It is a market of extremes, and this usually means emotions are extreme as well. So be careful what you read and believe as not all market "calls" are supported by the data.

A well known CNBC commentator/ personality and online columnist whose last name sounds like but is spelled differently than Jerry Seinfeld's neighbor in the TV show "Seinfeld" had this to say about the percentage of bulls in the Investors Intelligence data: "Key Indicator Flashes Buy".

Wrong!!!

The current percentage of bulls in the Investors Intelligence data is 36%, and this "low-ish" number indicates that there aren't too many bulls around. Contrarians would view this as bullish, and I would tend to agree with this assessment as the best and most accelerated gains typically occur when the majority of investors are on the sidelines.

However, as a single data point leading to a buying opportunity, the percentage of bulls from the Investors Intelligence data has a very mixed record at calling a market bottom. The effectiveness of this "key indicator" really depends upon the decade in which you believed the indicator to work. So in the 1970's a low percentage of bulls actually was a bad sign of a market bottom. In the 1980's and 1990"s, this indicator seems to work better at identifying a market bottom and buying opportunity.

Let's design a study where we buy the S&P500 when the percentage of bulls in the Investors Intelligence data is less than or equal to 36%. I looked at performance 4, 8, and 12 weeks after the original buy signal, but because the results are not materially different, I will only discuss the strategy where positions are exited after 12 weeks.

The equity curve for this strategy (i.e., buy when the percentage of bulls is less than or equal to 36% and sell 12 weeks later) is shown in figure 1 (next page). Since 1971, such a strategy generated 832 S&P500; buy and hold yielded 1250 points. There were 67 trades of which 63% were profitable. With the strategy, your time in the market was 41%. But looking at the equity curve, we note that the strategy really was not effective for the first 35 trades (x - axis of graph). This corresponds to the period from 1971 to 1983. After this period and coinciding with the subsequent 20 year bull market, this was the perfect strategy, and in fact, since 1995, the equity curve has gone parabolic.

Figure 1. Equity Curve

So is 36% bulls a buy signal? It isn't if you lived and traded in the 1970's, and looking at the individual trades from this time period would suggest that it was actually a very poor indicator of a market bottom and buying opportunity. The MAE graph for this strategy is shown in figure 2.

Figure 2. MAE Graph

As a refresher, MAE stands for Maximum Adverse Excursion, and this graph shows every trade from the strategy. What does the MAE measure? If you are like most traders, you put on a trade, and then you watch that position move adverse to your entry point as very few individuals always pick the exact bottom tick. MAE is how much in percentage terms the trade loses before it turns around to become a winner or is closed out as a loser.

For example, the trade in the MAE graph with the blue box around it had a MAE of almost 9% (x- axis); this trade did recover, and it was closed out for a 10% winner (y- axis). We know it was a winner because it is a green caret.

So what does this MAE graph tell us? Out of the 67 trades from this strategy, there were 14 trades that lost over 8% before being closed out, and these are the ones to the right of the blue vertical line. Of the 14 trades, 11 were from the1971 to 1983 time period. The other three dates are notable: 1) October, 1987 (market crash); 2) August, 1990 (recession); and 3) July, 2002 (bear market plus recession).

Several other take away points. Remember, each trade from this strategy is only in the market for 12 weeks at a time, so these are fairly significant draw downs (especially after a supposed "buy" signal). Furthermore, it is likely that a draw down in excess of 8% from the time of the signal would imply even further downside and a bad or failed signal. 12 out of the 14 trades that experienced excessive draw downs were closed out for losses, and 10 of these losses exceeded 10% from the time of the signal.

So should we believe that well known, highly vocal and emotional analyst? As far as believing what's his name, this is an instance of a trader placing more emphasis on events that have occurred recently. Yes, this key indicator has marked an important inflection point in recent pasts, but not all the time over the past 37 years. Now we all know that no one indicator works all the time, but that isn't the case here. This indicator (and data) is clearly decade dependent.

More importantly, will our signal continue its recent track record, or will it revert to the failures of the 1970's? This is probably the most difficult question to ascertain. Markets change, and history is littered with indicators that used to work and now are in the waste bin of technical and fundamental analysis. Furthermore, they don't ring bell to tell you "hey this doesn't work anymore!!"

At 36% is the current percentage of bulls in the Investors Intelligence data a buy signal? My belief is that this is only one data point that sheds light on the current market environment. It has its limitations (i.e., the 1970's) that our high profile CNBC does not take in to account or seem to recognize.

However, the data point still has value because if we are willing to bet against the "herd" of investors who are typically wrong, then we must understand and accept the risks inherent in this approach. Like most market indicators, this one is not perfect, and as we can see from the above analysis, its failures can be rather spectacular. Once again, it is my belief that we should be in the market during these times, but we should have a strategy (i.e., money management scheme) that reduces the risk. Remember, they won't ring a bell to tell you "this doesn't work anymore!!"

To learn more about our quantitative and disciplined investment approach please visit www.thetechnicaltake.com and sign up for our free weekly newsletter and downloads.