A museum of dead trading strategies
Most strategies die.
We show you how.
v0mma tests futures strategies until they break: textbook setups, viral claims, published papers and our own ideas. Every backtest is recorded with its data, its costs and its warning flags, and almost all of them end up here as a lesson.
Causes of death
Eleven ways a strategy dies
A backtest that makes money is the start of a question, not the answer. These are the reasons the answer usually turns out to be no.
The best of many is lucky
Test enough rules and one will look brilliant. The question is how brilliant the best of pure noise looks.
CostsThe edge is the toll
A gross edge the size of one round turn is not an edge. Tight stops make costs the biggest number in the trade.
Look-ahead biasTrading on tomorrow's news
One line of code that reads a finished candle at its open is enough to turn noise into a fortune.
PlaceboAny level would do
The same rule at random nearby prices or other times of day makes the same money. The special level is decoration.
Out-of-sampleIt only worked in one half
A rule that works in 2015-20 and not in 2021-26 is a story about 2015-20.
DecayPublished, then gone
Real anomalies get traded away. The paper's years are not your years.
Execution artifactProfit made of fills
How the simulator fills an order in the entry minute can create or erase an edge on its own.
Win-rate illusionA win rate you get for free
Win rate is mostly a property of the exit. Random entries reach the same number.
No edgeFamous, and empty
Classic setups and popular order-flow reads, measured carefully, carry nothing.
Filter backfiresThe filter selects against you
A filter can pick exactly the days a setup fails, and flipping a loser rarely makes a winner.
HindsightPerfect in hindsight
The prize is easy to measure afterwards. Forecasting it in real time is the whole problem.
All 25Walk the Graveyard
Every grave has the rule, the test, the cause of death, the lesson and the run behind every number.
From the Graveyard
Three graves worth visiting
Academy
Ten chapters on honest testing
R-multiples, the cost fraction, random-entry controls, placebo levels, selection bias, walk-forward, look-ahead traps, fat tails and prop-firm math.
Lab
Break a strategy yourself
Search ten thousand random rules on data with no edge and watch the winner fail. Then see why your exit, not your entry, sets your win rate.
Ask Leo
An AI that thinks like a skeptic
Leo knows the graveyard and the academy. Bring a strategy idea and it will tell you how to try to kill it.
Education, not advice. Nothing on v0mma is investment advice or a recommendation to trade. Futures trading involves substantial risk of loss. Every result shown is a hypothetical backtest; see the full disclaimer below.