Archive for June, 2013

“Tornado Warnings” for the stock market? Can we? Should we?

Wednesday, June 5th, 2013

When those with expert knowledge detect that a potentially calamitous event threatens, but they aren’t sure, human nature runs in two directions. We are torn between wanting to issue a warning so that folks can take precautions, versus wanting to keep everyone calm to prevent the damage that could be caused by mass panic.

Meanwhile, those being warned are faced with their own difficult choice. Do they ignore the warning, accept the risks, and go about business as usual? Do they drop everything and take shelter or flee for safety? What level of precaution is appropriate?

Consider first the case of actual tornados. According to Wikipedia, early forecasters were banned from making actual tornado warnings, because of the fear of public panic.

According to NOAA data, tornado warnings have been about 70% accurate since the mid-1990s. Now, we need to be careful when measuring “accuracy”, because even when a tornado warning correctly predicts a tornado touchdown, the tornado doesn’t touch everyone in the warning area. And even if a tornado doesn’t touch down, the severe thunderstorm can still do damage with high winds, hail, and heavy rain leading to flash floods.

But despite not being a perfect warning system, we know that the warnings (a) provide people with valuable time to take shelter, (b) do not disrupt activities excessively, and (c) save lives when tornadoes do touch down and wipe out buildings and vehicles. The issue with mass panic has been addressed through decades of careful public education in tornado-prone areas, training people to react appropriately.  People have come to accept and value the warning system, most recently in Oklahoma City where 16 minutes’ advance notice saved thousands of lives as the mile-wide EF5 tornado (an extremely bad one!) wiped out thousands of buildings. The warning didn’t prevent the tornado, but the community will recover much more quickly because they reacted well.

But now consider a warning that might, by itself, trigger human reactions which bring about the calamitous event that it is warning about! What if, whenever it is issued, it causes a certain part of the business community serious damage, while saving others from grave harm? And of course it is going to cause anxiety among those who are warned, and people don’t like that unless it’s really justified.  This is the situation one faces when considering a “Tornado Warning” system for the stock market.

There is a stock market “measurement” (technical indicator) which has a 30-year record of making predictions which have a 77% accuracy in anticipating a market decline of 5% or more. Let’s call any 5% decline a “market tornado”. The measurement does not predict every “market tornado”, but it has triggered before all the really bad tornadoes (with drops in the DOW of 20% or more) and nearly all of the “medium size” ones (10% or more).

Note that in a “market tornado”, as with a real tornado, the damage is far greater in some places than in others. Thanks to inflation and growth, the market as a whole recovers eventually, just as communities repair their damage after real tornadoes. But in a 5% “market tornado” decline, some stocks will fall much farther than the average, just as some homes suffer more damage than others in a real tornado. And if the market tornado is bad enough, some stocks get destroyed (bankruptcy) if the company isn’t in good financial condition, just as weaker houses get destroyed by very bad real tornadoes.

However, unlike in a tornado, a “stock market tornado warning” induces more people to try to sell, which tends to push prices down, and that might exacerbate the market decline that the warning is all about. Businesses involved in the market will be damaged by loss of customers, decline in revenue from remaining customers, and potentially losses on their own trading positions.  We must expect the psychology of market participants to be deeply resistant to warnings which cause them immediate harm, even if the warning provides a prudent opportunity to prevent the harm from being far worse! (For some reason I’m reminded of the scene in Jaws where the shark is spotted and the warning is issued, but the folks running the beaches refuse to spread the word because they don’t want to scare people having fun… and lose profits…)

On the other hand, perhaps when a market has run up too far, a “Market Tornado Warning” might actually mitigate the subsequent inevitable correction?  Perhaps with proper public education, market participants could simply take precautions, rather than panicking?  A warning gives those caught “offsides”, overextended on leverage, a chance to pull in their positions, helping the market to be less overextended and less at risk of a deep crash. Those looking to buy in might hold back a bit, expecting a dip in prices ahead. By slowing the rally and creating a pool of folks willing and able to “buy the dip”, the market tornado might be far smaller than it would otherwise have been. This would stabilize the market, by preventing so many from buying too soon and then becoming sellers at the worst time for all.

In addition, with a reasonable warning system, businesses catering to market participants would have a chance to issue polite warnings that actually help their clients. They would have to accept that some short-term damage to portfolios is very likely, but they would have a critical window of opportunity to work constructively to minimize that damage. And that could also minimize the risk of deep damage to the business.

Perhaps, when the market is due for a correction, the warning actually allows the correction to take place in a less-damaging manner?  That is a tantalizing thought, because the horrific economic damage created by deep market declines affects millions or billions of people…

Oh, and the name of that “Market Tornado Warning” technical indicator?  It’s usually referred to as the Hindenburg Omen.  I believe that’s a poor choice of name, since the name alone induces more panic than it should.  And the indicator itself isn’t perfect and could probably be made better. But more about the Hindenburg later…