High Win Rate Backtests Can Be Fragile

Validation verdict for win-rate and payoff asymmetry

Win rate tells you how often a strategy wins, not whether the wins are large enough to pay for the losses. A high-win-rate system is fragile when its profit depends on keeping the left tail unusually quiet.

All figures in this article are educationally constructed examples created to explain the failure mode. They are not real strategy results or recommended thresholds.

What the headline metric obscures about win-rate and payoff asymmetry

The headline number invites a simple story: “This strategy is right most of the time.” That story feels safer than a 45% win-rate trend system, even when the high-win-rate strategy earns only a fraction of one risk unit on each winner and gives back several units on a loser.

The danger is psychological as well as statistical. A long sequence of small wins makes the equity curve look calm, so position size often rises just before the rare loss arrives. The backtest then reports a high success frequency while hiding that almost all downside is compressed into a handful of trades.

How win-rate and payoff asymmetry enters the backtest

Frequency replaces magnitude

The tester foregrounds the share of winning trades, while the average win, average loss and tail loss determine whether that frequency has economic value.

Stops are structurally wider than targets

Mean-reversion and premium-selling logic can bank many small exits while carrying positions through the very move that invalidates the setup.

Rare losses are under-sampled

A short backtest may contain only a few adverse episodes. The observed loss frequency can therefore look better than the strategy’s true exposure.

Costs attack small winners first

Spread, commission and slippage consume a larger share of a 0.22R winner than of a 2R winner, shrinking the cushion before the tail loss.

Compact reconstruction of win-rate and payoff asymmetry

CASE 01 · high win rate backtestFocused analysis · educationally constructed educational figures

Scenario	Wins	Losses	Avg win	Avg loss	Net result
Reported backtest	92	8	+0.22R	−2.20R	+2.64R
One additional ordinary loss	91	9	+0.22R	−2.20R	+0.22R
Two additional ordinary losses	90	10	+0.22R	−2.20R	−2.20R
Losses 20% larger	92	8	+0.22R	−2.64R	−0.88R

The reported 92% win rate is unchanged in the fourth row, yet the strategy moves from positive to negative because loss severity rises. The “edge” is therefore not the win rate. It is the assumption that the rare losses will remain both rare and capped at the historical size.

The test that can overturn the win-rate and payoff asymmetry verdict

Rebuild the verdict around payoff asymmetry, not frequency. Keep the same trade list and ask how little the loss tail must worsen before expectancy and Profit Factor cross below acceptable territory.

Rank every trade by loss size and show the worst 1%, 5% and 10% contribution to total loss.

Recalculate expectancy after adding one and two losses equal to the historical median tail loss.

Stress spread and slippage because small winners have the least room to absorb friction.

Compare median trade, average trade and trimmed mean; a wide gap signals that a few outcomes dominate.

Run sequence simulations at the intended position size to see whether clustered tail losses breach the capital threshold.

What trade-list analysis can and cannot identify about win-rate and payoff asymmetry

Export-level red flags for win-rate and payoff asymmetry

Win rate above 80% with payoff ratio below 0.25
Net profit is smaller than two ordinary tail losses
Median winner is close to estimated round-trip cost
The largest loss is several times the 95th-percentile loss
Position size increased after a long winning streak

What the export reveals about win-rate and payoff asymmetry

Payoff ratio, expectancy, Profit Factor and the full profit/loss distribution
Tail-loss contribution, outlier concentration and sensitivity to one or two additional losses
Cost stress, losing-streak simulations and risk-of-ruin under a stated threshold
Whether the smooth equity curve is built from many tiny wins and a small number of deep drops

What win-rate and payoff asymmetry still requires from settings, code, or market data

A CSV cannot prove why a stop was wide or whether live liquidity will enlarge the tail; inspect the strategy logic and execution context.
The historical loss distribution is not the maximum possible loss. Gap risk, outages and market discontinuities can exceed it.

ACADEMIC VALIDATION DOSSIER

Turn high win rate and payoff asymmetry into a falsifiable backtest diagnosis.

Case file 01/20 · WR-TAIL · one failure mechanism, one falsifiable protocol

Research abstract: win-rate and payoff asymmetry

Case file 01/20 · WR-TAIL · one failure mechanism, one falsifiable protocol

This article tests one central proposition: treating win rate as a proxy for quality makes a negatively skewed strategy that repeatedly earns small gains and occasionally gives them back look stable. The question is not merely whether the displayed net profit or win rate was arithmetically calculated. The deeper identification problem is whether we know what constitutes one observation, what information was available at the decision time, which assumptions are necessary for the profit to exist, and how much of the conclusion survives when those assumptions are perturbed. The research object is therefore not one performance table; it is the linked data-generation, fill-generation, estimation, selection, and capital-allocation process.

The primary estimand is the true net expectancy per trade and the sustainability of capital growth after left-tail losses. The observation unit is defined as a closed trade, with split fills from one signal consolidated into a single decision unit. Without this definition, split fills, duplicated signals, common events, synthetic prices, or timestamp conversions can be double-counted as independent evidence. A larger row count does not necessarily contain more independent information. An academically defensible analysis fixes the relationship between the observation unit and the estimand before it reports sample size, standard error, or statistical confidence.

The principal sensitivity axes are average-loss multiplier, tail-loss frequency, and losing-cluster length. The hidden state is the low-frequency, high-severity left tail, temporal clustering of losses, and payoff asymmetry. In particular, gap losses beyond the ordinary stop can remain absent from the sample while win rate keeps rising. Means and medians alone are incapable of describing that mechanism, so the analysis combines central estimates with lower quantiles, expected shortfall, sign stability, boundary-hitting frequency, and contribution concentration. The objective is not to find one pessimistic number, but to map the full region in which the original conclusion changes sign or ceases to be economically usable.

The conclusion does not attempt to prove that a backtest is good. It separates the component that remains after attempted falsification from the component that disappears when assumptions are reconstructed. The governing decision principle is to gate the strategy on net expectancy, payoff ratio, skewness, expected shortfall, and block-resampled ruin probability rather than win rate alone. This is not trading advice; it is a research procedure for measuring how much evidentiary weight a TradingView trade export can carry. Liquidity not present in the file, broker-specific rules, future regimes, outages, and gaps require separate evidence, and statistical survival never guarantees future profit.

The numerical values illustrate the method for high win rate and payoff asymmetry; they are not a real strategy, client record, or forecast.

Hypotheses and identification target for win-rate and payoff asymmetry

the true net expectancy per trade and the sustainability of capital growth after left-tail losses

Null hypothesis / H₀

H₀ for win-rate and payoff asymmetry: The reported performance is not materially dependent on the suspected failure mechanism and survives reasonable perturbations.

Alternative hypothesis / H₁

H₁ for win-rate and payoff asymmetry: The reported performance depends materially on the suspected failure mechanism and deteriorates after reconstruction, perturbation, or dependence-aware resampling.

Estimand

the true net expectancy per trade and the sustainability of capital growth after left-tail losses

Observation unit

a closed trade, with split fills from one signal consolidated into a single decision unit

Latent mechanism

the low-frequency, high-severity left tail, temporal clustering of losses, and payoff asymmetry

Stress axes

average-loss multiplier, tail-loss frequency, and losing-cluster length

Formal estimands for win-rate and payoff asymmetry

Definitions precede inference.

E[R] = p·μ₊ − (1−p)·|μ₋| − cNet expectancy including win frequency, average win, average loss, and per-trade cost.

p* = (|μ₋|+c)/(μ₊+|μ₋|)Break-even win rate under the stated cost convention.

ESα(L) = (1/(1−α))∫_α^1 VaR_u(L)duExpected loss in the upper tail of the loss distribution, equivalent to the left tail of returns.

Win rate91.2%

Average win+0.18R

Average loss−1.74R

Skewness−3.10

Add two tail losses黒字→赤字 / profit→loss

Illustrative recomputation design for win-rate and payoff asymmetry

For the high-win-rate asymmetry reconstruction, table values are illustrative calculations used to expose a verdict reversal; they are not a user’s observed TradingView result.

ID	Recomputation layer	Operation	Comparison	Diagnostic purpose
S0	Reported result	Restate the Strategy Tester aggregate	Base	Apparent conclusion
S1	Unit reconstruction	a closed trade, with split fills from one signal consolidated into a single decision unit	Reassess count and dependence	Information correction
S2	Independent recomputation	Rebuild price, size, cost, and currency row by row	Separate reconciliation error	Measurement validity
S3	Local stress	average-loss multiplier, tail-loss frequency, and losing-cluster length	Perturb one factor only	Causal sensitivity
S4	Tail injection	gap losses beyond the ordinary stop can remain absent from the sample while win rate keeps rising	Recompute lower quantiles and boundary hits	Capital preservation
S5	Dependence-aware resampling	Generate paths across several block lengths	Intervals and sign stability	Estimation uncertainty
S6	Selection adjustment	Log search, OOS review, and exclusions	Correct maximum-selection bias	Generalization
S7	Full gate	gate the strategy on net expectancy, payoff ratio, skewness, expected shortfall, and block-resampled ruin probability rather than win rate alone	Compare with predeclared thresholds	Pass / hold / reject

The illustrative recomputation for win-rate and payoff asymmetry changes one processing layer at a time, then combines only predeclared layers. S0 is never treated as ground truth; it is the statement to be audited. S1 and S2 ask whether the exported unit and arithmetic are coherent. S3 and S4 identify local sensitivity and tail failure. S5 changes the uncertainty model rather than the trade list. S6 adjusts for the search that preceded publication. S7 applies the same gate to every version. This order prevents an adverse result from being explained away by simultaneously changing several assumptions.

In the high-win-rate asymmetry figures, color and position encode diagnostic sensitivity only; they do not represent statistical significance or future P&L.

Diagnostic figures specific to win-rate and payoff asymmetry

Four separate visual tests; no decorative chart reuse.

Figure 1. Primary diagnostic for high win rate and payoff asymmetry. Values are methodological illustrations, not estimates of a real strategy or future return.

Figure 2. Break-even frontier for win rate and average-loss multiple. Even a rising win rate has a finite break-even boundary for average loss size. Values are illustrative recomputations, not observed performance or forecasts.

Figure 3. Tail-loss exceedance curve. Preserving loss clusters thickens the adverse tail and weakens the comfort implied by a high win rate. Values are illustrative recomputations, not observed performance or forecasts.

Figure 4. Bow-tie map from small wins to the capital boundary. Win rate is an aggregation node that can conceal several failure causes and several capital consequences at once. Values are illustrative recomputations, not observed performance or forecasts.

The primary diagnostic decomposes net expectancy, break-even win rate, payoff ratio, skewness, expected shortfall, and losing-cluster length along a causal axis. Read slope, curvature, and the first decision-boundary crossing as “average-loss multiplier, tail-loss frequency, and losing-cluster length” changes, not merely the height of the favorable point.

The two-dimensional surface exposes interaction among “average-loss multiplier, tail-loss frequency, and losing-cluster length.” Color is a normalized margin to a predeclared gate, not an empirical probability. A broad connected pass region is different evidence from a narrow isolated island.

The resampling statistic is net expectancy. Compare an IID benchmark with trade blocks that preserve losing runs and regime persistence across several block lengths, reporting the 2.5th, 50th, and 97.5th percentiles and verdict-reversal rate. Save seeds and repetitions.

The causal map traces “many small wins → an under-observed large loss → a high win-rate headline → overconfident adoption → proximity to a capital boundary.” A displayed metric is an intermediate product, not the first cause; perturb the input or assumption, rebuild trades and capital boundaries, and return to the predeclared gate.

Multi-layer audit questions for win-rate and payoff asymmetry

A result is only as strong as its weakest unresolved layer.

AUDIT LAYER 0101 · Fix the estimand

From an audit perspective, fix the estimand as “the true net expectancy per trade and the sustainability of capital growth after left-tail losses.” Do not substitute net profit, win rate, or a visually smooth curve for that target. Declare the horizon, account currency, included frictions, and operating-stop boundary before calculation. Any post-result change creates a new hypothesis and version, preventing the question from being selected after the answer is known.

AUDIT LAYER 0202 · Reconstruct the observation unit

Reconstruct the observation unit as “a closed trade, with split fills from one signal consolidated into a single decision unit” before treating rows as independent evidence. Report raw rows, parent trades, decisions, event clusters, and the denominator used for each average or standard error. Recompute net expectancy, break-even win rate, payoff ratio, skewness, expected shortfall, and losing-cluster length under more than one defensible aggregation rule so that a larger export is not mistaken for a larger information set.

AUDIT LAYER 0303 · Preserve provenance and settings

Preserve the hash of the TradingView export and the symbol, timeframe, session, timezone, order-processing settings, costs, account currency, and Pine version. For high win rate and payoff asymmetry, the low-frequency, high-severity left tail, temporal clustering of losses, and payoff asymmetry directly affects reproducibility. Keep immutable source, normalized, and analysis layers separate, with every join, deletion, imputation, and conversion recorded in a transformation ledger.

AUDIT LAYER 0404 · Separate identification from assumption

The export identifies only what can be rebuilt from recorded time, price, quantity, and P&L. left-tail events not visible in the export, including gaps, spread explosions, and unfilled limits requires additional evidence. Mark each causal link as observed, bounded by assumption, or externally unverified. This prevents the low-frequency, high-severity left tail, temporal clustering of losses, and payoff asymmetry from being presented as a confirmed fact when the available data support only an interval or conditional conclusion.

AUDIT LAYER 0505 · Reconcile row-level arithmetic

Do not adopt the platform summary as ground truth. Independently rebuild the net R series and separate win frequency from average magnitude, lower quantiles, and expected shortfall. Reconcile total and row-level differences by sign, date, symbol, and order type. If discrepancies concentrate in the exact state associated with high win rate and payoff asymmetry, treat that concentration as a primary finding rather than dismissing it as rounding.

AUDIT LAYER 0606 · Quantify finite-sample uncertainty

Report net expectancy, break-even win rate, payoff ratio, skewness, expected shortfall, and losing-cluster length with intervals or resampling distributions, not point estimates alone. Match the uncertainty method to sample size, skewness, heavy tails, censoring, and selection history. If normal, quantile, and dependence-aware methods disagree on the sign, classify the edge as unidentified and show the minimum detectable effect and lower decision bound.

AUDIT LAYER 0707 · Preserve serial and cluster dependence

Do not narrow uncertainty with an IID shuffle alone. Resample trade blocks that preserve losing runs and regime persistence using several fixed block lengths and stationary bootstrap. Preserve random seed, repetition count, wrap rule, and missing-data treatment. For each block specification, report the distribution of net expectancy, break-even win rate, payoff ratio, skewness, expected shortfall, and losing-cluster length, the rejection-side tail mass, and the rate at which the verdict changes sign.

AUDIT LAYER 0808 · Measure tails and operating boundaries

Interrogate the mechanism “the low-frequency, high-severity left tail, temporal clustering of losses, and payoff asymmetry” with lower quantiles, expected shortfall, influence, cluster length, and boundary-hitting measures. Historical maximum loss is not a loss cap. Define several absorbing or operating boundaries—capital, margin, mandate drawdown, and recovery time—and record which boundary fails first under each stress.

AUDIT LAYER 0909 · Model execution and market frictions

A flat commission deduction is not an execution model for high win rate and payoff asymmetry. Allocate spread, slippage, financing, borrow, roll, conversion, rounding, and rejected orders to the relevant unit. Recompute net expectancy, break-even win rate, payoff ratio, skewness, expected shortfall, and losing-cluster length under base, upper-quantile, and crisis states while preserving the possibility that costs and losses worsen together.

AUDIT LAYER 1010 · Count the complete search path

Count the complete population of periods, symbols, timeframes, parameters, exits, filters, and metrics that were tried. Do not detach the attractive result for high win rate and payoff asymmetry from rejected candidates, interim changes, or repeated validation reviews. Where appropriate, use PBO, SPA, and a Deflated Sharpe Ratio, and treat an unrecorded trial count as a material audit limitation.

AUDIT LAYER 1111 · Condition on market regimes

Test whether high win rate and payoff asymmetry is concentrated in one trend, volatility, liquidity, rate, or session state. Define regimes prospectively or on training data only. Report statewise net expectancy, break-even win rate, payoff ratio, skewness, expected shortfall, and losing-cluster length, occupancy, transition probabilities, and costs, then reweight the mixture to adverse but realistic future compositions.

AUDIT LAYER 1212 · Separate path, inception, and sizing

For the win-rate and payoff asymmetry case, the same trade set can follow different capital paths under another inception date, order, initial balance, rounding rule, or stop condition. Separate fixed quantity, fixed R, and percentage sizing, then use circular shifts and block orderings to recompute drawdown, recovery, and boundary hits. Equal terminal P&L does not imply equal path risk.

AUDIT LAYER 1313 · Design counterfactual stress tests

Perturb “average-loss multiplier, tail-loss frequency, and losing-cluster length” one axis at a time before creating a joint sensitivity surface. Add the negative control “symmetrize payoff magnitudes while preserving win/loss labels to test whether asymmetry, rather than win frequency, creates the result.” Predefine the grid and crisis rule so that neither the most favorable nor the most damaging cell is selected after inspection. Save the slope, curvature, and exact point where the decision boundary is crossed.

AUDIT LAYER 1414 · Verify through an independent implementation

Have a second implementation rebuild the net R series and separate win frequency from average magnitude, lower quantiles, and expected shortfall, then compare critical row-level outputs. Regression fixtures should include empty files, duplicate timestamps, extreme costs, reverse ordering, missing values, and boundary cases. Agreement between implementations is insufficient if they share the same bad input, so separate data construction and review roles where feasible.

AUDIT LAYER 1515 · Use a predeclared decision gate

Predeclare the decision rule. This case passes only if “both net expectancy and expected shortfall remain above predeclared limits across several block lengths.” Near a boundary, disclose interval width and economic materiality rather than a binary badge. If only one favorable block length, cost state, or implementation passes, classify the result as assumption-sensitive rather than robust.

AUDIT LAYER 1616 · Maintain a reproducibility ledger

The evidence ledger must store the input hash, code version, settings, exclusions, “average-loss multiplier, tail-loss frequency, and losing-cluster length,” block lengths, random seed, repetition count, and every scenario output. Keep exploratory and confirmatory results in separate namespaces and retain failed trials. When new TradingView data arrive, create a new version and track average loss, lower-one-percent loss, losing-run length, and distance from break-even win rate rather than overwriting the old result.

AUDIT LAYER 1717 · Translate statistics into capital impact

Translate statistical changes into capital consequences. A shift in expectancy, lower quantile, recovery time, or boundary risk caused by high win rate and payoff asymmetry should be mapped to trade count, capital, margin, and continuation. A small per-trade difference can compound under high turnover, while a rare loss can be decisive near an absorbing boundary.

AUDIT LAYER 1818 · Separate roles and enforce stop conditions

Separate hypothesis design, implementation, independent recalculation, and approval where practical. Stop automatically on material reconciliation error, unresolved missing data, non-reproducibility, or a predeclared threshold breach. Audit the chain “many small wins → an under-observed large loss → a high win-rate headline → overconfident adoption → proximity to a capital boundary,” and monitor average loss, lower-one-percent loss, losing-run length, and distance from break-even win rate prospectively without turning a historical pass into a promise of future profit.

Falsification protocol for win-rate and payoff asymmetry

gate the strategy on net expectancy, payoff ratio, skewness, expected shortfall, and block-resampled ruin probability rather than win rate alone

Freeze the TradingView source for the win-rate and payoff asymmetry audit

Store the export without alteration and record its hash, export time, strategy, symbol, timeframe, and settings. Preserve every column relevant to high win rate and payoff asymmetry; deletions and imputations belong only in derived tables.

Reconstruct the observation unit for win-rate and payoff asymmetry

Aggregate rows into “a closed trade, with split fills from one signal consolidated into a single decision unit,” and report raw rows, parent trades, events, and independent clusters. Recompute the critical result under another defensible aggregation.

Independently recompute the displayed win-rate and payoff asymmetry result

Independently rebuild the net R series and separate win frequency from average magnitude, lower quantiles, and expected shortfall. Reconcile row-level and aggregate outputs with Strategy Tester and preserve where discrepancies concentrate.

Isolate the win-rate and payoff asymmetry mechanism

Treat high win rate and payoff asymmetry as the principal mechanism and move “average-loss multiplier, tail-loss frequency, and losing-cluster length” one axis at a time while holding other settings fixed.

Map the operating boundary for win-rate and payoff asymmetry

Combine the primary and interacting axes on a predeclared grid and recompute net expectancy, break-even win rate, payoff ratio, skewness, expected shortfall, and losing-cluster length. Record the width and connectivity of the acceptable region and every boundary crossing.

Resample the dependence structure relevant to win-rate and payoff asymmetry

Use trade blocks that preserve losing runs and regime persistence with several fixed block lengths and stationary bootstrap. Save every random seed, repetition count, and block specification.

Inspect influence points and operating boundaries for win-rate and payoff asymmetry

For the high-win-rate asymmetry influence test, remove the largest contributor, top-k contributors, selected periods, and relevant regimes in sequence; then recompute lower-tail measures and the operating boundary.

Apply negative controls and conservative bounds to win-rate and payoff asymmetry

Symmetrize payoff magnitudes while preserving win/loss labels to test whether asymmetry, rather than win frequency, creates the result. Bound left-tail events not visible in the export, including gaps, spread explosions, and unfilled limits as unobserved factors rather than elevating the optimistic value into the final answer.

Apply the predeclared gate to win-rate and payoff asymmetry

Do not move the threshold after seeing results. Compare with “both net expectancy and expected shortfall remain above predeclared limits across several block lengths,” and distinguish pass, hold, and reject. Any unresolved material mismatch causes a hold.

Save a reproducible evidence package for win-rate and payoff asymmetry

Bundle the source, transformation ledger, formulas, figures, all scenarios, failure logs, and code version for rerun in another environment. Prospectively monitor average loss, lower-one-percent loss, losing-run length, and distance from break-even win rate.

Decision gate for win-rate and payoff asymmetry

Reject the story before trusting the curve.

How to read the win-rate and payoff asymmetry figures and equations

The figures for win-rate and payoff asymmetry use illustrative recomputations constructed to expose this specific failure mode. Do not infer statistical significance from line position or color alone; first verify the estimand, units, denominator, censoring rule, and cost sign defined by the equations. A sensitivity surface is not a causal estimate. It shows how a conclusion changes only within the stated assumptions. Resampling should compare an IID shuffle with stationary and block bootstrap procedures across several block lengths so that loss clustering and regime persistence are not silently destroyed. Store the random seed, iteration count, block length, bandwidth, and missing-data treatment, and claim reproducibility only after an independent implementation reproduces the same aggregates.

This case passes only if “both net expectancy and expected shortfall remain above predeclared limits across several block lengths” across reconstructed values, local perturbations, joint sensitivity, dependence-preserving resampling, and the negative control, with no material sign reversal or unresolved reconciliation error. A pass is limited evidence against the stated failure mode, not certification of future profit.

The estimand and observation unit were fixed before outcomes were reviewed
For high-win-rate asymmetry, any material disagreement between reported and independently recomputed values must be resolved or explicitly explained.
The high-win-rate asymmetry claim passes this gate only when its acceptable stress region is broad and connected rather than one isolated favorable island.
The sign of the high-win-rate asymmetry estimate must remain stable across defensible block lengths, saved seeds, and reasonable interval methods.
For win-rate and payoff asymmetry, economic margin remains after deleting the largest and top-five contributors and key regimes
For win-rate and payoff asymmetry, conservative cost, fill, and capital-boundary scenarios remain inside the stopping mandate

6/6required gates · not a performance forecast

Limitations, external validity, and reproducibility of the win-rate and payoff asymmetry audit

Every inference has a boundary.

The first limitation is that a trade export does not contain the complete market state. If order-book depth, queue position, network latency, rejected orders, broker liquidity, or realized financing history is absent, the true net expectancy per trade and the sustainability of capital growth after left-tail losses remains model-mediated. Model outputs should be displayed as scenario ranges and must not be formatted as though they were directly observed facts.

A second limitation specific to the win-rate and payoff asymmetry analysis is structural change. A long historical sample does not guarantee a common population when market rules, participants, volatility, rates, spreads, data construction, or Pine execution semantics change. Do not increase nominal sample size by indiscriminately pooling old periods. Estimate rolling and regime-conditioned behavior and test parameter stability around detected changes.

A third limitation specific to the win-rate and payoff asymmetry analysis is reuse of the diagnostic battery. Applying these tests repeatedly to the same data and editing the strategy until it passes turns the diagnostic process itself into another optimizer. Every post-test edit starts a new model version and requires untouched or prospective evidence. A test chosen after reading the outcome belongs to exploration and cannot be counted as independent confirmation.

A fourth limitation for the win-rate and payoff asymmetry analysis is the distinction between statistical survival and operational suitability. Behavioral tolerance, locked capital, tax, regulation, outages, account terms, order-size limits, market-order restrictions, and liquidity discontinuities cannot be resolved from a CSV alone. The lab is a diagnostic for discovering hidden failure risk earlier; it is not investment advice, a performance warranty, or a guarantee of bounded loss. User-specific constraints remain a separate decision layer.

LIMIT 01Identification boundary

The estimand “the true net expectancy per trade and the sustainability of capital growth after left-tail losses” is identified only within the columns present in the TradingView export and the stated assumptions. If left-tail events not visible in the export, including gaps, spread explosions, and unfilled limits cannot be observed, report bounds rather than a false point estimate.

LIMIT 02Structural change

Past estimates of high win rate and payoff asymmetry need not belong to the same population after changes in rules, participants, volatility, costs, or data specifications. Track average loss, lower-one-percent loss, losing-run length, and distance from break-even win rate in rolling and regime-specific windows.

LIMIT 03Reuse of the diagnostic

For high-win-rate asymmetry, repeatedly applying the same diagnostic battery and editing until it passes turns verification into another optimizer. Every post-audit change therefore creates a new model version and requires untouched evidence.

LIMIT 04Operational suitability

Even if both net expectancy and expected shortfall remain above predeclared limits across several block lengths, the analysis does not establish tax, regulatory, behavioral, liquidity, order-size, or systems suitability. Separate statistical diagnosis from live-operating approval.

LIMIT 05Missing data and anomalies

Deleting observations related to the low-frequency, high-severity left tail, temporal clustering of losses, and payoff asymmetry may improve the result. Compare no deletion, conservative imputation, and worst-case imputation, and display how net expectancy, break-even win rate, payoff ratio, skewness, expected shortfall, and losing-cluster length changes.

LIMIT 06Negative controls

Run the control “symmetrize payoff magnitudes while preserving win/loss labels to test whether asymmetry, rather than win frequency, creates the result.” If the control performs similarly, suspect processing rules or common market drift before attributing performance to the strategy.

LIMIT 07Prospective monitoring

After a provisional pass, log average loss, lower-one-percent loss, losing-run length, and distance from break-even win rate sequentially and stop on persistent departures from the predeclared predictive range. Diagnose implementation drift before reoptimizing history.

LIMIT 08Common-mode failure and reporting

Multiple methods can agree because they share the same bad input or the same mechanism “the low-frequency, high-severity left tail, temporal clustering of losses, and payoff asymmetry.” Give lower-tail outcomes, failed scenarios, and unresolved mismatches the same visual prominence as favorable results; test count is not proof of correctness.

10A

Independent and adversarial findings for win-rate and payoff asymmetry

The high-win-rate asymmetry case has a separate review line for formulas, chart encodings, data definitions, and falsifiability so agreement on one layer cannot mask failure on another.

The formula audit checks numerator, denominator, sign, unit, domain, and every conditioning assumption as one system. The material caution for this case is: The expectancy equation treats mean win μ₊ and absolute mean loss |μ₋| as positive magnitudes and c as a per-trade deduction. If cost depends on side, size, or outcome, use conditional terms. Expected shortfall is defined by the quantile integral over the upper 1−α tail of the loss variable L, rather than a naive conditional mean when discrete outcomes place probability mass at VaR. This is equivalent in direction to the left tail of returns. A correct symbolic expression can still calculate the wrong quantity when a column, currency, time unit, or fee sign is misdefined, so those mappings are part of the mathematical audit.

The figure audit assigns distinct jobs: Figure 1 diagnoses high win rate and payoff asymmetry; Figure 2 maps joint sensitivity; Figure 3 shows the dependence-preserving distribution of net expectancy; Figure 4 traces causal propagation. Color denotes distance to a predeclared gate, not probability or observed performance. Axis units, zero, quantiles, censoring, and bounds must agree with captions and tables. A smooth SVG line is explanatory geometry, not evidence of estimation precision.

The adversarial test does not cherry-pick one hostile scenario. It uses the negative control “symmetrize payoff magnitudes while preserving win/loss labels to test whether asymmetry, rather than win frequency, creates the result,” resamples trade blocks that preserve losing runs and regime persistence at several block lengths, and bounds left-tail events not visible in the export, including gaps, spread explosions, and unfilled limits as unobserved factors. Repetitions, seeds, exclusions, block specifications, and plotting range are frozen before results so the implementer cannot tune the audit after seeing the answer.

The independent conclusion is restricted to whether “both net expectancy and expected shortfall remain above predeclared limits across several block lengths.” It does not certify a good strategy or future profit. Any material reconciliation error, formula-domain violation, table-figure contradiction, sign reversal across defensible block lengths, or failure to outperform the negative control produces hold or reject. Prospectively, monitor average loss, lower-one-percent loss, losing-run length, and distance from break-even win rate.

Methodological references for win-rate and payoff asymmetry

Primary methods and official platform documentation.

References for the win-rate and payoff asymmetry case provide methodological context; they do not validate the synthetic numbers in this article or certify any backtest result. TradingView documentation is used for platform semantics, while statistical papers motivate uncertainty and selection controls.

Previous case← Why 500 Trades May Not Be 500 Independent Observations
Backtest Analysis Articles
Next caseThe One Trade That Made Your Backtest Look Profitable →

Frequently asked questions about win-rate and payoff asymmetry

Is a 90% win rate automatically suspicious?

No. It is a prompt to inspect payoff asymmetry. A high win rate can be valid, but it must still cover average losses, costs and plausible tail expansion.

Which metric should replace win rate?

No single metric replaces it. Read win rate with average win/loss, expectancy, Profit Factor, tail contribution, drawdown and sequence risk.

Can the free health check detect this?

The free analysis can expose the basic payoff and distribution problem from a trade list. More advanced saving, comparison and robustness workflows depend on the selected plan and available columns.

Backtest Analysis

Can a backtest exposed to win-rate and payoff asymmetry be trusted?

Do not judge the high-win-rate asymmetry case from a finished equity curve alone. Use the TradingView trade list to inspect the mechanism-specific concentration, path, cost, timing, and dependence evidence shown on this page.

Analyze My Backtest Verification Status View Plans

Important limitations for the win-rate and payoff asymmetry analysis

This article provides educational, descriptive analysis of constructed backtest failure examples. It is not investment advice, a buy or sell signal, a forecast or a promise of performance. Backtest results depend on data, code, broker-emulator assumptions, costs, sizing and market structure. TradingView is a trademark of TradingView, Inc.; SG Group is independent and does not claim endorsement or sponsorship by TradingView.

Counterpart: 高勝率ほど危険？勝率がバックテストの脆弱性を隠す理由