Increasing trade minimum holding period reduces trades by about a factor of two in a daily long/short strategy and as a result the transaction cost. In this article we look at the impact on performance.
We consider the fixed rules strategy described in detail in this article. In the original strategy there was no minimum holding period imposed and as a result more frequent rebalancing was in effect. Although the closed trade holding period varies from 1 day to 58 days, nearly half of the trades were open for one day and the average holding period was 3.6 days.
If we increase the trade minimum holding period to two days the average trade holding period increases to 5.77, or by about a factor of 1.6. There is a significant gain from transaction cost due to this if the performance is not affected much.
Below is a table that compares the performance of the fixed rules strategy for trade minimum holding period (TMHP)) of 1 and 2 days in the period 01/02/2008 to 04/05/2019.
|TMHP = 2
|B&H (SPY TR)
No commissions are included in the results. These types of strategies are suitable only for those with direct access to a prime broker that offers low cost execution and liquidity search algos that minimize slippage. These are not strategies for retail traders because they try to exploit small idiosyncratic edges. For all practical purposes adding friction effects will not affect maximum drawdown much but will impact CAGR by about 200 basis points on the average. The objective here is to discuss the concept and not to provide final strategy details. There are a number of significant steps to convert the idea to an automated strategy that often involve hundreds or even thousands of lines of code depending on specifications.
It may be seen from the above table that the effect on performance is minimal with TMHP equal to 2 while the number of trades is reduced by a factor of about 1.85. This is a significant gain. Below is the equity curve for TMHP =2.
For more details about the strategy see the article Low Risk Equity Long/short with Features Based on Price Anomaly Detection. This article and related software DLPAL LS have generated significant interest in the professional trader and hedge fund community because of the potential of extracting high risk-adjusted returns from idiosyncratic factors with better prospects than using alternative data such as sentiment indicators.
If you have any questions or comments, happy to connect on Twitter: @priceactionlab