Backtesting on the crypto tokens that still trade

Question

A researcher backtests a momentum strategy on all cryptocurrency tokens currently listed on a major exchange, using five years of history. The results look outstanding.

Why are these results probably too optimistic, and how should the test be built instead?

Accepted Answer

Today's listings are the survivors. A token is listed today only because it did not collapse, get delisted, or fade to zero over those five years. Thousands of coins launched in that window failed, were abandoned, or were removed from the exchange, and every one of them is absent from your universe. By running the backtest only on current listings, you quietly threw out the disaster cases before the test even began. That is survivorship bias, and it flatters any strategy. Why it inflates returns. Momentum and many other strategies would have bought some of those doomed tokens on the way up an…

Backtesting on the crypto tokens that still trade

Your answer

Discussion

Learn the concepts

Related questions