Yesterday was one of those days the market opened higher but then closed lower. In fact, this has been a more frequent pattern since the correction started. This price action that is caused mainly by institutional trading offers a small edge to retail traders.
The S&P 500 has fallen 13.12% from it’s all-time high on September 20, 2018, and is now below the longer-term drawdown average at 11.58%, as measured since 1960.
However, since the all-time high on September 20, 2018, the sum of changes from the close to the open of the next day (red line) in SPY ETF is positive while the sum of changes during regular trading hours (RTH) is highly negative (green line.)
Specifically, SPY has gained a total of $10.70 since its all-time high due to overnight changes but has lost a total of $47.88 during RTH, as shown in the above chart. In fact, buying the ETF on the close of a day and selling at the following open would have resulted in a new all-time high price of 302.95 versus the actual 255.08. Of course, this is theoretical because in real trading there are transaction costs and other friction effects. Below is a backtest of the overnight trading anomaly that accounts for $0.02 commission per share since the start of this year.
Year-to-date return (as of 12/28/2018) is 11.44% for the overnight trading anomaly versus -3.16% for buy and hold. Risk-adjusted return is much higher since the drawdown is much lower. Some questions emerge about this price action anomaly:
- Why is it present in the first place?
- Why has it not been arbitraged out of the market?
The first question does not have an easy answer. Causes are very hard or impossible to prove in financial markets. My suspicion is that the anomaly exists because sell program of large institutions seek liquidity during RTH and as a result cause these frequent reversals.
The main reasons this has not arbitraged out is that this is a speculative edge that carries overnight risk and probably only exploited by retail traders, not even by small quant funds that during these times try to be minimize overnight risk. The focus of funds is how to sell and secure profits, not how to realize small idiosyncratic alpha. This in turn offers profit opportunities to some retail quant traders.
The overnight trading anomaly has been discussed extensively in this blog and it is present in both ETFs and individual stocks. Here is a more recent article.
If you found this article interesting, I invite you follow this blog via any of the methods below.
If you have any questions or comments, happy to connect on Twitter: @mikeharrisNY
Charting and backtesting program: Amibroker