Too Many Bulls or Not?

Bottom Line: If we adjust the Investor’s Intelligence bullish percentage for how well the market has performed and where interest rates are, things don’t look as overdone as if we took the numbers at face value.

In an intraday note this morning, I had a quick blurb about the Investor’s Intelligence sentiment survey showing the most bullish respondents in nearly 18 years. That data point, and others like it, is getting a lot of “play” as it gets passed around from trader to trader. On the surface, it’s a rather stunning statistic, but I have been asked (rightly so) if it is more a simple reflection of how well the market is doing as opposed to a display of excessive optimism.

The Investor’s Intelligence survey is made up of people who write newsletters. They can say they are bullish for any number of reasons, not the least of which is the idea that newsletters get more business if they are bullish than if they are bearish. Whether they end up being right or not is secondary – typically a newsletter can get more subscribers if they paint a bullish picture. However, certainly the performance of the market is a key ingredient when they determine their forecasts, and it is likely that interest rates are a part of that picture as well.

To try to weed out the effects of the market and interest rates, and to try to determine if the survey is showing excessive bullishness or not, we can look at the long history of the survey through various types of markets and rate cycles and look at how bullish this population was throughout. To do this, we can use something called multiple regression, which is a statistical concept that allows us to input one or more factors (e.g. market performance and interest rates) to determine what impact they have on another factor (e.g. bullish opinion).

Since 1977, there is a very clear, direct correlation between how the market has performed and bullish opinion, and an indirect correlation between interest rates and bullish opinion. That is, when the market does well, we tend to see high bullish opinion, and when interest rates are low, we also see high bullish opinion. Those are not surprises to be sure, but we can quantify this behavior and use it to determine what impact they are having now.

As an example, as of the last Investor’s Intelligence reading, I show that the S&P was about 5.6% above its 30-week moving average, and the yield on 30-year Treasury Bonds was 4.82%. Using those two inputs, historically we should expect 50.7% of the Investor’s Intelligence population to say they are bullish.

Let’s change the two inputs a bit to see just what impact they have on the bullish percentage. First, we’ll keep the S&P’s performance about where it is now, but we’ll change the Bond yield to see how many bulls we could expect.

If the S&P is…	And the Bond yield is…	Then we should expect this amount of bulls…
5% above its 30-week average	1.0%	64%
5% above its 30-week average	12.0%	40%

Now, we’ll keep the Bond yield about where it is now, but change the S&P’s level compared to its 30-week moving average.

If the S&P is…	And the Bond yield is…	Then we should expect this amount of bulls…
20% above its 30-week average	4.8%	59%
20% below its 30-week average	4.8%	36%

From the tables above we can clearly see that both market performance and bond yields have a large impact on the percentage of bulls in this survey. Using the current inputs, as I said above, we could reasonably expect to see Investor’s Intelligence to report that about 51% of its respondents were bullish. Instead, they have reported that over 62% are bullish, so around 11% more respondents are bullish than we should be seeing, given the performance of the S&P and level of interest rates.

Going back 18 years to 1987, the last time we saw such a high level of bulls, the S&P was 10% above its 30-week moving average, and long-term rates were at 7.4%. Given those inputs, we should have expected to see 49% of the survey respondents call themselves bullish, but instead we saw over 64% do so. In 1987, then, the excess bullish sentiment was 15% (i.e. 64% actual minus 49% expected), which is higher than the 11% we’re seeing now.

The chart below plots this “excess” bullish percentage over the past few years.

We can see a couple of standout periods. On October 11, 2002, the S&P was 15% below its 30-week average and rates were nearly exactly where they are now. Given that, we should have expected there to be about 39% bulls in the I.I. survey. Instead, there was only 28%, so we were actually seeing about 11% fewer bulls than we should have. Even in August of this year, we should have seen the survey report about 46% bulls but we only saw 40%, so again we were seeing this population be more negative than they had been historically given similar market scenarios.

With the latest data, there are about 11% more bulls than there “should” be given the inputs. This is not truly extreme, as the difference has been higher, but it is high. The table below shows how the S&P has performed going forward when we have seen both excessive pessimism (meaning the bullish percentage is at least 10% below where we would expect it to be) and excessive optimism (meaning the bullish percentage is at least 10% above what we should expect).

S&P 500 Performance After Excess Pessimism and Optimism 1977 - 2004
	4 wks later	8 wks later	16 wks later	26 wks later	52 wks later
When bulls are 10% or more below where they should be…
Avg Ret	0.6%	1.6%	4.7%	8.5%	18.3%
% Pos	64%	65%	77%	85%	93%
When bulls are 10% or more above where they should be…
Avg Ret	0.2%	0.1%	0.0%	1.8%	8.2%
% Pos	51%	53%	42%	54%	67%

It’s no real surprise that the market does better after extreme pessimism than after extreme optimism. The greatest disparity was after 16 weeks, where the S&P was higher 77% of the time after pessimism but only 42% of the time after optimism.

Looked at another way, over the following 16 weeks, when the difference was -10% or less (meaning excess pessimism), the average maximum gain was 7.4% while the average maximum drawdown was only 2.5%. However, when the difference was +10% or greater (meaning excess optimism), then the average maximum gain was 5.2% and the average maximum drawdown was -5.4% - not exactly a barn-burner of a risk/reward ratio from the long side.

While the “highest bullish percentage in 18 years” fact gets all the headlines, it doesn’t tell the complete story. The S&P has done well lately, and bond yields are still low, so we should expect to see a high number of bulls. Still, we shouldn’t expect them to be this high. The difference between what we are seeing and what we should be seeing is enough to be of concern from a risk/reward perspective, as the risks just about balance out.

In an intraday note this morning, I mentioned that Rydex traders were finally showing another spark of speculation, as they put money into the leveraged bullish Nasdaq 100 fund and took a good chunk of money out of the leveraged bearish Nasdaq 100 fund, despite the index being down on the day. If we look at the transfer of assets as a percentage of total assets in the two funds, we can see that there were 39 other occurrences in the past four years that equated to the current one. Ten days later, the NDX was lower 72% of the time with an average return of -2.7% and thirty days later it was also lower 72% of the time but the average return dropped to a large -8.0%. So whether we look at the transfer in absolute or relative dollars, the precedent of these traders trying to “buy the dip” suggests that it tended to not be successful for them.

My opinion has been that if we are going to see any type of tradeable decline, it is either going to come prior to December 23^rd or after the first week of the New Year. For shorts, the best chance for success will likely come if we have a mild up day on Friday. Expiration days have a very slight upward bias, but the following day(s) have been clearly negative. Out of the 11 option expirations so far this year, the Monday following them has been negative 8 times. Since the beginning of 2003, those Mondays have been negative 16 out of 23 times. That’s consistent enough for me to sit up and pay attention. I continue to prefer looking on the short side of the market for opportunities for the time being, but that will almost certainly stop by the middle of next week.

This disclosure is not intended as trading advice in any form. It is meant as a note to subscribers that the author may have a position directly affected by the market outlook reflected in the commentary. Although the author takes great pains to remain objective in any commentaries, it is only fair that readers should know that the author may have taken positions in accordance with his market outlook. Positions can and do change at any time, without notice to the reader.