Investor's Intelligence Sentiment Survey

Click here for chart of Bullish %  Bearish %  Bull Ratio

APPLICABLE TIME FRAME(S):  

INTERMEDIATE

UPDATE SCHEDULE:

Each Wednesday morning by 9:00 AM EST

REPORTING DELAYS:

Three days, so each Wednesday release will reflect opinion as of the previous Friday.

EXPLANATION:

The Investor's Intelligence sentiment survey is now administered by a private company called Chartcraft and is edited by Michael Burke.  The survey has been conducted weekly since 1963, and is considered the "grandaddy" of investor sentiment surveys.  The survey is constructed by polling 140 investment newsletters and determining whether the publisher is bullish, bearish or neutral.  Neutral in this case means generally bullish overall, but expecting a short-term downside correction.

While this survey is considered the most representative of general investor sentiment, in reality it is anything but.  Since the survey reviews investment newsletters, and not those who strictly invest or trade on their own, the population of the survey has a vested interest in how the market performs.  To be more specific, newsletter writers do better business when the market goes up than when it goes down.

Investors are more willing to buy a publication when it is bullish than when it is bearish.  Therefore, there may be a bullish bias inherent in this survey.  A glaring example of this is the fact that the bull ratio of the survey (defined below) recorded readings above 60% on 34% of the weeks during the 1990 bull market, when the S&P rose over 356%.

However, during the bear market which began in 2000, over 60% of the weeks recorded readings over 60%, despite a decline in the S&P of over 40%.  So even though price was falling, these publishers were consistently quick to re-establish their bullish stances, and very hesitant to become bearish as a group.  This does not necessarily make it less effective, it only means that we must consider the survey readings in the context of recent history.

Like most contrarian indicators, when the survey shows too many newsletter writers as being bullish, it very often corresponds to market highs.  Conversely, too many bears suggest that the market may soon find a low.  Lately, in particular, when we see bearish extremes in this survey, it has coincided very well with market lows (again, likely because of the inherent bullish bias of the survey population).

We follow the weekly "Bull Ratio" of this survey, which is calculated as follows:

II BULL RATIO = % BULLS / (% BULLS + % BEARS)

So, for example, if one week we have 35% bulls and 45% bears, the Bull Ratio would be 44%:

II BULL RATIO = 35% / (35% + 45%) = 35% / 80% = 44%

The point of using the Bull Ratio is that it isolates only those who express a definite opinion, and ignores the "neutral" respondents.  In this survey, "neutral" isn't really neutral - they are bullish, but expect a short-term correction, so it's hard to place them in one category or another.

We also plot the "correction" percentage in the lowest pane on the chart.  This is the percentage of newsletter writers who are essentially bullish, but somewhat on the fence shorter-term.  It can be either a positive for the market, or a negative, depending on how many others are expressing a more definite opinion, so we don't give strict bullish and bearish levels for this indicator. 

GUIDELINES:

Overall bullishness has risen during the bear market which began in 2000 compared to historical averages.  We consider this data to be generally bullish for the market when the bull ratio drops to 55% and below.  We consider it bearish for the market beginning with bull ratio readings of 67% and above.

The chart below shows two occurrences of the bull ratio reaching extremes.  After the panic selling which accompanied the re-opening of US equities markets in September 2001, the number of bullish newsletter writers shrank considerably, while the number of bearish publishers rose.  This pushed the bull ratio more than 2 standard deviations below its bear market mean, suggesting that the bearishness had reached an unsustainable extreme and we were likely about to put in a major low.  That did indeed occur, and we rallied consistently into the new year.  However, by early January the newsletter writers were back to their old habits, and once again recommending that their subscribers be heavily invested in equities.  This was inopportune, as the market peaked almost exactly at the time the bull ratio became extreme.

Although this is a real example and points out the value of following this information, we do not mean to intimate that the market ALWAYS peaks when the bull ratio becomes "too high", or troughs immediately after the bull ratio drops to 55% or under.  It is a guideline and not a trading system unto itself.

For purposes of the Indicator page, only the II bull ratio is used.  The thresholds used for the check marks are below:

STATS:

*Standard Deviation.  See below for a description of standard deviation for the bull ratio:

68% of readings (1 standard deviation) should be between 55% and 67%

95% of readings (2 standard deviations) should be between 49% and 73%

99% of readings (3 standard deviations) should be between 43% and 79%

In other words, we should expect a reading under 49% or over 73% approximately 3 times per year.  Since such a reading would be relatively unusual, it suggests that we may be seeing an unsustainable trend.  These figures assume a normal distribution curve.

ADDITIONAL RESOURCES:

Investor's Intelligence (www.investorsintelligence.com)

© 2005  Sundial Capital Research, Inc.  All Rights Reserved.