Subject: Re: funds mech?
Hi musselmant,

I think you were answering my question from a different thread so I only saw this now:

Just as an example, using Zacks backtester, the most recent 1 year ROE top 10% of zacks ranked stocks has a monthly traded CAGR since the end of 2008 to 3/20/2026 of 12% (16% if all zacks rank 1 stocks with 1 year ROE higher than their industry). The last 5 years ROE average instead is 15% (17.1% for zacks rank 1 stocks with 5 year ave ROE higher than their industry), and the top 10% by change in ROE of the two is 18%. If you buy those with zacks rank 1 with a ROE change higher than their industry it is 17.3%

Thanks. Just to confirm, when you refer to the change in ROE, I assume you are ranking the 10% by (Most recent 1 year ROE - The last 5 years Average ROE).

If so, do Zacks report negative ROE values or are they excluded by default? Just that in theory both could be negative and still be included in the greatest change top ranks.

Anyway, your test would support (1) the idea that a multi-year average is probably better than a single year and (2) an improving ROE is better again.