How To Use Multiple Timing Systems To Help Adjust Allocation To The Market To Reduce Risk And Enhance Returns

• May 31, 2001
• Mark Boucher

Over
the past year or so, TradingMarkets.com has published the daily
status of several of the more than two dozen timing systems that I follow on a
daily basis to help me decide how aggressively to position long and/or short.
The goal of any timing system is to improve the risk/reward ratio of
holding that asset on a buy-and-hold basis.

While all of the
timing systems we have published have indeed improved the risk/reward of
investing in the market over the period since we began publishing them (mainly
by being out of the market during at least some of the bear market of the past
year), the results, particularly of our combination timing system, have been
less than we would have hoped. Let us
therefore explore the problems and look at solutions — so that we can not only
continue to outperform buy and hold, but also to create a reasonable return
posture even during bear markets or trading ranges.

It should be noted
that many timing systems, even some of the best in the business (such as The
Chartist, AIQ, and MIRAT) have had unusually difficult periods over the last two
years. We’ll look at reasons for and
solutions to these problems.

Most timing systems
rely fairly heavily on breadth figures and interest-rate tools.
Independently, one can devise simple rules-of-thumb that beat buy and
hold using both breadth alone and interest rates alone.
Putting them together often creates a formidable timing system. 
But during a major top, oftentimes breadth will turn down while the major
averages keep rallying for up to 24 months. Breadth
peaked eight months prior to the high of 1929. It
peaked 14 months ahead of the peak in 1972.
Models that exited early allow investors to avoid the inevitable crash
following the dramatic peak, but they also leave them without the profits in the
run-up prior to that peak. 

Such was the case
for many timing models in 2000. In
addition, interest-rate troughs and rises in interest rates following those
troughs can lead stock market averages anywhere from weeks to up to a year.
Long-term interest rates troughed in October 1998, yet the market
averages continued to run up into March of 2000, a record divergence of nearly
17 months.  November
1999 – March 2000.  These systems
also bought prematurely, as interest rates had been dropping quite rapidly long
before any decent rally has yet to take hold of the major averages (bonds
bottomed in January 2001 and began an extremely strong rally at that point).

One of the most
frequent questions we get at TradingMarkets.com regarding timing systems is, “Since
a lot of the timing systems Mark Boucher uses have been a little off phase in
terms of market timing over the last year or so, why has his timing in his
weekly column been so darn good?”  The
brief answer to this is a concept I call “plurality.”Top
RS/EPS New Highs List or Bottom
RS New Lows List, volume, and thrust action. 

Only when the vast
majority of all of these indications give a signal, should one consider that the
plurality of timing tools is speaking strongly. In
November, 1999, most breadth models were negative, as were most interest-rate
models, and the market was overbought.  But
sentiment was not quite at extremes usually reached at a market top. 
And most importantly, what I call the internal dynamics, were still
positive — things like leadership, breakouts, and fuel stocks breaking to new
highs.

While the number of
groups participating in the rally was getting dangerously thin, the high-tech
group was still soaring, and as the leader of the bull market, until these gave
the first sign of breaking down, there could be no top. 
Finally in March 2000, the leaders took their first big hits after
sentiment had reached nearly record levels. Breakouts
to new highs in Top RS/EPS New High stocks plummeted.
Finally, almost all of the timing tools were in agreement, and at that
time I advised investors to take quick half profits on all longs and move
protective stops up very aggressively.

Thus the idea of
timing systems is to use them as gauges, as rules-of-thumb, as indications, not
as the God authority of what to do mechanically with your capital.
Don’t get me wrong, it is fine to follow a mechanical set of timing
systems with a small portion of your capital, but I think that by following the
markets oneself, one can substantially outperform a mechanical system — by
using one’s brain and keeping very closely on top of what the market is saying. 

Therefore, instead
of placing several mechanical systems before traders, we will 
present a group of indicators based on differing indications of market
action as an assistant to help traders follow and read the markets, much as I
use these models. 

However only by
looking at at least two or more systems — based on breadth, momentum, volume,
sentiment, overbought/oversold reversals, leadership, valuation, interest rates,
and a combination of these, along with the information gleaned from the Top
RS/EPS New Highs List and the Top RS/EPS New Lows List and the groups and
sub-groups that frequent these lists and breakouts from them — can an investor
really get a decent view of what the market is currently saying.

Thus, on a daily
basis, we will present traders with:

• five breadth
  based systems,

• two sentiment
  systems, two volume-based systems,

• six
  overbought/oversold reversal systems,

• three leadership
  based models,

• five
  momentum-based systems,

• two valuation
  models, and

• six combinations
  of various of these different groups of indicators. 
  

Each of these
systems has beaten the market in terms of return and risk over the last 30-plus
years, yet each is based on a different component of market analysis.
Investors may choose to mix and match these systems as they want, but the
broader selection of systems is designed to give investors much more insight
into what various schools of analysis are saying about current market action. 

Plurality is the
key. When most or all of the systems are
positive, odds are high that the market will continue to move up on an
intermediate-term basis, whereas when most systems are neutral or bearish, odds
are high that the market will move lower or in a flat trend.
Similarly, investors should watch for quick shifts in the number of
positive or negative models. 

The same goes for
negative shifts. We hope an expect that
such a broad array of models will help investors a great deal in timing the
markets in a much more accurate manner. Yet
we still advise that investors watch breakouts of 4-plus week consolidations of
stocks that nearly or fully meet fuel criteria on the upside or downside, as a
strong confirming tool to aid in determining when — and how heavily to allocate
— long or short.