In this case I'd expect the costs of trading in and out of the position everyday to be prohibitive. Even if your broker doesn't charge you anything (or you are

In this case I'd expect the costs of trading in and out of the position everyday to be prohibitive. Even if your broker doesn't charge you anything (or you are a prop trading desk with zero trading fees), just paying the spread every day would add up quickly over time. It might not be possible to perform such an experiment at all in the real world. It's still interesting from an academic standpoint though, if only to show how long term pricing inefficiencies can persist over long time periods even in otherwise efficient markets.