Strategy Validation Checklist

The risk questions to answer before trusting a strategy with real capital, so that a curve that looks good on paper does not quietly hide the flaw that empties the account.

Strategy Validation Checklist: Before committing capital, validate a strategy on risk grounds, not just returns: confirm the sample is large enough, that realistic costs and slippage are included, that the drawdown and risk of ruin are survivable, and that the results are not an artefact of overfitting or one market regime. Then prove it at small size live before scaling. A backtest is a hypothesis, not a promise. This is educational method, not a signal.

A strategy is trustworthy only after its risks are understood, and an attractive equity curve can hide the very flaw that ruins it, thin data, ignored costs, an unsurvivable drawdown, or a fit to one lucky regime. This checklist judges a strategy on whether its risk is real and survivable, because a positive backtest is a hypothesis to be stress-tested, never a forecast of profit. Once trading it, the earlier review lists take over.

Evidence and robustness

Confirm the test covers enough trades to be meaningful; a handful of trades cannot distinguish edge from luck.
Check the sample spans different market regimes, trending, ranging, high and low India VIX, not just one favourable stretch.
Confirm results are not driven by a few outsized trades whose absence would erase the edge.
Test robustness to small parameter changes; if a slight tweak collapses the results, the edge is likely overfit.
Reserve out-of-sample or walk-forward data the strategy was not tuned on, and confirm it still holds there.
Count the parameters: the more knobs tuned to past data, the higher the risk of curve-fitting noise.
Confirm the logic has an economic or behavioural rationale, not just a pattern that happened to fit history.

Cost and execution reality

Include realistic brokerage, STT, exchange fees and GST in every simulated trade, not gross prices.
Model slippage between signal and fill, and widen it for fast markets and less liquid instruments.
For options, account for bid-ask spread and thin liquidity in far OTM strikes, where exits can be costly or impossible.
Confirm the strategy's turnover is affordable; a high-frequency edge can vanish entirely once real costs are applied.
Check that entries and exits are actually executable at the assumed prices, not at unrealistic touches of the high or low.
Account for impact if position size is large relative to the instrument's liquidity.
Verify NSE lot sizes make the intended position size feasible for your capital.

Risk survivability

Measure the maximum drawdown in the test and ask honestly whether you could withstand it, in money and in nerve.
Translate that drawdown into the recovery required, and confirm it is climbable rather than crippling.
Estimate the risk of ruin at your intended position size, and reduce size until a realistic losing streak is survivable.
Identify the longest losing streak in the test and prepare for a live one at least as long.
Check the strategy's behaviour in its worst historical window, not just its average, since you must survive the worst case.
Confirm the sizing plan keeps per-trade and portfolio risk within your caps in live conditions.
Prove it with a small-size live trial first; live slippage, fills and psychology differ from any backtest.

A strategy that clears these risk tests is worth trusting with measured capital, not certain to profit, but unlikely to hide a fatal flaw. Scale it only gradually as live results confirm the backtest. For related depth, see the Risk Management Cheat Sheet.

Frequently asked questions

How many trades does a strategy need before I trust it?

There is no magic number, but a few dozen trades cannot separate edge from luck, and hundreds across varied conditions give far more confidence. What matters as much as the count is that the sample spans different regimes and is not dominated by a handful of outsized results. Treat a thin sample as an untested hypothesis rather than proof.

Why include costs in a backtest?

Because brokerage, STT, exchange fees, GST and slippage can turn a gross-profitable strategy into a net loser, especially at high turnover or in far-OTM options. A backtest on gross prices flatters the edge and hides whether the strategy can survive its own trading costs. Realistic costs are often the difference between a curve that works on paper and one that works in reality.

What is overfitting and how do I detect it?

Overfitting is tuning a strategy so closely to past data that it captures noise rather than a repeatable edge, which then fails live. Warning signs are many tuned parameters, results that collapse under small parameter changes, and strong in-sample performance that does not survive out-of-sample or walk-forward data. A clear economic or behavioural rationale is a further guard against fitting coincidence.

How does risk of ruin fit into validation?

It tells you whether your intended position size is survivable given the strategy's win rate and payoff. Even a positive edge can be ruined by betting too large through a normal losing streak, so estimate the risk of ruin and cut size until a realistic run of losses leaves capital and composure intact. Validation is incomplete without confirming the sizing is survivable.

Should I trade a validated strategy at full size immediately?

No. Prove it with a small-size live trial first, because live slippage, partial fills, latency and your own psychology differ from any backtest. Scale up only gradually as live results confirm the expected behaviour. Starting small caps the cost of any flaw the backtest missed while you gather real evidence.

Does passing this checklist mean the strategy will be profitable?

No. Validation reduces the chance of a hidden fatal flaw and confirms the risk is survivable, but it cannot promise future profit, since markets change and past results never guarantee future ones. The goal is a strategy whose risks you understand and can withstand, traded at a size that keeps you solvent while its edge, if real, has time to show.

Last reviewed 12 July 2026. Educational content only — not investment advice.