One of the reasons why trend-following works in markets is because of investor herding. If fact, one reason why markets are not always efficient is that fact that herding behavior exists in markets. It is not unusual and is a response to uncertainty.
Recent research has tried to define and focus on why herding occurs. See "Estimating a structural model of herd behavior in financial markets" in the AER January 2014 by Marco Cipriani and Antonio Guarino.
A good simple definition for herding is that it occurs when there is decision clustering. Everyone is making the same decision at the same time. However, there is complexity caused in this definition because there are two types of decision clustering. Clustering can occur when there is a common reaction to some new information. For example, the Fed makes an announcement and investors decide to change their bond allocations. This cluster can be considered spurious herding. Actions are being taken simultaneously by many but it is a perfectly rational response to new information. The herding that is less rational would be when an investor is willing to follow the crowd even though their private information is in disagreement.
Herding is more likely to occur if there is event uncertainty, that is, there not clarity on whether an information event has occurred. An investor does not know the cause of the decision cluster. When there is less market clarity, price adjustment will be slower so there will be serial dependence in price data.
Herd behavior will always exist in markets and create market inefficiencies because investors cannot untangle information events from liquidity trading or hedging. Hence, information traders may switch between herding and only using their private information. The authors develop a nice theoretical models that can be used to analyze transaction data. Herding is pervasive and dynamic and worthy of careful analysis by any market participant.
No comments:
Post a Comment