Why your crypto trading bot isn’t profitable — Part 2
In Part 1 we examined a common type of trading strategy and ran backtests on it to show how it might [hypothetically] produce profits. In this next section we will go deeper and look at precisely WHY these strategies come up short in live trading.
The 4 horsemen of profit erosion
Why these strategies come up short in real live trading.
(in no particular order though fees are least important to understand)
- exchange spread
- gap-effects
- trailing stop-loss opaqueness
- exchange fees
Not only will these erode profits in live trading but they will cause your backtesting results to diverge from live trading, increasingly so as your model frequency and complexity increases.
The Spread
By definition, you will not have bought at the price in historical data — you will have bought (in a market order) at the ASK (a higher price). That’s the definition of the market spread, right?
