it holds a digest of the values, the row\n", "count, the key columns, and the digest of each input this notebook read.\n", "\n", "A digest is a short string computed from the contents, so that two files holding the same\n", "values get the same string and any change in a value gets a different one. Computing it over\n", "the values rather than over the bytes of the file is what makes it usable here: rewriting the\n", "parquet with the rows in another order, or with different compression, leaves the digest\n", "alone, while a single corrected feature value moves it. Recording the inputs' digests\n", "alongside it is what makes the record a chain - a downstream stage can see not only that the\n", "matrix changed but that it changed because its input did." ] }, { "cell_type": "code", "execution_count": 27, "id": "5290287e", "metadata": { "execution": { "iopub.execute_input": "2026-08-10T18:24:39.325211Z", "iopub.status.busy": "2026-08-10T18:24:39.325141Z", "iopub.status.idle": "2026-08-10T18:24:39.591313Z", "shell.execute_reply": "2026-08-10T18:24:39.590760Z" }, "papermill": { "duration": 0.271291, "end_time": "2026-08-10T18:24:39.591663+00:00", "exception": false, "start_time": "2026-08-10T18:24:39.320372+00:00", "status": "completed" } }, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "Wrote case_studies/etfs/features/financial.parquet\n" ] } ], "source": [ "FEATURES_DIR.mkdir(parents=True, exist_ok=True)\n", "record = write_artifact(\n", " features,\n", " FEATURES_DIR / \"financial.parquet\",\n", " keys=[\"symbol\", \"timestamp\"],\n", " written_by=\"case_studies/etfs/03_financial_features.py\",\n", " inputs={\n", " \"eligibility.csv\": value_digest(eligibility),\n", " \"load_etfs\": value_digest(prices),\n", " \"load_macro:dgs10-dgs2\": value_digest(yield_curve),\n", " },\n", ")\n", "print(f\"Wrote {display_path(FEATURES_DIR / 'financial.parquet')}\")" ] }, { "cell_type": "markdown", "id": "af9dc69d", "metadata": { "papermill": { "duration": 0.007608, "end_time": "2026-08-10T18:24:39.605020+00:00", "exception": false, "start_time": "2026-08-10T18:24:39.597412+00:00", "status": "completed" }, "tags": [ "results" ] }, "source": [ "The matrix carries **57 features** on **404,500 rows** across **99 ETFs**, from **2007-01-03**\n", "to **2025-12-31**, under content digest **9c02a41ef4364257**. Cutting the redundancy tree\n", "leaves **22 clusters**, so well over half the columns repeat an ordering another column\n", "already carries." ] }, { "cell_type": "code", "execution_count": 28, "id": "cc9190f7", "metadata": { "execution": { "iopub.execute_input": "2026-08-10T18:24:39.614105Z", "iopub.status.busy": "2026-08-10T18:24:39.614033Z", "iopub.status.idle": "2026-08-10T18:24:39.619445Z", "shell.execute_reply": "2026-08-10T18:24:39.618953Z" }, "papermill": { "duration": 0.010059, "end_time": "2026-08-10T18:24:39.619664+00:00", "exception": false, "start_time": "2026-08-10T18:24:39.609605+00:00", "status": "completed" } }, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "57 features, 404,500 rows, 99 ETFs\n", "2007-01-03 to 2025-12-31, digest 9c02a41ef4364257\n", "22 redundancy clusters\n" ] } ], "source": [ "print(f\"{len(feature_cols)} features, {len(features):,} rows, {features['symbol'].n_unique()} ETFs\")\n", "print(f\"{features['timestamp'].min()} to {features['timestamp'].max()}, digest {record['digest']}\")\n", "print(f\"{len(set(clusters.values()))} redundancy clusters\")" ] }, { "cell_type": "markdown", "id": "eacbdff0", "metadata": { "papermill": { "duration": 0.004058, "end_time": "2026-08-10T18:24:39.628209+00:00", "exception": false, "start_time": "2026-08-10T18:24:39.624151+00:00", "status": "completed" } }, "source": [ "## Key takeaways\n", "\n", "- **State the timing contract before writing the feature.** The configuration fixes each family's\n", " lookback and lag in the register and every window the construction runs, so the timing figure\n", " and the warmup assertion read those declarations rather than numbers retyped in the code.\n", "- **Check that later dates were not read, by rebuilding rather than by inspecting.**\n", " Constructing the panel a second time with the holdout withheld and comparing the two value\n", " by value catches any transform that fits across the sample, including the ones nobody\n", " thought to flag.\n", "- **Rank inside the date.** A percentile taken within one timestamp is comparable across dates\n", " in a way a raw level, whose distribution drifts with the period's volatility, is not.\n", "- **Read the matrix before modelling it.** Distribution, dispersion, redundancy and decay each\n", " rule out a use: a feature with no cross-sectional spread cannot rank, and one whose ordering\n", " decays inside the rebalance cycle cannot be traded at that cadence.\n", "\n", "### Known limitations\n", "\n", "- The cross-asset regime feature is one pair, SPY against TLT. It describes the equity-bond\n", " relationship and says nothing about the commodity and currency funds the universe also\n", " holds.\n", "- The eligibility gate is annual, so an ETF that lost liquidity in June stays in the\n", " cross-section until December.\n", "- The yield-curve features carry the configured one-session availability lag, but they read\n", " the revised Treasury history rather than the initial release. A value revised later is not\n", " the value the decision could have seen, whatever its timestamp says.\n", "- Every feature here is a rule written in advance. `04_model_based_features` adds the features\n", " that are themselves model outputs, where the rule is estimated from the data." ] } ], "metadata": { "jupytext": { "cell_metadata_filter": "tags,-all" }, "kernelspec": { "display_name": "Python 3 (ipykernel)", "language": "python", "name": "python3" }, "language_info": { "codemirror_mode": { "name": "ipython", "version": 3 }, "file_extension": ".py", "mimetype": "text/x-python", "name": "python", "nbconvert_exporter": "python", "pygments_lexer": "ipython3", "version": "3.14.3" }, "papermill": { "default_parameters": {}, "duration": 65.162464, "end_time": "2026-08-10T18:24:42.234229+00:00", "environment_variables": {}, "exception": null, "input_path": "~/ml4t/public-rerun-etfs/case_studies/etfs/03_financial_features.ipynb", "output_path": "~/ml4t/public-rerun-etfs/case_studies/etfs/03_financial_features.ipynb", "parameters": {}, "start_time": "2026-08-10T18:23:37.071765+00:00", "version": "2.7.0" }, "ml4t_provenance": { "source_py_blob": "2bb88be846a9b69fe6d6badf96c0a0f77a439aa3", "executed_at": "2026-08-10T18:24:48.178974+00:00", "executor": "local-uv", "production": true, "parameters": {} } }, "nbformat": 4, "nbformat_minor": 5 }