Formulaic Alpha Trade Demo

Hi! I’m still on travel - and won’t return to my desk till next week, but I just wanted to give a short demo on how you can use the different components of quantpylib to run, say, a formulaic alpha on live.

As promised, it will be incredibly easy. We make some imports, set up our keys:

import os 
import pytz
import asyncio

from decimal import Decimal
from dotenv import load_dotenv
load_dotenv()
from datetime import datetime

import quantpylib.standards.markets as markets 

from quantpylib.standards import Period
from quantpylib.gateway.master import Gateway
from quantpylib.datapoller.master import DataPoller
from quantpylib.simulator.gene import GeneticAlpha

exc = "hyperliquid"
config_keys = {
    exc: {
        "alias" : exc,
        "address" : os.getenv("HYP_DEMO"),
        "hyp_key" : os.getenv("HYP_KEY"),
    }
}
datapoller = DataPoller(config_keys=config_keys)
gateway = Gateway(config_keys=config_keys)
interval = Period.DAILY

if __name__ == "__main__":
    asyncio.run(main())

Now let’s write the main method - we retrieve our capital, get the contracts available. and get OHLCV:

async def main():
    capital = (await gateway.account.account_balance(exc=exc))[markets.ACCOUNT_EQUITY]
    contracts = await gateway.exchange.contract_specifications(exc=exc)
    tickers = list(contracts.keys())[:20]

    period_start = datetime(2016,1,1, tzinfo=pytz.utc)
    period_end = datetime.now(pytz.utc)

    ticker_dfs = [datapoller.crypto.get_trade_bars(
        ticker=ticker,
        start=period_start,
        end=period_end,
        granularity=interval,
        kline_close=True,
        src=exc
    ) for ticker in tickers]
    dfs = {ticker:df for ticker,df in zip(tickers,ticker_dfs)} 

As usual, we set up our genetic alpha, write the fee parameters, inertia, volatility target and so on…then we run the backtest on say, the bollinger score momentum:

    strat_configs = {
        "dfs":dfs,
        "granularity":interval,
        "instruments":tickers,
        "execrates": [0.0005] * len(tickers),
        "longswps": [0.1] * len(tickers),
        "shortswps": [-0.1] * len(tickers),
        "weekend_trading":True,
        "positional_inertia": 0.1
    }
        
    alpha = GeneticAlpha(genome="tszscre_12(close)",portfolio_vol=0.25,**strat_configs)
    df = await alpha.run_simulation()

Last but not least we obtain held positions from the gateway, and compute the contracts difference between target and held - then we use the gateway to submit orders:

    held = await gateway.positions.positions_get(exc=exc)
    take = alpha.get_positions_for_capital(capital=capital,held={k:v[markets.AMOUNT] for k,v in held.items()})
    trade = {inst : Decimal(str(take[inst])) - (Decimal('0') if inst not in held else held[inst][markets.AMOUNT]) for inst in  tickers}
    trade = {inst : round(amount,int(contracts[inst][markets.SYMBOL_QUANTITY_PRECISION])) for inst,amount in trade.items()}
    
    for inst,amount in trade.items():
        res = await gateway.executor.market_order(exc=exc,ticker=inst,amount=amount)

And just by changing the formulaic string - you can just about run about any trading strategy encoded formulaically….

Say you want to run this every day, on a new candle, we can schedule it on cron:

TZ=UTC
1 0 * * * python3 /path/alpha.py

This runs it at 00:00:01 UTC every day. Get access to quantpylib here:

HangukQuant Community Github Repo

Will be back next week to superboost quantpylib again…so get ready!