No. of Recommendations: 14
Bill2m has been posting the the results of weekly trading MI screens for a number of years. Quite an accomplishment.

I wanted to see how the best performing screens have done, and the worst. The test used the Trailing Twelve Month results for consistency, and looked at 5-stock monthly trading. Only long screens were considered and only those whose results could be found at gritton's backtester.

Here are the results for 2003 through 2011:

Best 5 Worst 5 SPY
2003 24% 58% 29%
2004 14 33 9
2005 44 14 3
2006 12 20 14
2007 15 - 4 5
2008 -62 -54 -38
2009 31 49 32
2010 31 12 15
2011 -23 - 4 2
.
CAGR 3% 8% 6%

DB2
Print the post Back To Top
No. of Recommendations: 0
If I am understanding exactly what you did. First at the end of each year you found the best 5 and worst 5 ranked by TTM CAGR. Next you found the average performance for the selected best and worst performers over the next 12 months. If that is the case there appears to be a mean reversion with respect to a 12 month look back.

If this is what you did why did you need gritton's backtester? Isn't all the data you need in Bill2m's downloadable spread sheet?

RAM
Print the post Back To Top
No. of Recommendations: 5
First at the end of each year you found the best 5 and worst 5 ranked by TTM CAGR. Next you found the average performance for the selected best and worst performers over the next 12 months. If that is the case there appears to be a mean reversion with respect to a 12 month look back.

My motivating question was whether it is worthwhile to select screens from the top of the weekly ledger listing.

Gritton's backtester has monthly screen granularity, and it shows the beginning date of each cycle. I used Bill's weekly ledger results for the same date (or the nearest earlier one) the find out which screens had the best and worst performance over the previous 12 months. The backtester then showed me the returns for monthly cycle.

This was basically a test of What's Working Lately and What's not Working Lately. Using a 12-month lookback does not appear to add value; perhaps a shorter period would.

DB2
Print the post Back To Top
No. of Recommendations: 3
My What's Working Lately endeavors would make a nice chapter in Fooled by Randomness. For several years I ran a genetic optimizer optimizing on different look back periods, Sharp . . . and everything else, similar to Zee's postings around '05. For a while 4 to 9 month CAGR to GSD performance looked promising. Lately I simply been choosing screens each month with the best Long term CAGR / (GSD ^1.6). This along with an independent consensus of market timing signals seems to give decent performance with a livable ulcer index. (although not up to Zee's finesse)!

RAM
Print the post Back To Top