Skip to main content

Category guide · Institutional investing

What is alpha capture?

A hand places an idea record into an archive, connected by a red thread to its source papers.

A system for recording investment ideas, tracking their subsequent performance, and evaluating the signals and contributors behind them.

The category began with submitted trade ideas and later expanded to disclosed holdings. Alpha Capture is being designed to take that discipline upstream: from public research, through the investor's interpretation, to a review of what happened next.

Start with the definition

The short definition

An alpha capture system turns investment recommendations into structured, measurable records.

The category is evolving

From capturing a call to understanding a decision.

Conventional systems begin when a recommendation or position enters the record. Alpha Capture is being designed to begin one step earlier, with the research that changed the investor's view.

Traditional alpha capture

Make external recommendations comparable.

  • Capture submitted ideas with an instrument, direction, rationale, timeframe, and conviction
  • Or infer ideas from disclosed positions, subject to the limits of the disclosure
  • Monitor subsequent performance
  • Evaluate ideas and contributors over time

Alpha Capture's design direction

Preserve the research and reasoning behind the record.

  • Surface possible trade ideas and emerging trends within public research
  • Keep the publisher's thesis distinct from the investor's interpretation
  • Record how a view might be expressed and over what horizon
  • Review the thesis, timing, expression, and source usefulness separately

The unit of learning

The record—not the return—is the unit of learning.

A return is observed at the end. The useful evidence begins earlier, with a dated account of what entered the process, what the investor understood, and what they chose before the outcome was known.

Origin

Preserve the claim as it entered the process.

Source
The contributor or research item, with its publication date and provenance.
Thesis
The view being advanced, its rationale, horizon, and stated level of conviction.

Decision

Separate the published view from the investor's own judgement.

Interpretation
What the investor accepted, rejected, or changed after applying their own context.
Expression
The instrument, direction, and horizon that could test the interpreted view.

Evidence

Bring the outcome back to the decision that preceded it.

Outcome
What happened over the declared review window—not a result chosen after the fact.
Review
Which parts of the thesis, timing, expression, and process held up under scrutiny.

Why return is not enough

One number cannot explain an investment decision.

A good return can conceal a weak process. A poor return can condemn a sound thesis that was expressed badly or assessed too early.

01

Thesis

Did the claimed mechanism occur, and was the view directionally sound? A positive return does not show which part of the reasoning was right.

02

Expression

Did the chosen instrument isolate the intended exposure, or did unrelated risks dominate the result?

03

Timing

Was the horizon appropriate? A sound long-term view and a stale short-term signal should not be judged through the same window.

04

Decision quality

Was the conclusion coherent given the information available then—even if the eventual outcome was unfavourable?

Source intelligence

A source is not simply “good” or “bad.”

A source can be accurate but too late for the relevant horizon. It can be original but difficult to express. A contributor can be useful in one domain and uninformative in another. Data quality, reporting lag, originality, time decay, and repeatability answer different questions; one league-table score cannot explain them all.

See what Alpha Capture is being built to do

Where alpha capture fails

Systematising judgment creates its own risks.

A serious implementation must make these failure modes visible rather than pretending that more data automatically means better decisions.

Selective capture

Recording only promoted or memorable ideas makes every source look better than the complete intake would suggest.

Retrospective rewriting

Once the outcome is visible, a tidy explanation can replace what was actually known and believed at the time.

Stale or incomplete inputs

Reporting lags, amendments, and missing positions can make an apparently precise record materially wrong.

Hidden context

A disclosed holding may be one side of a pair, hedge, or corporate-action trade rather than a standalone positive view.

Mismatched horizons

An idea with fast time decay cannot be assessed fairly using delayed data or an arbitrary review window.

False comparability

Different assets, horizons, and expressions do not become equivalent merely because each produces a return.

Sources and scope

Established category. Proposed extension.

The category description above is grounded in the alpha capture system overview and Mirlesse and Lhabitant's Alpha Capture Systems: Past, Present, and Future Directions. The research-to-decision model is Alpha Capture's design direction; it is not presented as the definition of every alpha capture system.

Help shape the workflow

Does this reflect how research moves your view?

Alpha Capture is still being designed. We want early users to challenge the method before the product is released.