Tuesday, March 17, 2015

Herd Behavior In Financial Markets


 "Men, it has been well said, think in herds; it will be seen that they go mad in herds, while they only recover their senses slowly and one by one." Charles Mackay (1841)
This is the opening quote in the paper Herd Behavior in Financial Markets by Sushil Bikhchandani and Sunil Sharma published as an International Monetary Fund staff paper in 2001. Marco Cipriani and Antonio Guarino decided to take another look at this paper, published by the Federal Reserve Bank of New York, to see if its conclusions can help us to better understand the market today.

Herd Behavior Defined

First, let's define herding. Herding is when a trader disregards their own knowledge or trading plan to follow the behavior of the crowd. The reasons for the Fed's interest in the subject is clear -- to understand how to get ahead of, or put tools in place to counteract, contagion, specifically information contagion as discussed in the article Federal Reserve Bank Of New York: A Study On Contagion Theory.

The authors split the identification of "herding" from the use of data into two categories: spurious and real. Some herding, characterized by clustering in statistical data, may be the result of a public announcement rather than true herd behavior. In response to this the authors present another way to measure herd behavior through a theoretical model.

The Theoretical Herding Model

The model used to test the theory is based on an asset that is traded over a period of time. An event occurs at the beginning of each day the asset is traded. Some traders receive or find public or private information about the asset -- these are considered "informed" investors. All other traders are therefore considered to be uninformed and are therefore considered to be trading due to liquidity or re-balancing. If no event occurs, all traders are uninformed.

So how does this scenario generally play out. In a nutshell, the herd convinces the trader to put its theory over the traders own knowledge about the stock. Here's the thought process:
  1. The informed investor knows something happened to change the fundamental price of the asset.
  2. The investor realizes that the their position is the opposite of what's occurring in the market.
  3. The informed investor weighs the importance of their own private information or trading plan against the asset's movement in the market.
  4. If the market movement is deep enough the trader will go against her own plan in favor of the market. The rationale being that the information traders are trading on in the market must be better than what she knows.
In this way, herd trading is a made into a rational decision, at least in our minds.

Example: Ashland's Herd Traders

The authors use Ashland Inc. (NYSE: ASH) in 1995 to further illustrate the theory.
We find that herding on Ashland Inc. occurred quite often: on average, the proportion of herd buyers was 2 percent and that of herd sellers was 4 percent. Additionally, we find that not only did herding occur but also it was at times misdirected (that is, herd buying in a day when the asset's fundamental value declined and herd selling in a day when the asset's fundamental value increased).
The authors go on to find that "the price was 4 percent further away from its fundamental value than it would otherwise have been." This seems like a rather small percentage, but the data supports these findings and according to the VIX, contrary to perception, the market is no more volatile today than it was in 1995.
^VIX Chart
^VIX data by YCharts

So What

What are the implications of this for the Fed and for the individual investor? It's hard to make definitive conclusions about the application of this data until we have a way to measure a stock's "herd" appeal.
  • Perhaps companies with a higher degree of volume or volatility also have a higher percentage of herd traders. 
  • Perhaps this is the reason stock runs are often followed by corrections. 
  • Perhaps stocks with a high P/E have a higher degree of herd buyers? 
A "Herd Index", theoretically, would be able to provide buy and sell signals that were even more reliable than P/E multiples in finding over- or under-priced stocks. As of this writing we are unaware of any such measure.