{
"cells": [
{
"cell_type": "markdown",
"metadata": {
"id": "xXvzVBhbupcT"
},
"source": [
"# Optimizing Risks for a Portfolio of Cryptocurrencies\n",
"> \"Intro into Optimizing Risks for a portfolio of cryptocurrencies using stochastic discount factor\"\n",
"\n",
"- author: Vladimir Eremin, Dmytro Karabash, Maxim Korotkov \n",
"- categories: [sdf, treasury, crypto]\n",
"- image: images/pixabay-game-of-thrones-4180794_640.jpg\n",
"- permalink: /cryptosdf/"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {
"id": "P6fjUwxHLMe4"
},
"outputs": [],
"source": [
"#install requirements if needed\n",
"!pip install cvxpy==1.0.31 pandas_datareader pandas matplotlib seaborn -q"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {
"colab": {
"base_uri": "https://localhost:8080/"
},
"id": "MrPfwTmGLJUc",
"outputId": "5338cdf9-e28e-4e87-8237-1625e42157b6"
},
"outputs": [],
"source": [
"#import modules\n",
"import cvxpy as cp\n",
"import pandas_datareader as pdr\n",
"import numpy as np\n",
"import matplotlib.pyplot as plt\n",
"\n",
"import pandas as pd\n",
"import math\n",
"import os.path\n",
"import time\n",
"\n",
"#select plotting style\n",
"import matplotlib\n",
"import seaborn as sb\n",
"matplotlib.style.use('fivethirtyeight')\n",
"\n",
"#seaborn-white fivethirtyeight\n",
"import sys\n",
"#Python versiond\n",
"print(sys.version)\n",
"\n",
"print(cp.__version__)"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {
"id": "LLJQtt2JR9Yi"
},
"outputs": [],
"source": [
"#read input dataset that was downladed from the binance \n",
"fuldf = pd.read_pickle('data.pickle')\n",
"y_label = fuldf.columns[0]\n",
"factors = (fuldf.columns[1:]\n",
" .tolist()\n",
" )\n",
"\n",
"\n",
"total_size = fuldf.shape[0]\n",
"# print('Combined DF size:', total_size,'\\nFull range: ', fuldf.iloc[0].name,fuldf.iloc[-1].name)\n",
"train_set = .8\n",
"\n",
"train_id = int( total_size * train_set)\n",
"df_train = fuldf.iloc[:train_id]\n",
"df_test = fuldf.iloc[train_id:]\n",
"split_index = fuldf.iloc[train_id].name\n",
"# print('Training period: ', df_train.iloc[0].name,'-',df_train.iloc[-1].name,'shape:',df_train.shape[0],'\\nTesing period:',df_test.iloc[0].name,df_test.iloc[-1].name,'shape:',df_test.shape[0],)\n"
]
},
{
"cell_type": "markdown",
"metadata": {
"id": "GNz7_FGEQj3R"
},
"source": [
"## Data Acquisition\n",
"\n",
"---\n",
"\n",
"The data for this exercise was obtained from the Binance API using the Python API client `python-binance`.\n",
"\n",
"An example of the data acquisition code:\n",
">Note: Install the module if needed using pip:\n",
">#pip install python-binance\n",
"\n",
"```python\n",
"from binance.client import Client\n",
"binance_api_key = 'YOUR-API-KEY'\n",
"binance_api_secret = 'YOUR-API-SECRET'\n",
"binance_client = Client(api_key=binance_api_key\n",
"\t\t\t\t,api_secret=binance_api_secret)\n",
"klines = binance_client.get_historical_klines(symbol\n",
"\t\t\t\t ,kline_size, date_from, date_to)\n",
"data = pd.DataFrame(klines, columns = [COLUMNS])\n",
"```\n",
"\n",
"Following daily pairs was downloaded and formatted to pd DataFrame:\n",
"1. BTCUSDT\n",
"2. ETHUSDT\n",
"3. BNBUSDT\n",
"4. LTCUSDT\n",
"\n",
"For each timepoint, Binance provides conventional OHLC (Open, High, Low, Close) and Volume data. In this exercise, we used only the `Close` column. It's possible, though to consider the combination of all five values and came up with a more reliable metric, for example: the weighted average price.\n",
"\n",
"The joint time-series of four crypto assets looks the following:"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {
"colab": {
"base_uri": "https://localhost:8080/",
"height": 237
},
"id": "GItfRqgqkC1b",
"outputId": "2e185e13-b114-4bbe-cc16-2f3fb6686b35"
},
"outputs": [],
"source": [
"df_train.head()"
]
},
{
"cell_type": "markdown",
"metadata": {
"id": "oOOWsE-4MS2g"
},
"source": [
"Each trading pair has a different amount of data available: The oldest tradable crypto instruments on Binance are BTCUSDT and ETHUSDT - data points are available from _2017-08-17_. For BNBUSDT and LTCUSDT first trading days are _2017-11-06_ and _2017-12-13_, respectively.\n",
"\n",
"We joined these time series together on the earliest common trading date: _2017-12-13_ up to _2021-06-14_. The full joint dataset size is equal to 1280 samples.\n",
"\n",
"In order to properly infer the SDF, We split the dataset into training and testing sets at _2020-10-02_.\n",
"\n",
"Training interval: _2017-12-13_ to _2020-10-02_ (1024 data points).\n",
"\n",
" \n",
"\n",
"Below is the joint plot of the full dataset in the logarithmic scale. The dotted line represents the train-test split at _2020-10-02_"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {
"colab": {
"base_uri": "https://localhost:8080/",
"height": 516
},
"id": "KrT57ZAljK45",
"outputId": "153addb5-9a71-4f89-c7a5-08670d2cb1a7"
},
"outputs": [],
"source": [
"plot_data =np.log(fuldf)\n",
"fig , ax = plt.subplots(figsize=(12,7))\n",
"# ax1 = ax.twinx()\n",
"# plot_data.drop(y_label,axis=1).plot(ax=ax)\n",
"plot_data.plot(ax=ax,alpha=0.8)\n",
"ax.legend(loc=0)\n",
"ax.set(title='Assets',xlabel='Date',ylabel='price, log scale');\n",
"\n",
"ax.axvline(split_index,color='grey', linestyle='--', lw=2);"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {
"id": "oldD9O-nW1ds"
},
"outputs": [],
"source": [
"#Transform raw data to log-return format:\n",
"lret_data = np.log1p(df_train.pct_change()).dropna(axis=0,how='any')"
]
},
{
"cell_type": "markdown",
"metadata": {
"id": "tCjpd5h6W51t"
},
"source": [
"## II. Data Transformation\n",
"---\n",
"In order to be properly trained, input time series has to be transfored into the log-return format:\n",
"\n",
"$r_t = \\ln{\\frac{P_t}{P_{t-1}}}$\n",
"\n",
"Where $P_t$ is a price of the asset at time $t$.\n",
"\n",
"The SDF model needs two vectors to perform the optimization: \n",
"\n",
"1. Feature vector $I$\n",
"2. Target vector $R$\n",
"\n",
"The $R$ vector contains values shifted to $t+1$ compared to the $I$ vector:\n",
"\n",
"$R = \\begin{pmatrix}\n",
"r_{1}\n",
"\\\\ \\vdots\n",
"\\\\r_{t}\n",
"\\end{pmatrix}\n",
"I = \\begin{pmatrix}\n",
"r_{0,0} & \\dots & r_{0,l} \\\\ \n",
"\\vdots & & \\vdots \\\\ \n",
"r_{t-1,0} & \\dots & r_{t-1,l} \n",
"\\end{pmatrix}\n",
"$\n",
"\n",
"Where $l$ is the number of features in the model.\n",
"\n"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {
"colab": {
"base_uri": "https://localhost:8080/"
},
"id": "hW48zI-ILJUf",
"outputId": "b501e795-14ec-41b5-9fdf-e8ebb106b90b"
},
"outputs": [],
"source": [
"R = lret_data[y_label].shift(1).iloc[1:].values.reshape((-1,1))\n",
"I = lret_data[factors].iloc[1:].values\n",
"# assert R.shape[0] == I.shape[0] and I.shape[1]==3\n",
"\n",
"print('R:',R.shape,'\\n',R[:10]\n",
",'\\n I:',I.shape,'\\n',I[:10,:])"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {
"colab": {
"base_uri": "https://localhost:8080/",
"height": 175
},
"id": "mov7bsEVb5Fb",
"outputId": "a534fa17-1817-4635-e79a-dad9d07f521d"
},
"outputs": [],
"source": [
"#correlation matrix\n",
"F= pd.concat([pd.DataFrame(R),pd.DataFrame(I)],axis=1)\n",
"F.columns = [y_label]+factors\n",
"F.cov()"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {
"colab": {
"base_uri": "https://localhost:8080/",
"height": 650
},
"id": "pYqE9uELlNzj",
"outputId": "676f58b4-81e1-4acf-da45-28cac12ffa5c"
},
"outputs": [],
"source": [
"f = plt.figure(figsize=(10, 10))\n",
"cov_data = F.cov()\n",
"mask = np.triu(np.ones_like(cov_data))\n",
"dataplot = sb.heatmap(cov_data.corr(), cmap=\"YlGnBu\", annot=True, mask=mask)\n",
"plt.title('Covariance Matrix', fontsize=16);"
]
},
{
"cell_type": "markdown",
"metadata": {
"id": "lfEcVEmJapAz"
},
"source": [
"## Model \n",
"\n",
"---\n",
"\n",
"Our goal is to explain the differences in the cross-section of returns $R$ for individual stocks.\n",
"\n",
"Let $R_{t+1}$; denote the return of BTCUSDT at time $t + 1$. The fundamental no-arbitrage assumption is\n",
"equivalent to the existence of a stochastic discount factor (SDF) $M_{t+1}$ such that for any return in excess of the risk-free rate \n",
"\n",
"$R^e_{t+1} = R_{t+1}- R^f_{t+1}$ it holds:\n",
"\n",
"$E_t[M_{t+1} R^e_{t+1}]= 0$\n",
"\n",
"\n",
"For demonstration purpose and in seak of simplicity, we assume that\n",
"$R^e_{t} = R_{t}$: excesses return of the BTCUSDT is the same as \n",
"the market return.\n",
"\n",
"We consider the SDF formulation as:\n",
"\n",
"$M_{t+1} = 1 - \\sum_{i=1}^N w_{t,i} R_{t+1} = 1 -w_t^T R_{t+1}$\n",
"\n",
"\n",
"The Generalized Method of Moments (GMM) was used to approximate the SDF $M$ vector. \n",
"The first moment condition equation can be expressed as following:\n",
"\n",
"$ \\underset{M}{\\text{min}}\\underset{t}{\\sum} || M(I_{t}) R^e_{t+1}||^2 $\n",
"\n",
"Since the above problem is convex, we can estimate the SDF for BTC using the `cvxpy` framework."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {
"colab": {
"base_uri": "https://localhost:8080/"
},
"id": "HFAjd_WHLJUf",
"outputId": "a6b57e33-0fa4-479d-85f1-072d2b2030e2"
},
"outputs": [],
"source": [
"#4 - Solve GMM problem\n",
"n = R.shape[0]\n",
"#define weights for the SDF\n",
"w = cp.Variable(shape=(I.shape[1]\n",
" ,1),nonneg=True)\n",
"\n",
"#define SDF \n",
"M = I @ w\n",
"\n",
"#Define expression\n",
"Sigma = M @ R.T\n",
"\n",
"#Construct the problem\n",
"prob = cp.Problem(cp.Minimize( cp.norm(Sigma) )\n",
" # , [cp.geo_mean(M) >= 1 , ]\n",
" , [cp.sum(w) == 1 \n",
" ,cp.sum(M) == 1 ]\n",
" \n",
" )\n",
"\n",
"prob.solve(verbose=True\n",
" #, max_iters = 300\n",
" )\n",
"\n",
"print('SDF weights from BTC:\\n',dict(zip(factors,w.value)))"
]
},
{
"cell_type": "code",
"execution_count": 13,
"metadata": {
"id": "sWfKMQpnWXgh"
},
"outputs": [],
"source": [
"def sharpe(x):\n",
" return (x.mean() / x.std() * np.sqrt(365))\n",
"\n",
"\n",
"\n",
"def plot_return(data,title):\n",
" datac = data.copy()\n",
" fig , ax = plt.subplots(figsize=(13,5))\n",
"\n",
" y_return = (datac[y_label].pct_change().fillna(0)+1)\n",
" \n",
" p_return = ((datac[factors].pct_change().fillna(0) + 1)\n",
" .apply(lambda x: (x @ w.value)[0] ,axis =1)\n",
" .rename('SDF Portfolio')\n",
" )\n",
" \n",
" \n",
" \n",
" p_return_c = p_return.cumprod()\n",
" y_return_c = y_return.cumprod()\n",
" portfolio_benchmark = pd.concat([y_return_c,p_return_c],axis=1)\n",
"\n",
" portfolio_benchmark.plot(ax=ax)\n",
" ax.legend(loc=0)\n",
" ax.set(title=title,ylabel='growth factor',xlabel='date')\n",
" \n",
" return portfolio_benchmark\n",
"\n",
"def calc_metrics(data):\n",
" #y_return = data[y_label].pct_change().fillna(0)\n",
" \n",
" p_return = ((data[factors].pct_change().fillna(0))\n",
" .apply(lambda x: (x @ w.value)[0] ,axis =1)\n",
" .rename('SDF Portfolio')\n",
" )\n",
" y_return = data.pct_change().fillna(0)\n",
" y_return['SDF Portfolio'] = p_return\n",
" \n",
" #portfolio_ret = pd.concat([y_return,p_return],axis=1)\n",
" \n",
" return '\\nSharpe ration: '+str(portfolio_ret.apply(sharpe).round(3).to_dict()) \n",
" "
]
},
{
"cell_type": "markdown",
"metadata": {
"id": "ym4-09mgHS-i"
},
"source": [
"## Results\n",
"Below is the result of computing the SDF portfolio on the training dataset and plotting it against the index portfolio (BTCUSDT):"
]
},
{
"cell_type": "code",
"execution_count": 14,
"metadata": {
"colab": {
"base_uri": "https://localhost:8080/",
"height": 450
},
"id": "qSUOksp2bafc",
"outputId": "a4247541-1e12-4567-ed2a-6245c9a0c24b"
},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"Portfolio growth (x times):\n",
"{'BTCUSDT': -0.34, 'SDF Portfolio': 2.92}\n",
"Sharpe ration: {'BTCUSDT': 0.222, 'SDF Portfolio': 0.997}\n"
]
},
{
"data": {
"image/png": "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\n",
"text/plain": [
"