Skip to main content
ForexNews24
RUEN
Topic · testing

Strategy Testing: Backtesting Without Self-Deception

A strategy without a check on history is faith, not method. The route shows how to backtest honestly: account for costs, separate overfitting from a genuine edge and read the metrics. Alongside are open data and the portal’s real runs that you can reproduce yourself.

22 articles in the path · read in order
  1. 1Drawdown: What It Is in Simple TermsDrawdown explained simply: what a decline in capital is, how it differs from a single loss, and why its shape matters more than the raw number.Glossary
  2. 2The Sharpe Ratio: What It Is in Simple TermsThe Sharpe ratio explained simply: what this metric shows, why returns must be viewed risk-adjusted, and where its blind spots lie for traders.Glossary
  3. 3How to Tell Your Trading System Is BrokenHow to tell your trading system is broken: separating a normal drawdown from real edge decay, the signs of degradation, and what to do next.Forex Basics
  4. 4Risk-On Regime: When the Market Wants Yield AgainRisk-on is the market mood where optimism returns and capital chases yield. Learn what triggers it and how it moves currencies as the mirror of risk-off.Forex Basics
  5. 5Overfitting: the main enemy of backtestingOverfitting is why a perfect backtest deceives: a flawless curve is a warning sign. Here's how to tell a robust system from one fitted to history.Backtesting
  6. 6In-sample and out-of-sample: why split the data in a backtestIn-sample and out-of-sample: why you split the data in a backtest and how it catches overfitting. A simple method for honestly testing a trading system.Backtesting
  7. 7Walk-forward analysis: methodology and examplesWalk-forward analysis: how to test a system's robustness across rolling periods. The methodology and the essence of a method that mirrors real trading.Backtesting
  8. 8Sample Size: Why a Small Sample DeceivesSample size: why a small sample of trades deceives and how many you need for conclusions. The role of sample size in evaluating a strategy.Backtesting
  9. 9Statistical Significance: When a Result Is Genuinely Non-RandomStatistical significance: when a result is genuinely non-random rather than luck. How to tell a real edge from a lucky streak when evaluating a strategy.Backtesting
  10. 10Overfitting in Statistics: When Numbers Are Fit Too WellOverfitting is when numbers fit the data too tightly and lose predictive power. See why results that look too good fail and how to avoid it.Backtesting
  11. 11Hypothesis Testing: How Not to Build Trading on GuessesHypothesis testing means checking trading ideas with data instead of trusting hunches. Learn the scientific approach that protects you from self-deception.Backtesting
  12. 12Backtest Bias: When Testing History Misleads YouBacktest bias is the set of errors that make a strategy look better than reality. Learn the main distortions that inflate results and how to avoid them.Backtesting
  13. 13Survivorship Bias: Why You See Only SurvivorsSurvivorship bias means you see only the survivors, which distorts how you judge strategies and success. Learn the survivor's error and how it deceives.Backtesting
  14. 14Walk-Forward Analysis: Testing a Strategy in MotionWalk-forward analysis tests a strategy on rolling periods that mimic real trading with re-optimization. Learn why this method is stricter than a backtest.Backtesting
  15. 15Out-of-Sample Testing: Checking on Unfamiliar DataOut-of-sample testing checks a strategy on data it has never seen. Learn this key method for an honest evaluation of a real trading advantage.Backtesting
  16. 16Edge Decay: Why an Advantage Fades Over TimeEdge decay explains why a real trading advantage weakens over time. Learn how to recognize a fading edge and adapt to an ever-changing market.Backtesting
  17. 17The Day-of-Week Effect in Forex, TestedIs Monday really quiet and Friday wild? Median daily ranges by weekday across 12 pairs over two years, and which piece of folklore survived.Backtesting
  18. 18Slippage: Backtest Versus RealityWhy backtests overstate results: the asymmetry between stops and limit entries, how to model slippage, and the sensitivity test that matters.Backtesting
  19. 19Why You Need a Buy-and-Hold BenchmarkA return figure means nothing on its own. What a benchmark shows, what holding returned on our data, and where the comparison becomes dishonest.Backtesting
  20. 20Look-Ahead Bias: Four Ways to Peek at the FutureHow future information slips into a backtest unnoticed, the four typical leaks, and three checks that catch them.Backtesting
  21. 21Data Mining Bias: How Testing Enough Ideas Fools YouTest enough variants and the best will look excellent even with no edge at all — demonstrated on our own ten backtests.Backtesting
  22. 22MACD Signal Line Crossover: Backtest on 12 PairsThe classic MACD signal line crossover tested on 12 currency pairs over 518 daily bars. Zero profitable pairs, and the reason is visible in two other numbers.Backtesting

Topic FAQ

What is a backtest in plain words?

A backtest checks the rules of a strategy on historical data: we apply the signals to past quotes and see what would have happened. It is a way to weed out clearly non-working ideas before risking real money — but the past does not guarantee the future.

Why does a good backtest fail live?

Usually because of overfitting (parameters tuned to history), unaccounted costs (spread, commission, slippage) or look-ahead bias in the code. An honest test charges costs and is validated on a held-out sample.

How can I verify a backtest myself?

Take an open sample from the "Data" section, apply the described rules and costs, and compare the figures with the run on the site. If the result reproduces, it can be trusted; if not, the method is in question.

From research to application

In our Allocation product we implemented these algorithms with all the nuances covered across the portal.

Learn about Allocation