Historical Backtesting
Replaying a strategy's rules over past market data to see what it would have done. It is the cheapest way to reject a bad idea, and it is only as trustworthy as the assumptions underneath it about fills, fees and data quality.
Open DappSoon
Learn
Proving a strategy before it costs you.
A backtest is easy to produce and easy to fool yourself with. Most impressive backtests are wrong, not because the maths failed, but because the test quietly used information the strategy would not have had at the time.
Validation is the discipline of trying to break your own result before the market does it for you.
Walk-Forward Testing
Optimise on one window, test on the next unseen window, roll forward, repeat. It mimics how a strategy is actually used, since in reality you always tune on the past and trade the future. A strategy that survives walk-forward is far more credible than one with a single strong backtest.
Out-of-Sample Testing
Holding back a portion of history the strategy was never tuned on, and only looking at it once. The moment you tune against that held-back data it stops being out-of-sample, which is why serious testing treats it as spendable exactly once.
Slippage Modeling
Assuming realistic fills rather than perfect ones. A backtest that fills at the mid price will flatter any strategy that trades often. Modelling the spread, the fee and the market impact of your own size is usually the difference between a profitable backtest and an unprofitable strategy.
Survivorship Bias
Testing only on assets that still exist today. Backtest a rotation strategy across the current top fifty coins and you have implicitly excluded everything that went to zero, which is most of what was in the top fifty a few years ago. The result will look far better than it would have been.
Look-Ahead Bias
Using information at a decision point that would not have been available yet. Common causes are using a daily close to trade during that same day, restated fundamental data, or an indicator that silently peeks at future bars. It is the most frequent reason a backtest is spectacular and live trading is not.
Overfitting Detection
Recognising when a strategy has learned the history rather than the market. Warning signs are many parameters relative to the number of trades, results that collapse when a parameter moves slightly, and performance that is excellent in-sample and mediocre out-of-sample. A strategy should sit on a plateau of parameters that all work, not a single peak.
Liquid Edge
How Liquid Edge handles this
The platform gives you the tools to test honestly. It cannot stop you deceiving yourself, which is why paper trading sits between a backtest and live capital.
- Backtest against historical data with fees and slippage assumptions applied rather than ignored
- Paper trade against live markets afterwards, which catches assumptions no historical test can
- Change one parameter and re-run to see whether results sit on a plateau or a lucky spike
Liquid Edge
Validate the idea before the market charges you for it.
Paper trade it first, go live when the numbers convince you.
- Historical Backtesting
- Walk-Forward Testing
- Out-of-Sample Testing