Learn · Topic
EAs and automated trading
How EAs work, how to backtest them and how to set them up for your account.
10 articles
Test Your Entry Against Random Entries: Does Your Signal Beat a Coin Flip?
Backtest profit can come from your exits and the market's trend, not your entry. How to compare with 1,000 random-entry runs, get a p-value, and build a fair placebo.
12 min readHow to Judge a Copy-trading, Signal or EA Track Record Before You Follow
How to check a signal seller, copy-trading provider or EA seller - verified accounts, account age, equity drawdown, martingale signs, luck, deposits and copier slippage.
9 min readMonte Carlo Strategy Testing: Reshuffle Trades to See the Drawdown to Plan For
A backtest gives one path. Monte Carlo reshuffles and resamples the trade list thousands of times to show the range of drawdown, streaks, net profit and risk of ruin.
9 min readWalk-forward Analysis: Test an EA on Rolling Windows Without Fooling Yourself
Fit settings on one stretch, test them on the next stretch you didn't use, then roll forward. Rolling vs anchored, walk-forward efficiency, and a worked example.
10 min readEquity Curve Trading: Does Switching Off Below the Equity MA Really Help?
Switch a system off when equity drops below its own moving average, on again when it recovers. Simulations show when it helps, when it hurts, and the test it must pass.
10 min readData Snooping: Test 200 Strategies and the Best One May Be Pure Luck
The more variants you test, the more fake winners you find. Best-of-N Sharpe inflation, Bonferroni, Reality Check, t > 3, Deflated Sharpe and a 200-strategy simulation.
9 min readPairs Trading and Cointegration: Why Correlation Is Not Enough
Trading the spread between two assets: hedge ratio by regression, z-score, half-life and what happens when the relationship breaks, with costs on both legs.
12 min readRepainting Indicators and Lookahead Bias: Why Great Backtests Fail Live
ZigZag redraws its last leg until it confirms and fractals need future bars. How future data leaks into backtests and EAs, with a worked example of the fake profit.
9 min readWhat Is Overfitting? Pick EA Settings on a Plateau, Not a Peak
Tuning an EA until the backtest looks great may be pure luck. Pick values whose neighbours also work, split search and confirm years, and beat a plain lot increase.
7 min readWhy live results usually fall short of the Backtest, and how much to allow for
A great Backtest doesn't mean live trading will match it. The main causes: tick data quality, spread spikes, slippage and picking the best of hundreds of settings.
6 min read