Frequency Effects on Predictability of Stock Returns

Abstract

We propose that predictability is a prerequisite for profitability on financial markets. We look at ways to measure predictability of price changes using information theoretic approach and employ them on all historical data available for NYSE 100 stocks. This allows us to determine whether frequency of sampling price changes affects the predictability of those. We also relations between price changes predictability and the deviation of the price formation processes from iid as well as the stock's sector. We also briefly comment on the complicated relationship between predictability of price changes and the profitability of algorithmic trading.