Part 16 min read
I built a futures trading bot. The honest first month.
Part 1 of building a Binance futures bot in public. The strategy, the bugs that cost real money, and a first month that started down 13%. No fake P&L.
Read part 1Build in public · Real capital
The journey
I'm building and running a real Binance futures trading bot with my own money. This is the honest log: the strategy, the bugs that cost me money, and the live results, good and bad. No hype, no fake P&L.
- Part 26 min read
Survivorship bias killed my 1H bot (part 2)
Part 2 of the build-in-public bot series. A holdout test on 11 unseen coins revealed the 1H strategy barely beat a coin flip. The 4H version survived.
Read part 2 - Part 38 min read
The backtest that lied: same-bar re-entry artifact (part 3)
A backtest artifact made one coin look like +5,032% in a year. Fixing the same-bar re-entry bug dropped returns 30-50x and turned 5-7 winning strategies into zero.
Read part 3 - Part 48 min read
Three bugs that a backtest can't catch (part 4)
Three live bugs a backtest never shows: shared-margin sizing, Binance -4120 algo order migration, and stale-candle stop rejection -2021. Real fixes, real cost.
Read part 4 - Part 58 min read
ADX regime filtering: when the best trade is no trade (part 5)
How an ADX regime filter cut chop losses from a Binance futures trend bot. Trade above 25, sit flat below 20, hysteresis in between. Why doing nothing beats both alternatives.
Read part 5 - Part 68 min read
Risk sizing is a linear dial, no free lunch (part 6)
Why 2x risk per trade means 2x return and 2x drawdown, always. How I size a live Binance futures bot for the bad case, not the good one, and why 25-35% drawdowns are normal.
Read part 6