VALIDATE / PIPELINE

How to test strategy robustness: holdout, WFO, DSR and PBO.

Separate candidate selection from evidence: frozen holdout, walk-forward, DSR, PBO, synthetic and cross-asset checks raise the bar for an attractive result.

Open laboratory Explore the details ↓

EVIDENCE / VERDICTRESILIENT

UNSEEN

HOLDOUTPASSDSR0.93PBO18%

Interface schema · values are illustrative

SYSTEM DOMAIN03 / VALIDATE

Challenge the candidate before risk

DISCOVER
OPTIMIZE
VALIDATE
RUN
OBSERVE

METHOD / OBSERVATION / LIMIT

A result is a record, not a green badge.

The page follows the same order as the application: declare the question, inspect the run, then state what the data cannot prove.

01
Frozen holdout
Reserve data that parameter search never sees.
02
DSR
Adjust Sharpe evidence for trial count, sample length and return shape.
03
PBO
Estimate the probability that the in-sample winner fails out of sample.
04
Walk-forward
Test stability across successive training and validation windows.

Frozen holdout

Reserve data that parameter search never sees.

DSR

Adjust Sharpe evidence for trial count, sample length and return shape.

PBO

Estimate the probability that the in-sample winner fails out of sample.

Walk-forward

Test stability across successive training and validation windows.

DECISION FRAMEWORK

Which independent test could still reject the candidate?

Treat holdout, time stability, multiple-testing correction and stress paths as separate evidence. Agreement matters more than one green metric.

Protect unseen data

Record every trial

Accept FAIL and insufficient data

OUTPUT

Validation verdict

An honest verdict

Not every result should be green

A status describes the current validation procedure. It never guarantees a future result.

PASS

No blocking signal was found by the current test set.

YELLOW

Evidence is incomplete or unstable.

FAIL

A critical validation test failed.

INSUFFICIENT DATA

There are too few observations or trades for a conclusion.

Important

Historical, statistical and model results do not promise future returns. kquant provides computational tools and technical execution, but does not assess whether a specific trade is suitable for a user.