Skip to content
Formica ZetaTrading
Menu
NewsSignal roomTest benchSealed envelopesGuidesServicesAboutContact
IT
Services

Strategy audit

Your strategy goes through the lab's five gates, with real data and real costs. You receive a report that says whether it holds, where it breaks and what would change things.

Price
Custom quote

What you get

A written report, in plain words and with the numbers, answering five questions in the order they should be asked:

  1. Do costs eat the stop? Spread and swap compared with the stop distance, instrument by instrument and timeframe by timeframe. Many strategies die here, before any other measure.
  2. Does it work on data it has never seen? The rule is tuned on one period and tested on another. If the edge disappears, it was memory, not method.
  3. Does it beat chance? The same rule with levels, hours or signals drawn at random: if a random rule gets the same result, yours is adding nothing.
  4. Does it hold across the years? Year by year, regime by regime (zero rates, high rates, crises). An edge that lives in one period only is an edge that is already over.
  5. Can it be executed? The rule runs bar by bar as a bot would, without knowing the future. This is the gate that exposes «eyeballed» chart tests.

For each gate: the number, the threshold declared before measuring, passed or not passed. At the end the verdict: it holds, it is a coin flip, or it holds only under certain conditions, and which.

How it works

  1. You write the rule in an executable way (I help you make it so: it is often the most useful part of the whole job).
  2. I declare in writing the gates and thresholds I will use. You see them before the results, so nobody can choose after the fact.
  3. I run the measures on the broker's data (bars and ticks where needed) and write the report.
  4. A call to read it together and decide what to do with it.

Who it is for

For people about to put real money behind an idea who want to know beforehand, not afterwards. For people who bought a course or a «system» and want an independent opinion, with the numbers.

Who it is not for

For people who want to be told that their strategy works. Most of those I have measured so far do not pass: that is normal, and knowing it beforehand is the biggest saving this service can give you.

The site's public test bench does a short version of this work, free and in a queue. The service is the complete, private version, on your schedule.

All services