Friday, January 09, 1998

The Daily Workshop Report
by Robert Sheard (TMF Sheard)

LEXINGTON, KY. (Jan. 9, 1998) -- A number of readers have wondered if the Keystone model rankings of late are demonstrating a problem with the approach. For example, the stock ranked #1 in our 1998 model rankings at the end of last year, Fifth Third Bancorp <% if gsSubBrand = "aolsnapshot" then Response.Write("(Nasdaq: FITB)") else Response.Write("(Nasdaq: FITB)") end if %>, disappeared from the rankings altogether the next week.

Is the 26-week return screen too sensitive? Did Fifth Third lose its timeliness ranking? Was the company acquired? None of the above.

Let me backtrack and review how the screen works. I start with all of the U.S. stocks ranked either 1 or 2 for Timeliness by Value Line, and then focus on the largest thirty (by market cap). Those thirty are sorted by 26-week return in order to choose the stocks with the best relative strength.

Imagine now a stock that just barely made it into that group of thirty (based on market cap). It might well have the best relative strength of the thirty and be ranked high in the Keystone rankings, yet be teetering on the edge of the group itself.

If Value Line upgrades a couple of large stocks from Timeliness rankings of 3 to a ranking of 2, those stocks would join the thirty Keystone stocks and push out a couple of smaller stocks, even though nothing has changed fundamentally or technically for the smaller stocks.

That's what happened to Fifth Third last week. It dropped to #32 on the market-cap list of all stocks ranked 1 or 2 for Timeliness. That doesn't mean the stock is any less likely to perform well over the next year than it was a week earlier, but with any artificial cut-off point based on a set number of stocks, there will be some fluctuations around the fringes of the group. It just happened the first week we began following Keystone officially, that's all.

I hope that straightens out any concerns you had about stocks "disappearing."

A footnote on today's new database. For some unexplained reason, this week's Value Line data doesn't include a value for the Projected EPS Growth Rate for the bulk of our stocks. Your guess is as good as mine. Fool on!