Why Inverting a Losing Trading Model Needs Fresh Data

Summary

The document asks whether a trader can reverse every signal from a model that performed poorly in a backtest. Its central lesson is that a failed test of the original strategy does not provide valid evidence for the opposite strategy. Choosing to trade the inverse after seeing the backtest results is another data-driven strategy selection, and the same sample has already influenced that choice.

The proposed sounder approach is to state the inverse strategy as a new hypothesis and evaluate it on new, independent market data. The discussion offers statistical reasoning rather than performance results or a worked example. It does not specify how much new data is sufficient, or address practical matters such as transaction costs, changing market conditions, or risk controls. The inverse may be worth investigating, but the original backtest alone cannot establish that it will work out of sample.

Key ideas

• Rejecting the original strategy does not establish that its inverse is profitable.
• Selecting a direction after inspecting backtest results can undermine the validity of the statistical test.
• Treat the inverse as a new hypothesis and evaluate it on data not used in the original backtest.
• The document provides a testing principle, not evidence that any particular inverse strategy succeeds.

Tags

# Doing opposite of what the model says


# Doing opposite of what the model says












Is it considered a viable trading strategy to do the opposite of a consistently losing model?

That is, whenever the model says short, you go long, and vice versa.

Disclaimer: I would never do this. I am just interested in the opinion of other members of this community.

## Answer by Alex C (score 15)

https://quant.stackexchange.com/a/35906

If you do this, you would destroy the value of the statistical tests that you performed on the backtest. You had a hypothesis that the strategy would make money, but the hypothesis was rejected. You cannot say "I will accept the hypothesis that the opposite strategy is successful"; no statistician would agree with this conclusion. In that case, you might as well test strategies at random and trade them in whatever direction (direct or opposite) they seem to work, but it would be an unsound procedure from a statistical point of view, with low chance of success in out of sample data. Some people do that, but I would not consider it valid statistical trading.

What you could do is formulate a new hypothesis, but that hypothesis cannot be accepted yet. It would need to be tested ON NEW DATA, not the data that you used in the backtest. Perhaps, you would monitor the market for a while and see how the opposite strategy does. At some point, you may have enough data to conclude that the opposite strategy works and you can trade it for real.

Shown in full with attribution under the source's licence. Licence: CC BY-SA 4.0 (Stack Exchange)

This summary was written by Stratmill's research agent from the original; it is not a copy of the source.