Thread by @quant_arb

Lag error:

I’ve totally made up this word but there’s not word for the concept it captures and I’ve been using it for over a year regardless.

If you use bollinger bands to trade mean reverting portfolios your lag error is the loss of alpha from the deterministic component.

This comes in 3 forms:

Jump risk:
Large jumps in the mean will take time for your mean to move to and cause errors because moving averages are lagged. This is a regime shifting ish problem and is aided by unsupervised learning models with conservatism controls.

…

The next is mismatched period:

If there is a sin wave with white noise we may attempt to use an MA to trade the noise part. This will give us lag error as we will not be accounting for the broader sim function and get lag error, hurting our PnL. Mismatched timeframe

…

Finally we have trend based lag error:

If your mean reverting portfolio is trending up and you are using bollinger bands your upper bands will be more likely to get hit, BUT they have negative alpha inherently built in because you are betting against the trend.

…

This form is annoying because you are more likely to take a negative alpha trade because of the trend.

Lag error always works against you because you get out at the mean, not after it (it mean reversion and you expect it to continue is just silly).

Lag error is a form of…

Persistence or non-mean reversion on a scale you fail to account for, or your model (like with jump error) believes something that is just a huge deviation is part of a persistent trend because of its size and the mean moves.