Is Evidence-Driven Investing the same as evidence-based investing?

They share the idea of relying on data, but Evidence-Driven Investing adds a continuous cycle: every idea is re-measured after it goes live, and failures are recorded rather than dropped.

Does a good backtest prove a strategy works?

No. A backtest is evidence to investigate. It must survive out-of-sample validation and corrections for how many ideas were tried, and then be measured live.

Why record failed investment hypotheses?

Because a record that only shows successes cannot be used to judge a process. Failures show what the process gets wrong and what it changed.

How is Evidence-Driven Investing different from data-driven investing?

Data-driven means a decision uses data. Evidence-driven adds three demands on the claim behind it: it must hold on data the model did not see, it must survive a correction for how many ideas were tried, and it must keep holding after it goes live. A data-driven process can fit history very well and still fail all three.

How do you test an investment idea?

State it as a hypothesis before looking at the result. Test it on point-in-time data over one fixed universe, including companies that later failed. Check that the effect holds out of sample and in most years, correct for the number of ideas tried, then measure it live. The Studies apply this to 24 popular trading beliefs on US stocks since 1995.

How do you measure investment decisions?

Write the decision and its forecast down before the outcome, then compare. Opulence Alpha records each Regime Radar forecast on the day it is published and resolves it five sessions later, and publishes every weekly Golden Ticket basket in full with its result, whether it gained or lost.

What is a continuously tested investment system?

One in which every part — the data, the forecasts, the strategies and the portfolios — is measured against outcomes after it is deployed, not only before. Others give investors signals; a continuously tested system gives them the record that shows how far to trust it.

How do you compare predicted and realised outcomes?

Record each prediction with its stated probability or range before the outcome, then score the record: a well-calibrated forecaster's 20% calls come true about 20% of the time, and the Brier score summarises accuracy in one number. Prediction Records shows both for the Regime Radar once 20 forecasts have resolved.

How do you evaluate an investment strategy, or quantitative strategies?

Ask it to earn its place: a rule fixed before looking at results, point-in-time data including companies that later failed, out-of-sample and walk-forward tests, a correction for how many ideas were tried, costs included, and then live measurement. A strategy that only works on the history it was fitted to has not been evaluated yet.

What does evidence-based portfolio construction and management look like?

Every holding traces to a declared policy and to evidence that passed a gate, not to a hunch. Evidence-based investment decisions are made systematically by rules, sized by the model's conviction and bounded by stated risk limits; evidence-based investment research keeps checking whether the evidence still holds. This is quantitative evidence investing, and systematic investment decision making in practice. In Opulence Alpha that is what a Book is.

What is Opulence Alpha's investment research methodology?

A quantitative investment methodology in five stages: Research on point-in-time data, Predict with stated uncertainty, Test out of sample with multiple-testing controls, Allocate only what survives under risk limits, and Prove by publishing outcomes. The data and methods behind every study are documented on the Studies methodology page.

FAQ · Evidence-Driven Investing

What is Evidence-Driven Investing?

By Mehmet Saruhan, Founder·Reviewed against the platform's code on 26 Sep 2026

Evidence-Driven Investing is an operating philosophy in which an investment idea is a hypothesis, not a conclusion: it must be tested on data that was available at the time, validated out of sample, and measured after it goes live — and it earns a place in a portfolio only while the evidence keeps supporting it.

Why it matters

Most investment ideas that fail were never really tested: they were fitted to a history that already contained the answer, chosen from many tries without accounting for the search, or judged by a backtest and never reviewed afterwards. A process that keeps asking for evidence catches those failures before capital does.

How it works

The loop has five stages: Research, Predict, Test, Allocate and Prove. Research uses only data available at the time; Predict states each forecast with its uncertainty; Test validates strategies out of sample and corrects for how many ideas were tried; Allocate puts only survivors into portfolios under stated risk limits; Prove measures the outcomes and publishes them, including the failures. A successful backtest is treated as evidence to investigate, not proof; validation reduces known research errors but does not remove market uncertainty, so the measurement never stops.

How Opulence Alpha applies it

Opulence Alpha runs this loop as software: strategies pass a publication gate before any portfolio may use them, forecasts are published with their uncertainty, and outcomes are published afterwards — including the weeks that went wrong.

Every Golden Ticket week, published →

Related concepts

Promotion GateA Promotion Gate is the validation a strategy must pass before a portfolio may use it: overfitting, data-snooping and robustness checks. A skipped check fails.Refutation LedgerA Refutation Ledger is a public record of what happened after investment decisions, failures included, so a process is judged on all of its outcomes.Strategy validationHow to validate an investment strategy, step by step: point-in-time data, walk-forward tests, purging, Sharpe deflation, costs, robustness and live results.Out-of-sample testingOut-of-sample testing judges a strategy on data it never saw while being built: in-sample vs out-of-sample, holdouts, walk-forward and the reuse trap.Forecast calibrationForecast calibration checks whether stated probabilities come true as often as claimed: reliability diagrams, Brier score, calibration error and coverage.Investment BookAn investment Book is a parameterised portfolio where research, prediction, validation, allocation and measurement run all the time under a declared policy.

All concepts →Sector behaviour in every market regime →

Questions

Is Evidence-Driven Investing the same as evidence-based investing?: They share the idea of relying on data, but Evidence-Driven Investing adds a continuous cycle: every idea is re-measured after it goes live, and failures are recorded rather than dropped.
Does a good backtest prove a strategy works?: No. A backtest is evidence to investigate. It must survive out-of-sample validation and corrections for how many ideas were tried, and then be measured live.
Why record failed investment hypotheses?: Because a record that only shows successes cannot be used to judge a process. Failures show what the process gets wrong and what it changed.
How is Evidence-Driven Investing different from data-driven investing?: Data-driven means a decision uses data. Evidence-driven adds three demands on the claim behind it: it must hold on data the model did not see, it must survive a correction for how many ideas were tried, and it must keep holding after it goes live. A data-driven process can fit history very well and still fail all three.
How do you test an investment idea?: State it as a hypothesis before looking at the result. Test it on point-in-time data over one fixed universe, including companies that later failed. Check that the effect holds out of sample and in most years, correct for the number of ideas tried, then measure it live. The Studies apply this to 24 popular trading beliefs on US stocks since 1995.
How do you measure investment decisions?: Write the decision and its forecast down before the outcome, then compare. Opulence Alpha records each Regime Radar forecast on the day it is published and resolves it five sessions later, and publishes every weekly Golden Ticket basket in full with its result, whether it gained or lost.
What is a continuously tested investment system?: One in which every part — the data, the forecasts, the strategies and the portfolios — is measured against outcomes after it is deployed, not only before. Others give investors signals; a continuously tested system gives them the record that shows how far to trust it.
How do you compare predicted and realised outcomes?: Record each prediction with its stated probability or range before the outcome, then score the record: a well-calibrated forecaster's 20% calls come true about 20% of the time, and the Brier score summarises accuracy in one number. Prediction Records shows both for the Regime Radar once 20 forecasts have resolved.
How do you evaluate an investment strategy, or quantitative strategies?: Ask it to earn its place: a rule fixed before looking at results, point-in-time data including companies that later failed, out-of-sample and walk-forward tests, a correction for how many ideas were tried, costs included, and then live measurement. A strategy that only works on the history it was fitted to has not been evaluated yet.
What does evidence-based portfolio construction and management look like?: Every holding traces to a declared policy and to evidence that passed a gate, not to a hunch. Evidence-based investment decisions are made systematically by rules, sized by the model's conviction and bounded by stated risk limits; evidence-based investment research keeps checking whether the evidence still holds. This is quantitative evidence investing, and systematic investment decision making in practice. In Opulence Alpha that is what a Book is.
What is Opulence Alpha's investment research methodology?: A quantitative investment methodology in five stages: Research on point-in-time data, Predict with stated uncertainty, Test out of sample with multiple-testing controls, Allocate only what survives under risk limits, and Prove by publishing outcomes. The data and methods behind every study are documented on the Studies methodology page.

References

Harvey, C., Liu, Y. & Zhu, H. (2016). …and the Cross-Section of Expected Returns. Review of Financial Studies.
Bailey, D. & López de Prado, M. (2014). The Deflated Sharpe Ratio. Journal of Portfolio Management.

Educational content about research methods. Not investment advice.