OptiNod Academy
Backtest Validation
Browse all 9 parts in the Backtest Validation series and start from the section you need.
What this series covers
Find where a backtest and live trading diverge, remove trading costs and test-setup errors, filter out overfitting, and read the result metrics last.
Where to start
Backtests vs live trading — find the first trade that diverged
Later parts reuse the terms the earlier ones define, so reading in order leaves fewer gaps.
Learning path
1. Backtest versus live
Before validating anything, see where the two equity curves separate.
- Backtests vs live trading — find the first trade that diverged Match code, market, and execution settings first. Then compare signals, orders, fills, and costs to test possible causes instead of diagnosing an equity curve by appearance.
2. Trading costs and test setup
Apply fees, slippage and liquidity, then strip out lookahead and repainting.
- Slippage, Fees, and Liquidity: The Trading Costs Backtests Hide Calculate fees and execution costs on the same notional. Check double-counted spread, the denominator behind trade frequency, and whether the book can take the actual order size.
- Lookahead and repainting — verify when a signal became knowable Separate changing realtime values, unconfirmed higher-timeframe data, past plotting, and future leakage. Test when a signal first became available, not just where its marker appears.
3. Overfitting and procedure checks
Tell a lucky setting from a durable one, and validate the optimization procedure itself.
- Overfitting: Only Settings With a Plateau Around the Peak Survive Live Trading The parameter set with the best backtest return is often just a curve fitted to historical noise. In live trading, only plateau-shaped performance curves, where nearby settings also perform well, tend to hold up.
- Walk-Forward Analysis: How to Validate the Optimization Process Itself A score optimized over the full dataset is a score fitted with hindsight. Walk-forward analysis tests the process itself by selecting settings in-sample and evaluating them only on the next out-of-sample period.
4. Reading the result metrics
See what robustness, profit factor, Sharpe and Sortino each leave out.
- The Three Robustness Scores — How to Decide Which of Two Equal-Return Settings to Trust How OptiNod's three robustness scores (Neighborhood Stability, Edge Consistency, Top-N Consistency) are calculated, and how to read the overall score and parameter sensitivity when choosing between two settings with the same return.
- Profit Factor: The Sample Size and Distribution Hidden Inside One Number Do not trust profit factor on trade count alone. Together, check profit concentration, whether costs are included, win rate and realized average reward-to-risk, and periods left out of the test.
- Sharpe and Sortino: Performance Divided by Volatility, and the Problem With Penalizing Upside Comparing strategies by return alone ignores how differently they may behave on the way to the same result. This article explains the limits of the Sharpe ratio and why the Sortino ratio is often closer to real trading conditions.
5. Practice in TradingView
Set up the TradingView Strategy Tester and export your trade history to check the results.
- TradingView backtesting — from initial settings to CSV analysis Apply a strategy, set capital, fees, slippage and dates, diagnose zero trades, read results and analyze the trade CSV. Keep the conditions so the next test is comparable.