It seems you have JavaScript disabled.

Ummm.. Yeah... I'm going to have to ask you to turn Javascript back on... Yeah... Thanks.

A Brief History Of Contrarian Analysis at Trader’s Narrative

A Brief History Of Contrarian Analysis

Charles H Dow picture smallGranville said it best in his book, A Strategy of Daily Stock Market Timing:

When it’s obvious to the public, it’s obviously wrong.

Since we talk a lot about sentiment and contrarian sentiment, lets step back and review where this idea came from and how it developed from its origins.

Charles H. Dow
The main principle behind contrarian analysis and sentiment (two sides of the same coin) comes from Charles H. Dow’s work on distribution and accumulation. The same ideas that underpin the Dow Theory. I’m sure you’ll also notice the similarity between these ideas and Weinstein’s stage analysis which breaks up a movement of a security into four parts.

According to Dow Theory, major market movements start with an “accumulation” phase where insiders, and other knowledgeable traders or investors start to buy shares. Since at this point the average public sentiment towards the market is negative, they are able to accumulate shares without significantly pushing prices higher.

Eventually the general sentiment starts to tip as more and more people start to realize that something has changed. This is the stage at which trend followers jump on and start to push up prices further. The trend continues and feeds on itself, perpetuating until it reaches a crescendo.

At this point we reach “distribution”, the final phase of the trend where the reverse happens: insiders, institutions, or if you will, “smart” market participants begin to sell their holdings into a frenzy of indiscriminate public buying. Since a smaller number of players are in the know, their holdings must need be several magnitudes higher than the average retail participant.

This is why we see lopsided sentiment metrics. Since for every trade to occur, we need to have an equal number of shares bought and sold, if the minority are selling, then they must have the ability to supply the demand of the many who are buying (in a distribution phase). If we imagine, for instance, that 90% are bullish, then the average seller must be selling 10 times as large as the average buyer.

Garfield Albee Drew
In the 1940’s, Drew started to gather and study trading statistics from retail brokerage accounts and noticed that small traders or “odd lot” traders tended to sell when the market was bottoming and buy when it was topping. So he started to track odd lot trades on the NYSE and this now familiar metric was born.

I’ve mentioned this sentiment measure a few times before (Climbing the Wall of Worry). There is also the flip side: odd lot short sales ratio. But I suspect that the change in the market structure has eroded the usefulness of odd lot data. When Drew did his studies, odd lot volume was 15% of the NYSE, now it is less than 1%.

Drew garnered attention when he published “New Methods for Profit in the Stock Market” and later started an institutional service (for $95 a year back in the 1960’s) gaining thousands of clients.

Humphrey B. Neil

In 1954, Neil was arguably the first to introduce the concept of contrarian sentiment in his book: The Art of Contrary Thinking. Unfortunately, he didn’t really explain exactly what he meant, other than just doing the opposite of what others are doing.

Neither did he provide any quantitative methods for measuring sentiment to be able to not only put the ideas to the test, but to also come up with a framework that others could follow.

A. W. Cohen
investor's intelligenceThe task of quantification began in 1963 when Cohen started to compile statistics on a number of market newsletters to aggregate their recommendations. It was Cohen who laid the groundwork for moving sentiment and contrarian analysis from vague generalities to hard numbers and metrics. He established a famous sentiment measure that is now known as Investor’s Ingelligence (by ChartCraft) - along with the AAII, the most watched weekly sentiment data.

Cohen began to compile the sentiment data monthly in January 1963. A year later it was measured twice a month and in 1969 it changed to the now familiar weekly frequency. Cohen’s work is now carried on by Michael Burke. Cohen, you may also remember, was the major force behind the popularization of point and figure charting (which has nebulous origins somewhere in the early 1900’s).

R. Earl Hadady
earl hadady contrary opinion.jpgHadady refined much of the previous work already mentioned, as well as that of his one time partner, J. H. Sibbet - whose most important contribution was weighing each newsletter according to its reach and audience. Hadady delineated methods for both quantitative and qualitative measure of contrary sentiment in his book. He is also the developer of a sentiment measure you’re probably familiar with: Bullish Consensus (now provided by Market Vane).

Although Bullish Consensus is known for its weekly sentiment data on the US equity market, they also track 36 commodity futures markets. Hadady has written other books (both on the market and other subjects) but “Contrary Opinion” remains his masterpiece.

Before becoming Market Vane, Hadady Corp. used to publish charts which plotted sentiment below the major market index. The chart for the S&P 500 Index for 1987 is a great example: on August 25th 1987, Bullish Consensus reached 70% - the critical optimistic level for the first time in the year. On October 20th 1987, Bullish Consensus fell to the critical pessimism level of 25%. Between those two dates, the market provided one of the blackest swans we have ever seen.

The shocking volatility of the 1987 market crash lead Hadady to conclude that weekly numbers were not enough so in late 1988 his company started to compile and disseminate daily Bullish Consensus data.

Enjoyed this? Don't miss the next one, grab the feed  or 

                               subscribe through email:  

4 Responses to “A Brief History Of Contrarian Analysis”  

  1. 1 mak


  2. 2 Russ Abbott

    You quote Dow as saying that “If … 90% are bullish, then the average seller must be selling 10 times as large as the average buyer.” This marks the “distribution” stage. Complementary statistics mark the “accumulation” stage. This suggests that we should look for statistics about how many distinct buyers and sellers there are over a range of trading days. If there are more distinct buyers, we are in distribution; if more distinct sellers accumulation. Are any such statistics available?

    A second point, though, is that this assumes that the large buyers/sellers are right and that the small buyers/sellers are wrong. Do we have statistics to back that up?

  3. 3 ales


  4. 4 Babak

    Russ, I wasn’t quoting Dow, I was simply using an example to illustrate the point. I’m not sure if it is even possible to keep track of such data as you request. There are however, different ways of approaching it. Lowry for example, has their proprietary buying power and selling pressure index.

    re your second point, one of the reasons the ‘insiders’ are larger than the retail participants is that they know more, have more resources and are usually more right than wrong. Otherwise, their capital would soon dwindle to the size of retailers. Having said that, they aren’t always right.

    We can see that clearly today when firms like Goldman Sachs goes panhandling to Washington like a hobo. Of course, you could also argue that the fact that these firms continue to exist and even manage to come out of such a crisis shows that they are truly ‘insiders’ with not only capital but also connections that you and I can only dream about!

Leave a Reply