Alpaca-compliant algorithm for tracking a single asset. Underperforms most assets except in downturns - a conservative strategy.

track-ticker.py

from pylivetrader.api import *
import logbook


log = logbook.Logger('track-ticker')


def enter_play(context, data):
    s = context.ticker

    if not data.can_trade(s):
        return

    fast_sma = data.history(s, 'price', context.fast_sma_days, '1d').mean()
    slow_sma = data.history(s, 'price', context.slow_sma_days, '1d').mean()
    
    context.target_percentage = context.exit_percentage if fast_sma < slow_sma else context.enter_percentage

    print("Current target percentage: {.2f}".format(context.target_percentage))

    order_target_percent(s, context.target_percentage)


def initialize(context):
    # TODO: Try gradient descent on these parameters
    context.exit_percentage = 0.1
    context.enter_percentage = 1.0
    context.fast_sma_days = 2
    context.slow_sma_days = 15
    
    context.trade_at_minute = 30
    
    context.ticker = symbol('MSFT')

    if context.target_percentage:
        log.info("Current target percentage: {}".format(round(context.target_percentage * 100, 2)))
    else:
        log.info("Fresh context")

    schedule_function(
        enter_play,
        date_rules.every_day(),
        time_rules.market_open(minutes=context.trade_at_minute),
    )