Now there are multiple ways one can combine these edges so that the little edges/odds fit together to create a bigger edge. I think two crucial ingredients for a calendar strategy are (a) figuring the right combination of odds/edges and (b) detecting early on when the strategy edge fades (see red thick lines on chart).
Following is one simple way to combine couple of these edges. For example, from prior tests we know last part of the month has better positive edge and lower draw down. Similarly we saw returns were better in bull market regime compared to bear market regime. Similarly mean reversion days produced better returns than non mean reversion days. So how about we combine couple of these individual edges?
 |
| Day of Month performance profile - Trend & MR Regime |
Test:
Buy @ market the next day open if -
(a) Today is an MR day (and)
(b) We are in bull market i.e., Today's close greater than 200 Simple moving average.
(c) Today is near end of the month. (See Equity Curves image for top 5 days).
Sell @ market the next day open after 5 trading days.
Caveats: Results are friction-less i.e., no slippage & no commission.
Duration: 1970 - 2012 Sept.
Some notes...
The bar plot panel shows the performance profile for all days of the month. The days near the end of the month have better profile then other days of the month.
 |
| Performance Summary - Top 5 days |
Except for one, the max drawdown for these days is less than 10%. Actually one can reduce the drawdowns further by adding one additional filter we covered in prior posts. One red flag to watch in this test - performance of days 26-29 were not in top 5 days.
Finally, the above test is neither the only way nor the best way to combine the calendar odds to make the edge bigger/better. Prior posts on calendar day series provides the rules for each of the tests and lots of data for interested readers to experiment further.
Wish you all good health & good trading!
0 comments:
Post a Comment