Wednesday, June 25, 2008

A Breadth Indicator That's Suggesting A Bounce

In Gerald Appel’s book, “Technical Analysis – Power Tools For Active Investors” he discussed a breadth measure he uses to anticipate market rallies. Bascially, he takes a 10-day Exponential Moving Average of the NYSE advancing issues divided by the advancers + the decliners. In the book he discusses a system where a strong thrust upwards in this indicator frequently leads to rallies. I have found that strong downward moves also tend to lead to rallies.

According to my data provider, the 10-day Advancer EMA came in at 0.3714 on Tuesday. Below I looked at results for the S&P 500 following any time it dropped below 0.375:

These results are quite good, especially considering the averages had to absorb the max loss that occurred thanks to the Crash of ’87. You’ll notice I ran the test back to 1982. The reason being that prior to 1982 the system would have been a disaster. Look at the returns in the 70’s-1981:

While the indicator has not always worked, it has done a nice job over the last 25 years or so. It could also be used as a system parameter for a trade exit. Below I show the results of entering an index position when the indicator drops below 0.375 and then selling when it moves back above a certain number. Results here are from the 1982 – present period and are quite robust.


Dave M. said...

Rob, I think that since 2000, we have been in a 1970's to 1981 market. Yo-Yo type market. I'd like to see you re-run the test from 2000 to now.

Rob Hanna said...


From 2000 through the present the indicator has worked extremely well. No indication of the problems encountered in the 70's.


Marc said...

Ok, so we got our bounce. I must say it wasn't the most convincing bounce but there is a lot of uncertainty out there.

Where do we go from here? I'm looking for a longer term edge but scalping seems to work at the moment.


Mac said...

Hey, Rob, great work as always. I agree that we are oversold and a bounce is probably due, but this VIX is confusing me right now. Any studies comparing very oversold breadth indicators like the CBI along with low VIX numbers, relatively speaking? Thanks.

Anonymous said...

hey rob,
what data source did you use for your 10day EMA of advancers?

mojave said...

Oftentimes, strats that measure 'oversold-ness' are improved by adding a simple 'bull-market filter (for ex, 'if 100ema>200 ema, then begin'). That way, you filter out bear-market sucker buying (where oversold becomes more oversold). In this case, adding that filter worsens results for both historical periods you've chosen.

So, is the current market environment a continuation of the '82 primary bull-market, or are we entering a period similar to the '70s?

Beats me...