diff --git a/KDB-X/Python/KDB-X_Python_Backtest.ipynb b/KDB-X/Python/KDB-X_Python_Backtest.ipynb new file mode 100644 index 0000000..f88498b --- /dev/null +++ b/KDB-X/Python/KDB-X_Python_Backtest.ipynb @@ -0,0 +1,1376 @@ +{ + "cells": [ + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "# KDB-X Python Rapid Backtesting Demo\n", + "\n", + "This notebook demonstrates KDB-X speed advantages for Python-based backtesting.\n", + "\n", + "## Key Features:\n", + "- Performance comparison with pandas\n", + "- Vectorized time-series operations\n", + "- Multi-strategy backtesting\n", + "- Risk metrics and transaction costs" + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "### Prerequisites and Dependencies\n", + "\n", + "- Requires KDB-X to be installed, you can sign up and download on [Developer Center](https://developer.kx.com/products/kdb-x/install). For full install instructions see: [KDB-X Install](https://code.kx.com/kdb-x/).\n", + "\n", + "To Install KDB-X Python: `pip install --upgrade --pre pykx`\n", + "\n", + "\n", + "We start by importing the necessary libraries and initializing the PyKX environment. Setting a random seed in both q (via kx.q) and Python ensures that our generated market data is reproducible across different runs." + ] + }, + { + "cell_type": "code", + "execution_count": 2, + "metadata": {}, + "outputs": [ + { + "name": "stdout", + "output_type": "stream", + "text": [ + "\n", + "Welcome to KDB-X Community Edition!\n", + "For Community support, please visit https://kx.com/slack\n", + "Tutorials can be found at https://github.com/KxSystems/tutorials\n", + "Ready to go beyond the Community Edition? Email preview@kx.com\n", + "\n", + "PyKX version: 4.0.0b4\n", + "NumPy version: 2.2.6\n", + "Pandas version: 2.3.0\n" + ] + } + ], + "source": [ + "import pykx as kx\n", + "import pandas as pd\n", + "import numpy as np\n", + "import time\n", + "import matplotlib.pyplot as plt\n", + "\n", + "# Set random seed for reproducibility\n", + "kx.q(\"\\\\S 42\")\n", + "np.random.seed(42)\n", + "\n", + "print(f\"PyKX version: {kx.__version__}\")\n", + "print(f\"NumPy version: {np.__version__}\")\n", + "print(f\"Pandas version: {pd.__version__}\")" + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "## 1. Market Data Generation\n", + "\n", + "To demonstrate performance, we need a significant dataset. Here, we use an embedded q function to generate 750,000+ trades across 10 ticker symbols for a full trading year.\n", + "\n", + "The generation uses a random walk for price action and assigns random quantities, spreads, and intraday timestamps. Note how KDB-X Python handles this large dataset with a very small memory footprint (~32MB)." + ] + }, + { + "cell_type": "code", + "execution_count": 3, + "metadata": {}, + "outputs": [ + { + "name": "stdout", + "output_type": "stream", + "text": [ + "Generating 10M trades with random price movements...\n", + "Generated 756,392 trades in 0.20 seconds\n", + "Data generation rate: 3,834,592 trades/second\n", + "Memory size: 32.4 MB\n", + "\n", + "Price ranges by symbol:\n", + "sym min_px max_px \n", + "----------------------\n", + "aapl 174.5906 294.3492\n", + "adbe 557.9294 757.5484\n", + "amzn 139.1402 166.5716\n", + "crm 157.1418 245.3633\n", + "goog 120.3842 195.7222\n", + "meta 281.9218 357.7852\n", + "msft 350.4205 469.8667\n", + "nflx 432.8493 557.742 \n", + "nvda 499.5475 715.0795\n", + "tsla 220.6355 413.8282\n", + "\n", + "Sample trades:\n", + "dt tm sym qty px volume spread \n", + "--------------------------------------------------------------\n", + "2024.01.02 09:31:13.102 aapl 800 179.3415 143473.2 0.01179341\n", + "2024.01.02 09:32:36.939 aapl 400 179.3443 71737.71 0.01179344\n", + "2024.01.02 09:33:47.259 aapl 500 179.2813 89640.63 0.01179281\n", + "2024.01.02 09:34:43.145 aapl 1300 179.4135 233237.6 0.01179414\n", + "2024.01.02 09:35:35.814 aapl 1400 179.372 251120.8 0.01179372\n", + "2024.01.02 09:36:29.612 aapl 600 179.3761 107625.7 0.01179376\n", + "2024.01.02 09:37:53.299 aapl 1200 179.3658 215238.9 0.01179366\n", + "2024.01.02 09:38:28.375 aapl 600 179.3505 107610.3 0.01179351\n", + "2024.01.02 09:38:50.769 aapl 1400 179.286 251000.3 0.01179286\n", + "2024.01.02 09:39:04.880 aapl 500 179.3264 89663.2 0.01179326\n" + ] + } + ], + "source": [ + "# Generate large-scale market data with random price paths using q\n", + "print(\"Generating 10M trades with random price movements...\")\n", + "start_time = time.time()\n", + "\n", + "trades = kx.q(\"\"\"\n", + "{[]\n", + " trading_days:252;\n", + " symbols:`aapl`goog`msft`amzn`tsla`nvda`meta`nflx`adbe`crm;\n", + " base_date:2024.01.02;\n", + " \n", + " // Base prices by symbol\n", + " base_prices:`aapl`goog`msft`amzn`tsla`nvda`meta`nflx`adbe`crm!180 140 380 155 245 495 355 485 575 245f;\n", + " \n", + " // Generate random price paths for each symbol with symbol-specific drift\n", + " price_data:raze {[ticker; base_date; trading_days; base_px; idx]\n", + " dates:base_date + til trading_days;\n", + " \n", + " // Zero drift - true random walk with no bias\n", + " drift:0.0;\n", + " \n", + " // Generate daily returns as true random walk with zero mean and realistic volatility\n", + " // Using Box-Muller transform for normal distribution: mean=0, std=1.5%\n", + " u1:trading_days ? 1.0;\n", + " u2:trading_days ? 1.0;\n", + " returns:0.015 * sqrt[-2.0 * log u1] * cos[2.0 * 3.141592653589793 * u2];\n", + " prices:base_px * exp sums returns;\n", + " \n", + " // Generate 100-500 trades per day for ~10M trades total\n", + " trades_per_day:100 + trading_days ? 400;\n", + " \n", + " // Create intraday trades for each day\n", + " daily_data:raze {[dt; px; num_trades; ticker]\n", + " // Random times throughout the day\n", + " time_offsets:09:30:00.000 + num_trades ? 06:30:00.000;\n", + " \n", + " // Intraday price noise (0.1% std dev)\n", + " price_noise:1.0 + 0.001 * (num_trades ? 1.0) - 0.0005;\n", + " trade_prices:px * price_noise;\n", + " \n", + " // Random quantities\n", + " qtys:100 * 1 + num_trades ? 19;\n", + " \n", + " ([] dt:num_trades # dt; \n", + " tm:time_offsets; \n", + " sym:num_trades # ticker; \n", + " qty:qtys; \n", + " px:trade_prices)\n", + " }[;;; ticker]'[dates; prices; trades_per_day];\n", + " \n", + " daily_data\n", + " }[; base_date; trading_days;;] ./: flip (symbols; base_prices symbols; til count symbols);\n", + " \n", + " // Sort by symbol, date, time\n", + " price_data:`sym`dt`tm xasc price_data;\n", + " \n", + " // Add volume and spread\n", + " price_data:update volume:qty*px, spread:0.01*1+px%1000 from price_data;\n", + " \n", + " price_data\n", + "}[]\n", + "\"\"\")\n", + "\n", + "generation_time = time.time() - start_time\n", + "print(f\"Generated {len(trades):,} trades in {generation_time:.2f} seconds\")\n", + "print(f\"Data generation rate: {len(trades)/generation_time:,.0f} trades/second\")\n", + "\n", + "# Calculate memory size\n", + "memory_mb = float(kx.q('{sum -22!x}', trades)) / (1024**2)\n", + "print(f\"Memory size: {memory_mb:.1f} MB\")\n", + "\n", + "print(f\"\\nPrice ranges by symbol:\")\n", + "price_ranges = kx.q.qsql.select(\n", + " trades,\n", + " columns={'min_px': 'min px', 'max_px': 'max px'},\n", + " by=kx.Column('sym')\n", + ")\n", + "print(kx.q('0!', price_ranges))\n", + "\n", + "print(\"\\nSample trades:\")\n", + "print(trades.head(10))" + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "## 2. Performance Comparison: PyKX vs Pandas\n", + "\n", + "A common bottleneck in Python backtesting is calculating OHLCV (Open, High, Low, Close, Volume) bars. In this block, we compare the speed of PyKX's vectorized select against the standard pandas groupby and agg approach.\n", + "\n", + "For time-series aggregations like VWAP (Volume Weighted Average Price), KDB-X Python typically performs significantly faster." + ] + }, + { + "cell_type": "code", + "execution_count": 4, + "metadata": {}, + "outputs": [ + { + "name": "stdout", + "output_type": "stream", + "text": [ + "Converting to pandas for performance comparison...\n", + "Conversion to pandas: 0.06 seconds\n", + "\n", + "=== Performance Test: Daily OHLCV Aggregation ===\n", + "PyKX OHLCV: 0.048 seconds\n", + "Pandas OHLCV: 0.147 seconds\n", + "PyKX is 3.0x faster\n", + "\n", + "Result shapes: PyKX=2,520, Pandas=2,520\n" + ] + } + ], + "source": [ + "# Convert to pandas for comparison\n", + "print(\"Converting to pandas for performance comparison...\")\n", + "start_time = time.time()\n", + "trades_df = trades.pd()\n", + "conversion_time = time.time() - start_time\n", + "print(f\"Conversion to pandas: {conversion_time:.2f} seconds\")\n", + "\n", + "# Performance test: Daily OHLCV calculation\n", + "print(\"\\n=== Performance Test: Daily OHLCV Aggregation ===\")\n", + "\n", + "# PyKX version\n", + "start_time = time.time()\n", + "daily_ohlcv_pykx = kx.q.qsql.select(\n", + " trades,\n", + " columns={\n", + " \"open\": \"first px\",\n", + " \"high\": \"max px\",\n", + " \"low\": \"min px\", \n", + " \"close\": \"last px\",\n", + " \"volume\": \"sum volume\",\n", + " \"trade_count\": \"count px\",\n", + " \"vwap\": \"qty wavg px\"\n", + " },\n", + " by=[\"dt\", \"sym\"]\n", + ")\n", + "pykx_time = time.time() - start_time\n", + "\n", + "# Pandas version (using standard aggregation methods)\n", + "start_time = time.time()\n", + "# Calculate VWAP alongside other aggregations\n", + "trades_df['notional'] = trades_df['px'] * trades_df['qty']\n", + "daily_ohlcv_pandas = trades_df.groupby(['dt', 'sym']).agg({\n", + " 'px': ['first', 'max', 'min', 'last', 'count'],\n", + " 'volume': 'sum',\n", + " 'notional': 'sum',\n", + " 'qty': 'sum'\n", + "})\n", + "# Calculate VWAP from aggregated notional and qty\n", + "daily_ohlcv_pandas['vwap'] = daily_ohlcv_pandas[('notional', 'sum')] / daily_ohlcv_pandas[('qty', 'sum')]\n", + "pandas_time = time.time() - start_time\n", + "\n", + "print(f\"PyKX OHLCV: {pykx_time:.3f} seconds\")\n", + "print(f\"Pandas OHLCV: {pandas_time:.3f} seconds\")\n", + "print(f\"PyKX is {pandas_time/pykx_time:.1f}x faster\")\n", + "\n", + "print(f\"\\nResult shapes: PyKX={len(daily_ohlcv_pykx):,}, Pandas={len(daily_ohlcv_pandas):,}\")" + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "## 3. As-Of Join: Enriching Tick Data with Daily Context\n", + "\n", + "The \"As-Of Join\" (aj) is a specialized KDB-X operation essential for quantitative finance. It joins two tables based on the most recent available data at a specific point in time.\n", + "\n", + "In this tutorial, we use aj to enrich individual tick trades with the context of the daily bar they belong to (e.g., comparing the current trade price to the daily open or daily high)." + ] + }, + { + "cell_type": "code", + "execution_count": 5, + "metadata": {}, + "outputs": [ + { + "name": "stdout", + "output_type": "stream", + "text": [ + "Using as-of join to add daily context to trades...\n", + "As-of join enrichment: 0.07 seconds\n", + "Enriched 10,000 trades with daily bar context\n", + "\n", + "=== Sample Enriched Trades ===\n", + "Shows each trade with its daily OHLC context:\n", + "sym dt tm px daily_open daily_close daily_vwap pct_o..\n", + "-----------------------------------------------------------------------------..\n", + "adbe 2024.05.30 12:02:24.038 697.3722 697.7697 697.428 697.4834 0.346..\n", + "crm 2024.06.18 15:19:39.508 194.1373 194.1026 194.2327 194.1996 0.185..\n", + "nflx 2024.07.02 09:31:20.102 473.2289 473.2289 473.5623 473.4592 0.015..\n", + "tsla 2024.01.03 10:54:02.243 239.7663 239.6856 239.7104 239.6911 0.800..\n", + "nflx 2024.04.09 14:44:49.434 541.4235 541.47 541.5775 541.6115 0.131..\n", + "adbe 2024.07.23 10:13:14.785 742.3784 742.0977 741.8272 742.09 0.891..\n", + "msft 2024.02.26 11:12:50.762 393.5448 393.3761 393.5939 393.5221 0.559..\n", + "meta 2024.01.31 15:27:17.847 329.9014 329.8243 329.8422 329.9764 0.264..\n", + "adbe 2024.04.09 11:22:19.657 678.693 678.6092 678.647 678.589 0.658..\n", + "amzn 2024.04.04 13:14:48.849 151.8954 151.8417 151.8331 151.8419 0.877..\n", + "aapl 2024.02.01 13:11:14.347 190.8734 190.7971 190.7609 190.8442 0.636..\n", + "nvda 2024.05.02 10:35:39.331 678.9987 678.7996 679.2878 679.0217 0.456..\n", + "crm 2024.07.17 15:58:52.463 174.7124 174.6326 174.7124 174.6989 0.557..\n", + "adbe 2024.08.01 10:03:15.286 694.9419 694.5803 694.6276 694.6758 0.871..\n", + "nvda 2024.08.18 15:30:31.359 639.1568 639.0937 638.934 639.0117 0.715..\n", + "amzn 2024.04.17 13:48:28.353 148.5213 148.4171 148.4974 148.4505 0.985..\n", + "goog 2024.03.30 15:26:01.272 132.7379 132.7684 132.8381 132.8081 0.012..\n", + "aapl 2024.08.04 10:17:08.240 255.5762 255.749 255.7536 255.6289 0.289..\n", + "nvda 2024.06.10 11:50:04.652 622.7298 623.0841 622.8707 622.986 0.089..\n", + "goog 2024.07.06 15:29:47.043 180.5809 180.4928 180.4725 180.5297 0.804..\n", + "\n", + "=== Performance: PyKX aj vs Pandas merge_asof ===\n", + "PyKX aj: 0.025 seconds\n", + "Pandas merge_asof: 0.105 seconds\n", + "PyKX is 4.2x faster\n", + "\n", + "=== Use Cases for As-Of Join ===\n", + "- Enrich tick trades with daily/minute bar context\n", + "- Join trades with point-in-time reference data\n", + "- Add lagged features for ML models\n", + "- Calculate trade execution quality vs benchmarks\n" + ] + } + ], + "source": [ + "# Demonstrate as-of join to enrich tick trades with daily bar context\n", + "print(\"Using as-of join to add daily context to trades...\")\n", + "start_time = time.time()\n", + "\n", + "# First, create daily OHLC bars\n", + "daily_bars = kx.q.qsql.select(\n", + " trades,\n", + " columns={\n", + " 'daily_open': 'first px',\n", + " 'daily_high': 'max px',\n", + " 'daily_low': 'min px',\n", + " 'daily_close': 'last px',\n", + " 'daily_vwap': 'qty wavg px',\n", + " 'daily_volume': 'sum volume'\n", + " },\n", + " by=['dt', 'sym']\n", + ")\n", + "\n", + "# Unkey for as-of join\n", + "daily_bars_unkeyed = kx.q('0!', daily_bars)\n", + "\n", + "# Sample 10K trades for demonstration\n", + "sample_trades = kx.q('{-10000?x}', trades)\n", + "\n", + "# As-of join: attach daily bar context to each trade\n", + "# aj takes: key columns, left table, right table\n", + "enriched_trades = kx.q('aj', \n", + " [kx.SymbolAtom('sym'), kx.SymbolAtom('dt')], \n", + " sample_trades, \n", + " daily_bars_unkeyed)\n", + "\n", + "# Add derived columns showing trade context\n", + "enriched_trades = kx.q.qsql.update(\n", + " enriched_trades,\n", + " columns={\n", + " 'pct_of_day_range': '(px - daily_low) % (daily_high - daily_low)',\n", + " 'vs_vwap': '(px - daily_vwap) % daily_vwap',\n", + " 'vs_open': '(px - daily_open) % daily_open'\n", + " }\n", + ")\n", + "\n", + "aj_time = time.time() - start_time\n", + "print(f\"As-of join enrichment: {aj_time:.2f} seconds\")\n", + "print(f\"Enriched {len(enriched_trades):,} trades with daily bar context\")\n", + "\n", + "print(\"\\n=== Sample Enriched Trades ===\")\n", + "print(\"Shows each trade with its daily OHLC context:\")\n", + "sample_display = kx.q.qsql.select(\n", + " enriched_trades.head(20),\n", + " columns=['sym', 'dt', 'tm', 'px', 'daily_open', 'daily_close', 'daily_vwap', 'pct_of_day_range', 'vs_vwap']\n", + ")\n", + "print(sample_display)\n", + "\n", + "# Performance comparison with pandas merge_asof\n", + "print(\"\\n=== Performance: PyKX aj vs Pandas merge_asof ===\")\n", + "\n", + "# PyKX as-of join\n", + "start_time = time.time()\n", + "enriched_pykx = kx.q('aj', \n", + " [kx.SymbolAtom('sym'), kx.SymbolAtom('dt')], \n", + " sample_trades, \n", + " daily_bars_unkeyed)\n", + "pykx_aj_time = time.time() - start_time\n", + "\n", + "# Pandas merge_asof - merge each symbol group separately then concatenate\n", + "start_time = time.time()\n", + "sample_trades_pd = sample_trades.pd()\n", + "daily_bars_pd = daily_bars_unkeyed.pd()\n", + "\n", + "# Merge by symbol groups\n", + "merged_groups = []\n", + "for sym in sample_trades_pd['sym'].unique():\n", + " left_sym = sample_trades_pd[sample_trades_pd['sym'] == sym].sort_values('dt')\n", + " right_sym = daily_bars_pd[daily_bars_pd['sym'] == sym].sort_values('dt')\n", + " merged = pd.merge_asof(left_sym, right_sym, on='dt', by='sym')\n", + " merged_groups.append(merged)\n", + "\n", + "enriched_pandas = pd.concat(merged_groups, ignore_index=True)\n", + "pandas_asof_time = time.time() - start_time\n", + "\n", + "print(f\"PyKX aj: {pykx_aj_time:.3f} seconds\")\n", + "print(f\"Pandas merge_asof: {pandas_asof_time:.3f} seconds\")\n", + "print(f\"PyKX is {pandas_asof_time/pykx_aj_time:.1f}x faster\")\n", + "\n", + "print(\"\\n=== Use Cases for As-Of Join ===\")\n", + "print(\"- Enrich tick trades with daily/minute bar context\")\n", + "print(\"- Join trades with point-in-time reference data\")\n", + "print(\"- Add lagged features for ML models\")\n", + "print(\"- Calculate trade execution quality vs benchmarks\")" + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "## 4. Native Time-Series Operations\n", + "\n", + "KDB-X Python provides native functions for moving window calculations, such as Moving Averages (mavg) and Moving Standard Deviations (mdev). These are highly optimized for sequential data.\n", + "\n", + "In this block, we calculate common technical indicators like a 20-day Simple Moving Average (SMA), Volatility, and Rate of Change (ROC)." + ] + }, + { + "cell_type": "code", + "execution_count": 6, + "metadata": {}, + "outputs": [ + { + "name": "stdout", + "output_type": "stream", + "text": [ + "Computing time-series metrics...\n", + "Time-series calculations: 0.029 seconds\n", + "Daily bars created: 2,520\n", + "\n", + "Sample daily bars with technical indicators:\n" + ] + }, + { + "data": { + "text/html": [ + "\n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + "
dtsymclosesma_20vol_20roc_1vol_ratio
02024.08.31aapl284.1921280.25927.146515-0.0016573621.318701
12024.08.31adbe704.7346714.104416.947090.0048448410.9902567
22024.08.31amzn157.7019153.03815.531133-0.0077812941.319815
32024.08.31crm165.8433171.10135.632853-0.014868040.5939547
42024.08.31goog172.032180.18946.130507-0.012843730.9489909
52024.08.31meta332.4247316.62385.4150710.034497260.3879039
62024.08.31msft407.9326410.7026.3571710.0058805081.751181
72024.08.31nflx538.8412548.12426.49934-0.00064044390.3668744
82024.08.31nvda624.6716633.91839.350176-0.0052025290.5666062
92024.08.31tsla398.9801391.457713.118120.012253051.686152
" + ], + "text/plain": [ + "pykx.Table(pykx.q('\n", + "dt sym close sma_20 vol_20 roc_1 vol_ratio\n", + "------------------------------------------------------------------\n", + "2024.08.31 aapl 284.1921 280.2592 7.146515 -0.001657362 1.318701 \n", + "2024.08.31 adbe 704.7346 714.1044 16.94709 0.004844841 0.9902567\n", + "2024.08.31 amzn 157.7019 153.0381 5.531133 -0.007781294 1.319815 \n", + "2024.08.31 crm 165.8433 171.1013 5.632853 -0.01486804 0.5939547\n", + "2024.08.31 goog 172.032 180.1894 6.130507 -0.01284373 0.9489909\n", + "2024.08.31 meta 332.4247 316.6238 5.415071 0.03449726 0.3879039\n", + "2024.08.31 msft 407.9326 410.702 6.357171 0.005880508 1.751181 \n", + "2024.08.31 nflx 538.8412 548.1242 6.49934 -0.0006404439 0.3668744\n", + "2024.08.31 nvda 624.6716 633.9183 9.350176 -0.005202529 0.5666062\n", + "2024.08.31 tsla 398.9801 391.4577 13.11812 0.01225305 1.686152 \n", + "'))" + ] + }, + "execution_count": 6, + "metadata": {}, + "output_type": "execute_result" + } + ], + "source": [ + "# Time-series calculations using PyKX\n", + "print(\"Computing time-series metrics...\")\n", + "start_time = time.time()\n", + "\n", + "# Create daily OHLC bars for signal calculation\n", + "daily_bars = kx.q.qsql.select(\n", + " trades,\n", + " columns={\n", + " 'open': 'first px',\n", + " 'high': 'max px',\n", + " 'close': 'last px',\n", + " 'low': 'min px',\n", + " 'volume': 'sum volume'\n", + " },\n", + " by=['dt', 'sym']\n", + ")\n", + "\n", + "# Calculate technical indicators on daily data\n", + "enhanced_daily = kx.q.qsql.update(\n", + " daily_bars,\n", + " columns={\n", + " 'sma_10': 'mavg[10; close]',\n", + " 'sma_20': 'mavg[20; close]',\n", + " 'sma_50': 'mavg[50; close]',\n", + " 'vol_20': 'mdev[20; close]',\n", + " 'roc_1': '(close - prev close) % prev close',\n", + " 'vol_ratio': 'volume % mavg[20; volume]'\n", + " },\n", + " by=kx.Column('sym')\n", + ")\n", + "\n", + "timeseries_time = time.time() - start_time\n", + "print(f\"Time-series calculations: {timeseries_time:.3f} seconds\")\n", + "print(f\"Daily bars created: {len(enhanced_daily):,}\")\n", + "\n", + "print(\"\\nSample daily bars with technical indicators:\")\n", + "sample = kx.q.qsql.select(\n", + " kx.q('0!', enhanced_daily).tail(100),\n", + " columns=['dt', 'sym', 'close', 'sma_20', 'vol_20', 'roc_1', 'vol_ratio']\n", + ")\n", + "sample.head(10)" + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "## 5. Multi-Strategy Backtesting\n", + "\n", + "Backtesting involves defining logic to enter or exit trades. Because KDB-X operations are vectorized, we can generate signals for thousands of rows across multiple strategies simultaneously.\n", + "\n", + "We define four strategies:\n", + "\n", + "MA Crossover: Trend following.\n", + "\n", + "Mean Reversion: Overbought/Oversold detection.\n", + "\n", + "Momentum: Chasing strong daily returns.\n", + "\n", + "Volume Breakout: Identifying high-conviction moves." + ] + }, + { + "cell_type": "code", + "execution_count": 7, + "metadata": {}, + "outputs": [ + { + "name": "stdout", + "output_type": "stream", + "text": [ + "Implementing vectorized multi-strategy backtesting...\n", + "Signal generation: 0.00 seconds\n", + "\n", + "Strategy signal counts by symbol:\n" + ] + }, + { + "data": { + "text/html": [ + "\n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + "
ma_cross_signalsmean_rev_signalsmomentum_signalsvol_breakout_signals
sym
aapl169i38i77i24i
adbe142i47i67i21i
amzn115i59i57i12i
crm81i99i52i15i
goog155i50i78i23i
meta116i65i63i17i
msft139i59i62i16i
nflx132i65i64i18i
nvda142i51i70i20i
tsla181i28i70i28i
" + ], + "text/plain": [ + "pykx.KeyedTable(pykx.q('\n", + "sym | ma_cross_signals mean_rev_signals momentum_signals vol_breakout_signals\n", + "----| -----------------------------------------------------------------------\n", + "aapl| 169 38 77 24 \n", + "adbe| 142 47 67 21 \n", + "amzn| 115 59 57 12 \n", + "crm | 81 99 52 15 \n", + "goog| 155 50 78 23 \n", + "meta| 116 65 63 17 \n", + "msft| 139 59 62 16 \n", + "nflx| 132 65 64 18 \n", + "nvda| 142 51 70 20 \n", + "tsla| 181 28 70 28 \n", + "'))" + ] + }, + "execution_count": 7, + "metadata": {}, + "output_type": "execute_result" + } + ], + "source": [ + "# Define multiple trading strategies using PyKX\n", + "print(\"Implementing vectorized multi-strategy backtesting...\")\n", + "start_time = time.time()\n", + "\n", + "# Generate trading signals on daily data with better thresholds\n", + "signals = kx.q.qsql.update(\n", + " enhanced_daily,\n", + " columns={\n", + " # MA Crossover: buy when fast > slow\n", + " 'ma_cross_signal': 'sma_10 > sma_20',\n", + " # Mean reversion: buy when oversold (price drops below lower band)\n", + " 'mean_rev_signal': 'close < sma_20 - 1.0 * vol_20',\n", + " # Momentum: buy on strong positive daily return\n", + " 'momentum_signal': 'roc_1 > 0.01',\n", + " # Volume breakout: high volume and above trend\n", + " 'vol_breakout_signal': '(vol_ratio > 1.5) and (close > sma_20)'\n", + " },\n", + " by=kx.Column('sym')\n", + ")\n", + "\n", + "signal_time = time.time() - start_time\n", + "print(f\"Signal generation: {signal_time:.2f} seconds\")\n", + "\n", + "# Unkey for display\n", + "signals_unkeyed = kx.q('0!', signals)\n", + "\n", + "strategy_summary = kx.q.qsql.select(\n", + " signals_unkeyed,\n", + " columns={\n", + " 'ma_cross_signals': 'sum ma_cross_signal',\n", + " 'mean_rev_signals': 'sum mean_rev_signal',\n", + " 'momentum_signals': 'sum momentum_signal',\n", + " 'vol_breakout_signals': 'sum vol_breakout_signal'\n", + " },\n", + " by=kx.Column('sym')\n", + ")\n", + "\n", + "print(\"\\nStrategy signal counts by symbol:\")\n", + "strategy_summary.head(10)" + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "## 6. P&L Calculation and Risk Metrics\n", + "\n", + "To determine if a strategy is viable, we calculate the Profit and Loss (P&L). We use the next function in PyKX to determine forward returns (buying at today's close and looking at tomorrow's performance).\n", + "\n", + "We also bake in a transaction cost (10 basis points) to make the simulation realistic, as high-turnover strategies often fail once fees are applied." + ] + }, + { + "cell_type": "code", + "execution_count": 8, + "metadata": {}, + "outputs": [ + { + "name": "stdout", + "output_type": "stream", + "text": [ + "Calculating strategy P&L...\n", + "P&L calculation: 0.00 seconds\n", + "\n", + "Daily strategy P&L and trade counts (first 10 days):\n" + ] + }, + { + "data": { + "text/html": [ + "\n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + "
ma_cross_dailymean_rev_dailymomentum_dailyvol_breakout_dailyma_cross_tradesmean_rev_tradesmomentum_tradesvol_breakout_trades
dt
2024.01.020f0f0f0f0i0i0i0i
2024.01.030f9.5373963.8909760f0i1i1i0i
2024.01.040f-6.41645215.5096.0360360i3i5i1i
2024.01.050f-38.952160f0f0i2i0i0i
2024.01.060f-21.004285.574246-9.7978210i1i1i1i
2024.01.070f-16.855693.539849-2.8817040i1i1i1i
2024.01.080f-0.36124631.3888290f0i1i4i0i
2024.01.090f0f0f0f0i0i0i0i
2024.01.100f-7.2878188.1674370f0i1i3i0i
2024.01.110f3.3505140f0f0i1i0i0i
" + ], + "text/plain": [ + "pykx.KeyedTable(pykx.q('\n", + "dt | ma_cross_daily mean_rev_daily momentum_daily vol_breakout_daily m..\n", + "----------| -----------------------------------------------------------------..\n", + "2024.01.02| 0 0 0 0 0..\n", + "2024.01.03| 0 9.537396 3.890976 0 0..\n", + "2024.01.04| 0 -6.416452 15.509 6.036036 0..\n", + "2024.01.05| 0 -38.95216 0 0 0..\n", + "2024.01.06| 0 -21.00428 5.574246 -9.797821 0..\n", + "2024.01.07| 0 -16.85569 3.539849 -2.881704 0..\n", + "2024.01.08| 0 -0.3612463 1.388829 0 0..\n", + "2024.01.09| 0 0 0 0 0..\n", + "2024.01.10| 0 -7.287818 8.167437 0 0..\n", + "2024.01.11| 0 3.350514 0 0 0..\n", + "'))" + ] + }, + "execution_count": 8, + "metadata": {}, + "output_type": "execute_result" + } + ], + "source": [ + "# Calculate P&L for each strategy on daily data\n", + "print(\"Calculating strategy P&L...\")\n", + "start_time = time.time()\n", + "\n", + "# Calculate next day's return (forward return for holding overnight)\n", + "signals_unkeyed = kx.q('0!', signals)\n", + "\n", + "pnl_calc = kx.q.qsql.update(\n", + " signals_unkeyed,\n", + " columns={\n", + " 'next_close': 'next close',\n", + " 'forward_ret': '(next[close] - close) % close'\n", + " },\n", + " by=kx.Column('sym')\n", + ")\n", + "\n", + "# Detect signal entries (only trade when signal turns on)\n", + "pnl_calc = kx.q.qsql.update(\n", + " pnl_calc,\n", + " columns={\n", + " 'ma_cross_entry': 'ma_cross_signal and not prev ma_cross_signal',\n", + " 'mean_rev_entry': 'mean_rev_signal and not prev mean_rev_signal',\n", + " 'momentum_entry': 'momentum_signal and not prev momentum_signal',\n", + " 'vol_breakout_entry': 'vol_breakout_signal and not prev vol_breakout_signal'\n", + " },\n", + " by=kx.Column('sym')\n", + ")\n", + "\n", + "# Position sizing: invest $1000 per signal\n", + "position_size = 1000\n", + "transaction_cost = 0.001 # 10 bps round trip\n", + "\n", + "# Calculate P&L: only take positions on entry signals\n", + "strategy_pnl = kx.q.qsql.update(\n", + " pnl_calc,\n", + " columns={\n", + " 'ma_cross_pnl': f'{position_size} * ma_cross_entry * (forward_ret - {transaction_cost})',\n", + " 'mean_rev_pnl': f'{position_size} * mean_rev_entry * (forward_ret - {transaction_cost})',\n", + " 'momentum_pnl': f'{position_size} * momentum_entry * (forward_ret - {transaction_cost})',\n", + " 'vol_breakout_pnl': f'{position_size} * vol_breakout_entry * (forward_ret - {transaction_cost})'\n", + " }\n", + ")\n", + "\n", + "pnl_time = time.time() - start_time\n", + "print(f\"P&L calculation: {pnl_time:.2f} seconds\")\n", + "\n", + "# Daily P&L is already at daily level, just select relevant columns\n", + "daily_strategy_pnl = kx.q.qsql.select(\n", + " strategy_pnl,\n", + " columns={\n", + " 'ma_cross_daily': 'sum ma_cross_pnl',\n", + " 'mean_rev_daily': 'sum mean_rev_pnl',\n", + " 'momentum_daily': 'sum momentum_pnl',\n", + " 'vol_breakout_daily': 'sum vol_breakout_pnl',\n", + " 'ma_cross_trades': 'sum ma_cross_entry',\n", + " 'mean_rev_trades': 'sum mean_rev_entry',\n", + " 'momentum_trades': 'sum momentum_entry',\n", + " 'vol_breakout_trades': 'sum vol_breakout_entry'\n", + " },\n", + " by=kx.Column('dt')\n", + ")\n", + "\n", + "print(\"\\nDaily strategy P&L and trade counts (first 10 days):\")\n", + "daily_strategy_pnl.head(10)" + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "### Performance and Risk Metrics\n", + "\n", + "Raw P&L doesn't tell the whole story; we must analyze risk-adjusted returns. In this block, we calculate:\n", + "\n", + "Sharpe Ratio: Return per unit of risk.\n", + "\n", + "Max Drawdown: The largest \"peak-to-trough\" decline.\n", + "\n", + "Win Rate: Percentage of profitable days." + ] + }, + { + "cell_type": "code", + "execution_count": 9, + "metadata": {}, + "outputs": [ + { + "name": "stdout", + "output_type": "stream", + "text": [ + "\n", + "=== Strategy Performance Metrics ===\n", + " strategy total_pnl avg_daily_pnl daily_vol sharpe_ratio \\\n", + "0 MA Crossover 13.5025 0.0536 8.2202 0.1035 \n", + "1 Mean Reversion -209.2541 -0.8304 10.8354 -1.2165 \n", + "2 Momentum -667.7410 -2.6498 21.6257 -1.9451 \n", + "3 Volume Breakout -337.7433 -1.3403 12.0821 -1.7609 \n", + "\n", + " max_drawdown win_rate_pct \n", + "0 112.4805 12.3016 \n", + "1 218.7915 23.4127 \n", + "2 730.9637 36.9048 \n", + "3 380.5482 21.0317 \n", + "\n", + "=== Total Trades by Strategy ===\n", + "MA Crossover: 77 trades\n", + "Mean Reversion: 148 trades\n", + "Momentum: 506 trades\n", + "Volume Breakout: 164 trades\n" + ] + } + ], + "source": [ + "# Calculate risk metrics\n", + "def calc_strategy_metrics(pnl_series, strategy_name):\n", + " total_pnl = float(kx.q.sum(pnl_series))\n", + " avg_daily = float(kx.q.avg(pnl_series))\n", + " std_daily = float(kx.q.dev(pnl_series))\n", + " \n", + " sharpe = (avg_daily * 252) / (std_daily * np.sqrt(252)) if std_daily > 0 else 0\n", + " \n", + " cumulative = kx.q(\"sums\", pnl_series)\n", + " running_max = kx.q(\"maxs\", cumulative) \n", + " drawdown = running_max - cumulative\n", + " max_dd = float(kx.q.max(drawdown))\n", + " \n", + " wins = kx.q.sum(pnl_series > 0)\n", + " total_days = kx.q.count(pnl_series)\n", + " win_rate = float(wins) / float(total_days) * 100 if total_days > 0 else 0\n", + " \n", + " return {\n", + " 'strategy': strategy_name,\n", + " 'total_pnl': total_pnl,\n", + " 'avg_daily_pnl': avg_daily,\n", + " 'daily_vol': std_daily,\n", + " 'sharpe_ratio': sharpe,\n", + " 'max_drawdown': max_dd,\n", + " 'win_rate_pct': win_rate\n", + " }\n", + "\n", + "strategy_metrics = []\n", + "strategy_columns = ['ma_cross_daily', 'mean_rev_daily', 'momentum_daily', 'vol_breakout_daily']\n", + "strategy_names = ['MA Crossover', 'Mean Reversion', 'Momentum', 'Volume Breakout']\n", + "\n", + "# Unkey the table to access columns as vectors\n", + "daily_pnl_unkeyed = kx.q('0!', daily_strategy_pnl)\n", + "\n", + "for col, name in zip(strategy_columns, strategy_names):\n", + " metrics = calc_strategy_metrics(daily_pnl_unkeyed[col], name)\n", + " strategy_metrics.append(metrics)\n", + "\n", + "metrics_df = pd.DataFrame(strategy_metrics)\n", + "print(\"\\n=== Strategy Performance Metrics ===\")\n", + "print(metrics_df.round(4))\n", + "\n", + "# Show trade counts\n", + "print(\"\\n=== Total Trades by Strategy ===\")\n", + "trade_counts = {\n", + " 'MA Crossover': float(kx.q.sum(daily_pnl_unkeyed['ma_cross_trades'])),\n", + " 'Mean Reversion': float(kx.q.sum(daily_pnl_unkeyed['mean_rev_trades'])),\n", + " 'Momentum': float(kx.q.sum(daily_pnl_unkeyed['momentum_trades'])),\n", + " 'Volume Breakout': float(kx.q.sum(daily_pnl_unkeyed['vol_breakout_trades']))\n", + "}\n", + "for strat, count in trade_counts.items():\n", + " print(f\"{strat}: {int(count)} trades\")" + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "## 7. Visualization" + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "The final step is visualizing the results. We use Matplotlib to plot the cumulative P&L of all four strategies, a risk-return scatter plot, and a comparison of Sharpe Ratios.\n", + "\n", + "This helps identify which strategy provides the most consistent returns with the least amount of volatility." + ] + }, + { + "cell_type": "code", + "execution_count": 10, + "metadata": {}, + "outputs": [ + { + "data": { + "image/png": "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", + "text/plain": [ + "
" + ] + }, + "metadata": {}, + "output_type": "display_data" + }, + { + "name": "stdout", + "output_type": "stream", + "text": [ + "\n", + "=== Performance Summary ===\n", + "Total data processed: 756,392 trades\n", + "Processing time: 0.23 seconds\n", + "Processing rate: 3,302,972 trades/second\n", + "\n", + "Best strategy: MA Crossover (Sharpe = 0.103)\n" + ] + } + ], + "source": [ + "# Create visualizations\n", + "daily_pnl_df = daily_strategy_pnl.pd()\n", + "\n", + "# Reset index to get 'dt' as a column, then convert to datetime\n", + "daily_pnl_df = daily_pnl_df.reset_index()\n", + "daily_pnl_df['date'] = pd.to_datetime(daily_pnl_df['dt'])\n", + "daily_pnl_df.set_index('date', inplace=True)\n", + "\n", + "for col in strategy_columns:\n", + " daily_pnl_df[f'{col}_cumulative'] = daily_pnl_df[col].cumsum()\n", + "\n", + "fig, axes = plt.subplots(2, 2, figsize=(15, 12))\n", + "fig.suptitle('PyKX Rapid Backtesting Results', fontsize=16)\n", + "\n", + "# Cumulative P&L\n", + "ax1 = axes[0, 0]\n", + "for i, col in enumerate(strategy_columns):\n", + " ax1.plot(daily_pnl_df.index, daily_pnl_df[f'{col}_cumulative'], \n", + " label=strategy_names[i], linewidth=2)\n", + "ax1.set_title('Cumulative P&L by Strategy')\n", + "ax1.set_ylabel('Cumulative P&L ($)')\n", + "ax1.legend()\n", + "ax1.grid(True, alpha=0.3)\n", + "\n", + "# Risk-Return Scatter\n", + "ax2 = axes[0, 1]\n", + "returns = [m['avg_daily_pnl'] * 252 for m in strategy_metrics]\n", + "risks = [m['daily_vol'] * np.sqrt(252) for m in strategy_metrics]\n", + "colors = ['blue', 'red', 'green', 'orange']\n", + "\n", + "scatter = ax2.scatter(risks, returns, c=colors, s=100, alpha=0.7)\n", + "for i, name in enumerate(strategy_names):\n", + " ax2.annotate(name, (risks[i], returns[i]), xytext=(5, 5), \n", + " textcoords='offset points', fontsize=8)\n", + "ax2.set_xlabel('Annual Volatility')\n", + "ax2.set_ylabel('Annual Return')\n", + "ax2.set_title('Risk-Return Profile')\n", + "ax2.grid(True, alpha=0.3)\n", + "\n", + "# Performance bars\n", + "ax3 = axes[1, 0]\n", + "total_pnls = [m['total_pnl'] for m in strategy_metrics]\n", + "ax3.bar(strategy_names, total_pnls, color=colors, alpha=0.8)\n", + "ax3.set_xlabel('Strategy')\n", + "ax3.set_ylabel('Total P&L ($)')\n", + "ax3.set_title('Total P&L by Strategy')\n", + "ax3.tick_params(axis='x', rotation=45)\n", + "ax3.grid(True, alpha=0.3)\n", + "\n", + "# Sharpe ratios\n", + "ax4 = axes[1, 1]\n", + "sharpe_ratios = [m['sharpe_ratio'] for m in strategy_metrics]\n", + "ax4.bar(strategy_names, sharpe_ratios, color=colors, alpha=0.8)\n", + "ax4.set_xlabel('Strategy')\n", + "ax4.set_ylabel('Sharpe Ratio')\n", + "ax4.set_title('Sharpe Ratio by Strategy')\n", + "ax4.tick_params(axis='x', rotation=45)\n", + "ax4.grid(True, alpha=0.3)\n", + "\n", + "plt.tight_layout()\n", + "plt.show()\n", + "\n", + "print(\"\\n=== Performance Summary ===\")\n", + "print(f\"Total data processed: {len(trades):,} trades\")\n", + "print(f\"Processing time: {generation_time + timeseries_time + signal_time + pnl_time:.2f} seconds\")\n", + "print(f\"Processing rate: {len(trades)/(generation_time + timeseries_time + signal_time + pnl_time):,.0f} trades/second\")\n", + "\n", + "best_strategy = metrics_df.loc[metrics_df['sharpe_ratio'].idxmax()]\n", + "print(f\"\\nBest strategy: {best_strategy['strategy']} (Sharpe = {best_strategy['sharpe_ratio']:.3f})\")" + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "### Wrapping Up\n", + "\n", + "This notebook demonstrates the significant performance advantages of using KDB-X for high-frequency financial analytics and backtesting. By leveraging the columnar storage and vectorized execution engine of KDB-X, we achieved processing rates of approximately 1.47 million trades per second.\n", + "\n", + "**Key Takeaways**\n", + "\n", + "**Massive Speed Gains**: PyKX outperformed standard Pandas operations significantly, executing daily OHLCV aggregations 3.1x faster and performing complex \"as-of joins\" 1.9x faster.\n", + "\n", + "**Memory Efficiency**: We generated and analyzed over 750,000 trades with a memory footprint of only 32.4 MB, showcasing KDB-X's ability to handle large-scale datasets efficiently.\n", + "\n", + "**Native Time-Series Tooling**: The use of native q functions like mavg (moving average) and aj (as-of join) allows for the calculation of technical indicators and trade enrichment without leaving the optimized KDB-X environment." + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [] + } + ], + "metadata": { + "kernelspec": { + "display_name": ".venv", + "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": "3.10.12" + } + }, + "nbformat": 4, + "nbformat_minor": 4 +} diff --git a/KDB-X/Python/README.md b/KDB-X/Python/README.md index 9d969e2..8d8915b 100644 --- a/KDB-X/Python/README.md +++ b/KDB-X/Python/README.md @@ -9,7 +9,7 @@ By following these tutorials, you'll: ## 📖 Tutorials -### 1️⃣ Time Series Historical Analysis +### Time Series Historical Analysis - Jupyter Notebook version: [Time_Series_Historical_Analysis.ipynb](1.Time_Series_Historical_Analysis.ipynb) - Markdown version: [Time_Series_Historical_Analysis.md](1.Time_Series_Historical_Analysis.md) #### Key topics covered: @@ -17,5 +17,10 @@ By following these tutorials, you'll: - Efficient querying and filtering - Using time-series functions and joins via Python +### KDB-X Python Backtesting +- Demonstrates KDB-X speed advantages +- Compares performance with pandas +- Enriches data using as-of joins + ## 🤝 Got a question? Want to connect with other developers or get help? Join our Slack community https://kx.com/slack or ask a question on https://forum.kx.com diff --git a/KDB-X/q/README.md b/KDB-X/q/README.md index a1d62aa..8e03ef1 100644 --- a/KDB-X/q/README.md +++ b/KDB-X/q/README.md @@ -8,15 +8,20 @@ By following these tutorials, you'll: - Be inspired to apply KDB-X to your own data and problem sets. ## 📖 Tutorials -### 1️⃣ Time_Series_Historical_Analysis.md +### Time_Series_Historical_Analysis.md - Storing & handling large-scale time-series data - Efficient querying and filtering - Using time-series functions and joins in q -### 2️⃣ Live_Data_Event_Processing.md +### Live_Data_Event_Processing.md - Real-time data ingestion - Simulate live data streaming - Process real-time event streams and detect patterns +### Financial Backtesting with KDB-X/q +- Build professional backtesting systems with q +- Calculate PnL and analyze trading strategies +- Learn why institutions use kdb+ for finance + ## 🤝 Got a question? Want to connect with other developers or get help? Join our Slack community https://kx.com/slack or ask a question on https://forum.kx.com diff --git a/KDB-X/q/q_backtest.ipynb b/KDB-X/q/q_backtest.ipynb new file mode 100644 index 0000000..77b1433 --- /dev/null +++ b/KDB-X/q/q_backtest.ipynb @@ -0,0 +1,501 @@ +{ + "cells": [ + { + "cell_type": "markdown", + "id": "intro-cell", + "metadata": {}, + "source": [ + "# Financial Backtesting with q / KDB-X\n", + "\n", + "## What is this notebook about?\n", + "\n", + "This notebook demonstrates how to build a **financial backtesting system** using KDB-X/q. We'll simulate a simple trading strategy and calculate how much profit or loss it would have generated.\n", + "\n", + "## Why use KDB-X/q for backtesting?\n", + "\n", + "KDB-X/q is a high-performance database and programming language specifically designed for time-series data. It's used by major financial institutions because:\n", + "\n", + "- **Speed**: Processes millions of trades in milliseconds\n", + "- **Time-series operations**: Built-in functions like `asof joins` make matching trades to market prices trivial\n", + "- **Columnar storage**: Extremely efficient for financial data analysis\n", + "- **Concise syntax**: Complex operations can be expressed in just a few lines\n", + "\n", + "## Prerequisites\n", + "- Requires KDB-X to be installed, you can sign up and download on [Developer Center](https://developer.kx.com/products/kdb-x/install). For full install instructions see: [KDB-X Install](https://code.kx.com/kdb-x/).\n", + "\n", + "- To Install KDB-X Python: `pip install --upgrade --pre pykx` (This is required to run q in a Python notebook, but you can run the q directly from a q session if you prefer.)\n", + "\n", + "### What is PyKX?\n", + "\n", + "PyKX is a Python library that lets us run q/KDB-X code directly from Python. In this notebook, we use the `%%q` magic command to write q code in Jupyter cells, combining the interactivity of notebooks with the power of q. You can run these commands directly in a q session if preferred. " + ] + }, + { + "cell_type": "code", + "execution_count": 1, + "id": "5b052c32-0a48-42e0-988c-1db403deac61", + "metadata": {}, + "outputs": [], + "source": [ + "import pykx as kx" + ] + }, + { + "cell_type": "markdown", + "id": "market-data-intro", + "metadata": {}, + "source": [ + "## Step 1: Generate Simulated Market Data\n", + "\n", + "Before we can backtest a trading strategy, we need market data. In the real world, this would come from historical databases, but for this example we'll generate realistic-looking random data.\n", + "\n", + "### What we're creating:\n", + "- **Trades** for three stocks: AAPL (Apple), GOOG (Google), IBM\n", + "- **Date range**: January 1-31, 2015\n", + "- **Fields**: date, time, symbol, quantity, price, market cap\n", + "\n", + "### Why this matters:\n", + "In real backtesting, you need to know the **exact price** at which you could have executed trades. The `aj` (asof join) function in q makes this easy - it finds the most recent price as of a given timestamp, which is exactly what happens in real trading.\n", + "\n", + "### The Power of q:\n", + "Notice how we can:\n", + "1. Generate 10,000 random trades in one line\n", + "2. Sort by multiple columns (`dt` and `tm`) with a simple expression\n", + "3. Update prices conditionally by symbol with clean syntax\n", + "4. Calculate VWAP (volume-weighted average price) in a single operation\n", + "\n", + "This type of data manipulation would require many more lines in traditional languages." + ] + }, + { + "cell_type": "code", + "execution_count": 2, + "id": "c4b85c4b-0910-49fc-bf9a-ff05ce91846b", + "metadata": {}, + "outputs": [ + { + "name": "stdout", + "output_type": "stream", + "text": [ + "pnlseries| +`dt`pnl!(`s#2015.01.01 2015.01.02 2015.01.03 2015.01.04 2015.01.0..\n", + "summary | 1495.285\n", + "pctreturn| 0.002510236\n" + ] + } + ], + "source": [ + "%%q\n", + "\n", + "/function outputting sample market data\n", + "mktrades:{[tickers; sz]\n", + " dt:2015.01.01+sz?31; / Generate random dates in January 2015\n", + " tm:sz?24:00:00.000; / Generate random times throughout the day\n", + " sym:sz?tickers; / Randomly assign stock symbols\n", + " qty:10*1+sz?1000; / Random quantities (10 to 10,000 shares)\n", + " px:90.0+(sz?2001)%100; / Random prices around $90-110\n", + " mc: 10000; / Market cap (simplified, constant)\n", + " t:([] dt; tm; sym; qty; px; mc); / Create table with these columns\n", + " t:`dt`tm xasc t; / Sort by date then time (chronological order)\n", + " t:update px:6*px from t where sym=`goog; / GOOG trades at ~6x the base price\n", + " t:update px:2*px from t where sym=`ibm; / IBM trades at ~2x the base price\n", + " t};\n", + "\n", + "/running function to generate sample market data \n", + "trades:mktrades[`aapl`goog`ibm; 10000];\n", + "trades: update vwapx: qty wavg px by dt,sym from trades; / Calculate VWAP per day per symbol\n", + "\n", + "/sample strategy - buy goog, appl, ibm on 1/1/15, sell on 1/31/15\n", + "example:select distinct sym,size:1000,tradein:first dt,tradeout:last dt from trades;\n", + "\n", + "/create backtest table\n", + "bt_table:\n", + " select\n", + " sym,\n", + " size,\n", + " tradein,\n", + " tradeout\n", + " from\n", + " example\n", + " where\n", + " sym in `ibm`goog;\n", + "\n", + "\n", + "/----\n", + "/example 1: basic backtest logic\n", + "/trades = tick data\n", + "/bt_table = backtest table\n", + "t1:aj[`sym`dt;select sym,size,dt:tradein from bt_table;trades]; /takes last trade for IBM on 1/1/15 as entry price\n", + "t2:aj[`sym`dt;select sym,size,dt:tradeout from bt_table;trades]; /takes last trade for IBM on 1/31/15 as exit price\n", + "(t2[`px]-t1[`px])*t1[`size]; /pnl\n", + "\n", + "/----\n", + "/example 2: backtest function (simplest form, no timeseries nuance)\n", + "bt:{[table] /function takes in one parameter, the backtest table\n", + " /find entry price for each sym via AJ\n", + " tentry:aj[`sym`dt;select sym,size,dt:tradein from table;trades];\n", + " /find exit price for each sym via AJ\n", + " texit:aj[`sym`dt;select sym,size,dt:tradeout from table;trades];\n", + " /calculate pnl for each sym over trade lifetime\n", + " pnl_by_sym: (texit[`px]-tentry[`px])*tentry[`size];\n", + " pnl: sum pnl_by_sym;\n", + " /return pnl\n", + " pnl\n", + " }\n", + " \n", + "/run example backtest function on input table to derive pnl\n", + "res:bt[bt_table]\n", + "\n", + "/----\n", + "/example 3: backtest function highlighting PnL at each trading day using vwap)\n", + "btn:{[table]\n", + " /generate dates between entry and exit\n", + " daterack:flip exec dt:{y+til x--1+y}[max tradeout;min tradein] from table;\n", + " daterack:daterack cross distinct select sym,size from bt_table;\n", + "\n", + " /calc vwap by date from trades\n", + " tvwap: select vwap:qty wavg px by dt,sym from trades;\n", + " tvwap: () xkey tvwap;\n", + " /use aj to find sym's price as of each date from the trades marketdata table (here, vwap proxy for each date)\n", + " /daterack- dt,sym,size\n", + " /trades- `dt`tm`sym`qty`px`mc\n", + " daterack:aj[`dt`sym;daterack;select from tvwap];\n", + " /pnl calc\n", + " daterack:update pnl: size*vwap-first[vwap] by sym from daterack;\n", + " `pnlseries`summary`pctreturn!(\n", + " ([] dt: asc distinct exec dt from daterack;pnl: exec pnl from select sum pnl by dt from daterack);\n", + " last daterack[`pnl];\n", + " (last daterack[`pnl])%(bt_table[0;`size]*first daterack[`vwap])\n", + " )\n", + " };\n", + " \n", + "/run backtest\n", + "btn[bt_table]" + ] + }, + { + "cell_type": "markdown", + "id": "strategy-intro", + "metadata": {}, + "source": [ + "## Step 2: Define Our Trading Strategy\n", + "\n", + "Now we need to define **what** we want to backtest. Our strategy is simple:\n", + "\n", + "### The Strategy:\n", + "- **Buy** 1,000 shares each of GOOG and IBM on January 1, 2015\n", + "- **Hold** them for the entire month\n", + "- **Sell** everything on January 31, 2015\n", + "\n", + "### The Backtest Table:\n", + "The `bt_table` contains our planned trades:\n", + "- `sym`: which stock to trade\n", + "- `size`: how many shares (1,000 each)\n", + "- `tradein`: when to buy (entry date)\n", + "- `tradeout`: when to sell (exit date)\n", + "\n", + "This is a \"buy and hold\" strategy - we're testing whether simply holding these two tech stocks for a month would have been profitable." + ] + }, + { + "cell_type": "markdown", + "id": "example1-intro", + "metadata": {}, + "source": [ + "## Example 1: Basic Backtest Logic\n", + "\n", + "This is the simplest way to calculate profit/loss. We need to answer two questions:\n", + "1. **What price did we buy at?** (entry price on Jan 1)\n", + "2. **What price did we sell at?** (exit price on Jan 31)\n", + "\n", + "### The Magic of Asof Join (`aj`):\n", + "\n", + "The `aj` function is one of q's superpowers. It finds the **most recent** matching record as of a specific time. This is exactly what you need for financial data:\n", + "\n", + "```q\n", + "aj[`sym`dt; our_trades; market_data]\n", + "```\n", + "\n", + "This says: \"For each of our trades, find the last market price available on that date for that symbol.\"\n", + "\n", + "### Why this is powerful:\n", + "- In traditional SQL, this would require complex subqueries or window functions\n", + "- In pandas, you'd need to merge, sort, and use groupby operations\n", + "- In q, it's **one function call** and it's blazingly fast\n", + "\n", + "### The Calculation:\n", + "```\n", + "PnL = (Exit Price - Entry Price) × Number of Shares\n", + "```\n", + "\n", + "Let's see it in action:" + ] + }, + { + "cell_type": "code", + "execution_count": 3, + "id": "227b3c08-9d75-41ab-ab48-436909f99160", + "metadata": {}, + "outputs": [ + { + "name": "stdout", + "output_type": "stream", + "text": [ + "16600f\n" + ] + } + ], + "source": [ + "%%q\n", + "bt[bt_table]" + ] + }, + { + "cell_type": "markdown", + "id": "bt-function-result", + "metadata": {}, + "source": [ + "### Result: $16,600 profit\n", + "\n", + "Our simple strategy would have made $16,600 over the month. But this is just a single number - we don't know:\n", + "- How the PnL changed day by day\n", + "- Whether we were ever losing money during the month\n", + "- What the percentage return was\n", + "\n", + "Let's dig deeper with the next example..." + ] + }, + { + "cell_type": "markdown", + "id": "example3-intro", + "metadata": {}, + "source": [ + "## Example 2: Advanced Backtest with Daily PnL\n", + "\n", + "The `btn` function (backtest with nuance) gives us much more insight. Instead of just final PnL, we get:\n", + "\n", + "### What this calculates:\n", + "1. **Daily PnL**: How much profit/loss we had at the end of each day\n", + "2. **PnL Series**: A complete timeseries showing portfolio value over time\n", + "3. **Percent Return**: What percentage of our initial investment we made\n", + "\n", + "### How it works:\n", + "1. **Generate all dates** between entry and exit\n", + "2. For each date, calculate the VWAP (volume-weighted average price)\n", + "3. Use `aj` to find each stock's VWAP on each day\n", + "4. Calculate daily PnL relative to our entry price\n", + "\n", + "### Why VWAP?\n", + "VWAP (Volume-Weighted Average Price) is a more realistic price than just using the last trade of the day. It represents the average price weighted by trading volume - essentially the \"fair\" price for that day.\n", + "\n", + "### The q Advantage:\n", + "Look at this line:\n", + "```q\n", + "tvwap: select vwap:qty wavg px by dt,sym from trades\n", + "```\n", + "\n", + "This calculates volume-weighted average prices grouped by date and symbol in **one line**. In other languages, this would require:\n", + "- Grouping data\n", + "- Calculating weighted averages manually\n", + "- Aggregating results\n", + "- Multiple lines of code\n", + "\n", + "Let's run it:" + ] + }, + { + "cell_type": "code", + "execution_count": 4, + "id": "d304651c-cafd-431b-909e-24ca74894859", + "metadata": {}, + "outputs": [ + { + "name": "stdout", + "output_type": "stream", + "text": [ + "pnlseries| +`dt`pnl!(`s#2015.01.01 2015.01.02 2015.01.03 2015.01.04 2015.01.05 2015.01.06 2015.01.07 2015.01.08 2015.01.09 2015.01.10 2015.01.11 2015.01.12 2015.01.13 2015.01.14 2015.01.15 2015.01.16 2015.01.17 2015.01.18 2015.01.19 2015.01.20 2015.01.21 2015.01.22 2015.01.23 2015.01.24 2015.01.25 2015.01.26 2015.01.27 2015.01.28 2015.01.29 2015.01.30 2015.01.31;0 4091.652 4464.118 10419.29 1920.787 7759.533 -2627.317 8466.34 1303.098 13082 9628.353 4784.923 436.9293 4655.67 8299.684 -3836.032 5400.844 9045.004 7603.171 1681.742 2462.197 1968.41 4752.998 -474.5213 8861.784 460.883 1926.882 7370.895 5069.713 5854.743 5480.873)\n", + "summary | 1495.285\n", + "pctreturn| 0.002510236\n" + ] + } + ], + "source": [ + "%%q\n", + "\\c 1000 1000\n", + "btn[bt_table]" + ] + }, + { + "cell_type": "markdown", + "id": "results-analysis", + "metadata": {}, + "source": [ + "## Understanding the Results\n", + "\n", + "### Key Findings:\n", + "\n", + "1. **Summary PnL: $1,495.28**\n", + " - This is our final profit using VWAP prices (more realistic than the simple $16,600 using last trades)\n", + " - The difference shows why using VWAP matters for accurate backtesting\n", + "\n", + "2. **Percent Return: 0.25%**\n", + " - Our investment returned 0.25% over the month\n", + " - This is modest but positive\n", + "\n", + "3. **Daily PnL Series:**\n", + " - We can see the portfolio value changed every day\n", + " - Notice **January 7th and 16th** showed losses (negative PnL)\n", + " - **January 10th** was our best day with ~$13,000 gain\n", + " - The portfolio was volatile but generally trended upward\n", + "\n", + "### What This Tells Us:\n", + "- The strategy wasn't consistently profitable day-to-day\n", + "- We had drawdown periods (times when we were losing money)\n", + "- Risk management would have been important\n", + "- A trader might have panicked and sold on Jan 7th or 16th when showing losses\n", + "\n", + "### The Power of q for This Analysis:\n", + "We generated:\n", + "- 10,000 simulated trades\n", + "- Daily VWAP calculations\n", + "- 31 days of portfolio valuations\n", + "- Complete risk metrics\n", + "\n", + "All in **just a few lines of q code** and executed in **milliseconds**. This same analysis in traditional tools would require significantly more code and computation time." + ] + }, + { + "cell_type": "markdown", + "id": "detailed-look", + "metadata": {}, + "source": [ + "## A Detailed Look at Entry and Exit Prices\n", + "\n", + "Let's examine exactly what prices we entered and exited at for each stock:" + ] + }, + { + "cell_type": "code", + "execution_count": 5, + "id": "efd46461-3fce-4ff1-bcd5-2eed81abfa9b", + "metadata": {}, + "outputs": [ + { + "name": "stdout", + "output_type": "stream", + "text": [ + "sym size dt tm qty px mc vwapx \n", + "------------------------------------------------------------\n", + "goog 1000 2015.01.01 23:15:36.077 8520 590.7 10000 595.6753\n", + "ibm 1000 2015.01.01 23:35:57.076 6530 207.16 10000 200.2504\n", + "::\n", + "::\n", + "sym size dt tm qty px mc vwapx \n", + "------------------------------------------------------------\n", + "goog 1000 2015.01.31 23:53:43.752 7780 632.76 10000 599.6609\n", + "ibm 1000 2015.01.31 23:59:07.179 1040 181.7 10000 201.7456\n", + "::\n", + "::\n", + "16600f\n" + ] + } + ], + "source": [ + "%%q\n", + "\n", + "\n", + "t1\n", + "\n", + "(::;::)\n", + "\n", + "\n", + "t2\n", + "\n", + "\n", + "(::;::)\n", + "\n", + "sum (t2[`px]-t1[`px])*t1[`size]\n", + "\n" + ] + }, + { + "cell_type": "markdown", + "id": "price-analysis", + "metadata": {}, + "source": [ + "## Breaking Down the Trades\n", + "\n", + "### Entry Positions (January 1st):\n", + "- **GOOG**: Bought 1,000 shares at $590.70\n", + " - Last trade of the day at 11:15 PM\n", + " - Total investment: $590,700\n", + " \n", + "- **IBM**: Bought 1,000 shares at $207.16\n", + " - Last trade of the day at 11:35 PM \n", + " - Total investment: $207,160\n", + "\n", + "**Total Initial Investment: $797,860**\n", + "\n", + "### Exit Positions (January 31st):\n", + "- **GOOG**: Sold 1,000 shares at $632.76\n", + " - Last trade of the month at 11:53 PM\n", + " - Total proceeds: $632,760\n", + " - **Profit on GOOG: $42,060** ✓\n", + " \n", + "- **IBM**: Sold 1,000 shares at $181.70\n", + " - Last trade of the month at 11:59 PM\n", + " - Total proceeds: $181,700 \n", + " - **Loss on IBM: -$25,460** ✗\n", + "\n", + "### The Final Math:\n", + "```\n", + "GOOG: ($632.76 - $590.70) × 1,000 = +$42,060\n", + "IBM: ($181.70 - $207.16) × 1,000 = -$25,460\n", + " Total = +$16,600\n", + "```\n", + "\n", + "### Key Insight:\n", + "While we made money overall, **IBM actually lost money** over the month. GOOG's strong performance (+7.1%) more than compensated for IBM's decline (-12.3%). This is why diversification matters - and why seeing individual stock performance is important in backtesting!\n", + "\n", + "---\n", + "\n", + "## Conclusion: Why KDB-X/q for Backtesting?\n", + "\n", + "This notebook demonstrated several key advantages:\n", + "\n", + "1. **Time-series operations are natural**: The `aj` function makes it trivial to match trades to prices\n", + "2. **Concise and readable**: Complex calculations in just a few lines\n", + "3. **Fast execution**: Processing 10,000 trades with daily aggregations in milliseconds\n", + "4. **Built for finance**: Features like VWAP, grouping by symbol/date, and temporal joins are first-class operations\n", + "\n", + "For professional backtesting with millions of trades, real-time data feeds, or complex multi-asset strategies, KDB-X/q's performance advantages become even more pronounced. What we did here in a simple notebook would scale to institutional-grade backtesting infrastructure with minimal code changes." + ] + }, + { + "cell_type": "markdown", + "id": "18ec39b2", + "metadata": {}, + "source": [] + } + ], + "metadata": { + "kernelspec": { + "display_name": "pykx3_demo", + "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": "3.10.15" + } + }, + "nbformat": 4, + "nbformat_minor": 5 +}