Market Measures - August 20, 2021 - The Fallibility of Statistical Studies

published 1 month ago by tastytrade, Inc.

Historical backtests of strategies tell you solely what happened over that specific time period, and they should not be used to conclude the strategy’s future viability. The degree of the “statistical significance” of a claim depends on mainly two things:How likely is it that results are influenced by random noise compared to a known reason (correlation vs causation)?Would accepting the claim contradict previously proven claims?In this case, is GOOGL’s risk/reward pricing wrong, or was the GOOGL study just affected by a lot of variance over that timeframe? Both cannot be true.

more episodes from The tastytrade network