Feb 21, 2004...in the financial press, you'll hear writers speak of market "leaders".  These are the stocks that first make their upward moves, with the other, wimpier stocks following later.  Presumably, market leaders can also lead the pack downward, though we usually speak of a "leader" with admiration...it's a good thing to own a leader in a positive market.

Finding a "leader" is usually a somewhat mystical process.  It's not horribly difficult to find ways to quantify a leader (just write a program that identifies stocks whose upward moves tends to predict future upward moves in the general market), but the statistics aren't easily available.

Actually, though, we'd guess that stocks that are referred to as "leaders" are simply volatile stocks.  If you have good reason to believe that the market will prosper in the weeks ahead, would you rather own a volatile software stock or a dowdy floor tile manufacturer?  You grab the volatile stocks as quick as you can, knowing that they're likely to prosper above and beyond the general market.

Let's say that you're not particularly sure about the general market, but you do have reason to believe that the group of floor tile manufacturers will be a strong group in the weeks ahead.  What do you do?  You grab the most volatile floor tile company you can find...it should outperform the other manufacturers.

It seems that a fair number of technical guys look for these leaders to make further gains to confirm a bullish view of the market.  It might be more concise to simply say, "the market tends to go up when volatile stocks are the best performing group".  As a starter to this project, we've noted that on occasions where the most volatile stocks have topped our "10-slice" daily tables in January, the market has gone up 12 out of 13 times in the following session...

Feb 23, 2004...To continue the discussion we started on the 21st, we found that the market rose 98 out of 157 times (62.4%) when the top 10% of stocks in our "stdev", "pstdev", or "vstdev" categories was the best performing group in the previous day.  This compares to a slight general tendency for the market to finish up (51.6% of the time since 1992).

Feb 24, 2004...continuing some of the discussion above, we asked which indicators in our daily tables best predict positive or negative sessions in the next market session.  The following tables show the market gains that have followed when the given indicator is the top predictor of today's gains...

A few indicators actually showed more radical average gains, but they didn't appear often enough at the top of our tables to justify consideration.  

All the indicators in the table appeared at least 30 times.  The "high-close" indicator was the single most frequent...it appears 306 times at the top of our daily tables.  In the end, it does not serve as a strong indicator of gains or losses for the next market.

The average gain for all indicators was about 0.077%/day.  

Regarding our previous discussions on highly-volatile stocks leading the market to further gains, it would seem more to the point to say that when non-volatile stocks appear on the top of our winning stocks tables, they tend to signal losses for the general market in the session ahead.

Feb 25, 2004...below, we repeat the same sort of experiment as on the 24th, but this time we ask how the market reacts if particular groups appear atop our "worst performing stocks" tables.  The table shows the most interesting results we could dig up.  To be clear...if the group shown in the table was the worst performing group in the market, the following average gains ensued in the general market in the next session.

Note:  you can find more discussion relevant to market predictors here.

March 24...We've added an indicator that attempts to identify leaders.  We call the indicator "leadership".  The idea is if a leader goes up/down on day 1, other stocks should tend to go up/down on day 2...if this behavior plays out, you give a positive score (the absolute value of the total gain or loss that was predicted) to the stock in question.  If not, a negative score.  You look at this sort of behavior over a certain period of time with a large number of "reference" stocks (those for which the prospective "leader" is trying to make a prediction) and output a number.

The number we output is adjusted for volatility.  In other words, a non-volatile stock that does a good job of making predictions gets a higher score than a volatile stock that also does a good job.  In general, volatile stocks tend to do a good job of predicting market moves...we want to see if we can identify "leaders" regardless of volatility.

Obviously, there are a lot of variations on this indicator.  You could use it to focus on prediction of stocks within the leader's industry only.  You could experiment with different time frames.  You can de-adjust for volatility.  For the moment, we'll leave the indicator as it is.  If it finds its way into our tables with any frequency, we'll consider tinkering with it more.

Note that a negative "leadership" score doesn't necessarily mean a stock is a "follower".  It means that if the stock goes down, the other stocks tend to go up.  This, in a sense, is also a sort of "leadership".  Stocks that find themselves in the mid-range of our "leadership" scores are simply those that don't seem to predict tomorrow's market with any "force". 

So, what stocks are currently good "leaders" when looking at a prediction of tomorrow's market (as opposed to the direction of the market in the next month or yet another time frame)?  We've got "BE Aerospace" at the top of the list, followed by "Ixys Corp" and "Abercrombie and Fitch".  Microsoft is a weak (middle of the pack) leader, but Intel is rather strong.  Of the big name stocks, SBC Communications is probably the best leader.  Coca Cola gets a big negative rating...a strong performance tends to signal general losses in the next market.

Another possibility for the future would be to make a "market leaders" index of sorts and test whether overall gains in the group of strongly positive leaders or losses in the group of strongly negative leaders would tend to do a decent job of predicting tomorrow's market or not. 

Having said all this, we have no idea whether the leadership indicator has any worth at all.  We'll make a prediction regarding the viability of the indicator:  the indicator will indeed find its way into our tables with good frequency, but the results will be often be counterintuitive:  strong performance in positive leaders will occasionally signal a loss in the upcoming market.  It's as if the market is saying, "I'm sick of this paradigm...it's time to do a flip-flop".  We'll see.

Copyright © 2010 MarketSynopsis.com. All rights reserved.