As I suggested in my introductory post, the market crash on October 19, 1987 (at over 22% still the largest single day percentage drop on record) remains the most illustrative of market meltdowns.   The dynamics that led to the crash in 1987 are surprisingly transparent, and what happened that day exposed the fundamental dynamics which consistently appear when markets crash. 

Let’s go back to 1987.  There was no nationwide housing bubble, and no evidence of a massively over-leveraged economy.  There were only the most primitive forms of credit derivatives and Special Purpose Entities (SPEs).  Yet from its high in August, 1987 to the low on October 20, 1987 the market declined in value by roughly 40%.  Sound familiar?

The “villain” in the October, 1987 crash was a product called portfolio insurance.  Portfolio insurance was a system that seemed to offer institutional investors a way to “lock in” stock market gains by creating a synthetic put option.  That may sound like a complicated strategy, but in essence it is quite simple.  Recall that a put option is the right to sell an asset at a given price.  Here, the put option provides the holder a right to sell a market bundle of shares at a given price.  The synthetic aspect refers to the fact that there is no one on the other side of the trade agreeing to buy the market bundle of shares at a given price.  Instead, the institutions that “purchased” portfolio insurance put in place a trading strategy that replicates a put option.  What does such a trading strategy look like?  Sell as prices decline. 

Yes.  A holder of a put option is doing the opposite of what savvy investors are told to do, and this is just what numerous institutional investors committed to do.  Rather than buy low and sell high, the put option holder buys high and sells low.  The reasoning is as follows:  if share prices go up, then you are fully invested.  If share prices go down, then you have sold all of your holdings and have locked in your gains.  The only cost in executing such a strategy comes from volatility.  Volatility leads you to buy into a rising market and sell into a falling market, which can be costly over time (this is how the Black-Scholes model values an option).

The next step is to explain why investors purchasing a synthetic put option can cause a market crash, although it may be obvious at this point.  Who agreed to purchase the assets at a lower price in 1987?  No one.  Purchasers of portfolio insurance assumed that they could rely on a liquid market to sell their shares into as share prices declined.  But their assumption was based on a misunderstanding of what makes for market liquidity.  Markets are only liquid when supply equals demand, and if enough investors “purchase” portfolio insurance there are two quite different equilibriums at which supply equals demand.

By way of simple example, there could be one equilibrium point where the Dow is at 2,200 (this was twenty years ago), and institutions holding portfolio insurance are fully invested.  There could then be a second equilibrium point with the Dow at 1,700, and institutions holding portfolio insurance fully divested.  There is liquidity with either the Dow at 2,200 or with the Dow at 1,700, but not in between. 

The existence of two distinct equilibrium points is explained by the unusual shape of the aggregate demand curve for equity securities when enough investors follow a similar portfolio insurance strategy.  Usually demand curves are downward sloping: as prices decline, demand rises.  In the situation I’ve just described there is a range over which the aggregate demand curve for equity securities slopes upwards.  As prices decline, demand falls. Those familiar with microeconomics will recognize that I am describing what is known as a Giffen good.  We usually think of potatoes as one of those rare examples of a Giffen good, but in October, 1987 we saw that equities could also be a Giffen good. 

There is good reason to expect that the price of a Giffen will be unusually volatile.  In even a very simple economic system, if there are multiple equilibrium points it can be virtually impossible to predict the movement between different equilibriums.  Such a system is complex in that simple dynamics lead to outcomes that are unpredictable unless all of the details of the starting point are known.  Enough investors following a portfolio insurance strategy can create an aggregate upward slope demand curve, which, in turn, leads to a highly unstable market price.

The discussion above provides, of course, only a simplified description of the many different investors involved in 1987.  Nor does this account consider any of the various psychological dynamics that also may have affected behavior.  Setting aside these caveats, the stock market crash of 1987 remains a wonderfully revealing event.  For a brief period of time the trading strategies of major market players were sufficiently simple and transparent to enable us to make a pretty good guess at why the US stock market crashed that day. 

In my next post I will talk about the regulatory response to the October, 1987 stock market crash.  As a preview to that post, I will only say that I hope we do better this time around. In subsequent posts I will show why the crash of October 1987 is not a special case, but rather a window into the dynamics that lead to market crashes generally.

Financial Crisis | Bookmark

TrackBack URL for this entry:
https://www.typepad.com/services/trackback/6a00d8345157d569e2010535781efd970b

Links to weblogs that reference Why Markets Crash: Lessons from October, 1987: