Posts
Why There's No Such Thing As A Black Swan Event
Nassim Taleb maintains that much business forecasting is useless as events, positive or negative, that is deemed improbable yet causes massive consequences. Just because he and his fellow 'experts' cannot read the markets, it certainly does not mean others also cannot:-
(Rick Bookstaber is senior policy adviser at the SEC.)
All days in a normal market seem the same, but when a crisis occurs, it seems as if we have never seen the likes of it before. Each crisis brings evocations of Black Swans with twenty standard deviation tails swimming in the waters of a hundred year flood. But of course, we have seen it many times before, or at least some aspects of it. The cause might be different, the initial market from which it propagates might take us by surprise. But the path a crisis takes, at least in broad strokes, hardly differs from one case to the other.
We all know the limitations of standard risk management methods in dealing with times of market crisis. And we are starting to get a sense of what is needed beyond these methods in order to see a market crisis coming, things like understanding who is under pressure, what sorts of positions they hold (and thus might be forced to liquidate) and who else is holding those positions (and thus who might get caught up in the propagation of the forced selling).
Common Sense about Market Crisis
Unfortunately, although we can hope that this sort of information will end up with those regulating the markets, it is beyond the realm of anyone in the private sector. But here are a few common sense things we know about the way markets behave during a crisis:
Correlations rise because people don’t care much about the subtle characteristics of one instrument versus another. Everything is either high risk or low risk, high liquidity or low liquidity. I think of the market during a crisis like in high energy physics, where matter melds into a homogeneous plasma when the heat gets turned up.
Because liquidity becomes critical, the less liquid markets – emerging markets, low cap stocks and the like – take it on the chin more than their more liquid cousins.
(Oh, and what about gold? Sometimes it responds, sometimes it doesn't. There is nothing intrinsic about gold that makes it part of the crisis/no-crisis equation. If it is a flavor-of-the-month market, it will respond positively, otherwise, it will simply act like a commodity, responding to economics).
Knowing this, it is not hard to take steps to protect against a crisis. Just move away from equities, avoid being short volatility, stay away from credit-laden debt, focus on the liquid markets, and watch those carry trades. Also, don’t trust diversification, because those low correlations you are depending on will not be there when it matters.
Or, if you want to be more sophisticated about it, create a variance-covariance matrix predicated on these sorts of relationships, and be sure to add a constraint to your portfolio optimization so that you will not breach a specified risk level under this crisis-based matrix. For example, if your usual risk constraint is to keep you portfolio volatility below twelve percent, perhaps you also make sure it won’t be higher than thirty percent in the case of crisis. Or, because we know the direction of these market effects, to make life simpler you can add a simple scenario test, and not allow the portfolio to lose more than, say, ten percent in that scenario. Doing this will guard against the tendency to over-rely on diversification, over-lever or put too much exposure into the markets that are particularly sensitive to a crisis.
The problem with this advice is that it is exactly the opposite of what will make sense when the crisis has yet to emerge. More to the point, it is exactly the opposite of what makes money as the market is building into a crisis.
Before a crisis (I won’t say “during a bubble”):
Correlations are low because, in an attempt to find value in when every portfolio manager, trader and dentist is spending his time combing the hills for value, the slightest difference between otherwise similar instruments is worth mining. And with the languid pace set by the low volatility and money sloshing over the sides, people have the luxury of spending time in fine-tuning.
With so much money flooding into the market (and so much money means so much leverage), people start to scan the landscape for the less known – and less liquid – markets to find value.
Regime-switching models of market crisis
It might be reasonable to consider crises as hundred year flood events if we mistakenly treated them as being drawn from the same distribution as those normal market days.But they are not . It is following a different dynamic, a dynamic that we have seen enough to become familiar with. People sometimes look at periods of market crisis in the context of a regime switching model, and this gets more to the point. There are the normal times and then there are the crisis times. But what I am suggesting above is that there are (at least) three regimes. There are the crisis times, the normal times, and the pre-crisis times. The transition generally is not from normal to crisis, but rather from pre-crisis to crisis. And the move from the pre-crisis to the crisis regime is more gut-wrenching because in almost every dimension things are moving from one extreme to another.
The killer is that what protects you in a crisis is also what leaves money on the table pre-crisis. The best trades and market positions in the pre-crisis regime are the ones that cause the greatest losses in the crisis.
All days in a normal market seem the same, but when a crisis occurs, it seems as if we have never seen the likes of it before. Each crisis brings evocations of Black Swans with twenty standard deviation tails swimming in the waters of a hundred year flood. But of course, we have seen it many times before, or at least some aspects of it. The cause might be different, the initial market from which it propagates might take us by surprise. But the path a crisis takes, at least in broad strokes, hardly differs from one case to the other.
We all know the limitations of standard risk management methods in dealing with times of market crisis. And we are starting to get a sense of what is needed beyond these methods in order to see a market crisis coming, things like understanding who is under pressure, what sorts of positions they hold (and thus might be forced to liquidate) and who else is holding those positions (and thus who might get caught up in the propagation of the forced selling).
Common Sense about Market Crisis
Unfortunately, although we can hope that this sort of information will end up with those regulating the markets, it is beyond the realm of anyone in the private sector. But here are a few common sense things we know about the way markets behave during a crisis:
- Equities drop
- Volatility goes up
- Credit spreads widen
- Correlations rise
- Areas of low liquidity decline more than similar areas with high liquidity
- The yield curve flattens
Correlations rise because people don’t care much about the subtle characteristics of one instrument versus another. Everything is either high risk or low risk, high liquidity or low liquidity. I think of the market during a crisis like in high energy physics, where matter melds into a homogeneous plasma when the heat gets turned up.
Because liquidity becomes critical, the less liquid markets – emerging markets, low cap stocks and the like – take it on the chin more than their more liquid cousins.
(Oh, and what about gold? Sometimes it responds, sometimes it doesn't. There is nothing intrinsic about gold that makes it part of the crisis/no-crisis equation. If it is a flavor-of-the-month market, it will respond positively, otherwise, it will simply act like a commodity, responding to economics).
Knowing this, it is not hard to take steps to protect against a crisis. Just move away from equities, avoid being short volatility, stay away from credit-laden debt, focus on the liquid markets, and watch those carry trades. Also, don’t trust diversification, because those low correlations you are depending on will not be there when it matters.
Or, if you want to be more sophisticated about it, create a variance-covariance matrix predicated on these sorts of relationships, and be sure to add a constraint to your portfolio optimization so that you will not breach a specified risk level under this crisis-based matrix. For example, if your usual risk constraint is to keep you portfolio volatility below twelve percent, perhaps you also make sure it won’t be higher than thirty percent in the case of crisis. Or, because we know the direction of these market effects, to make life simpler you can add a simple scenario test, and not allow the portfolio to lose more than, say, ten percent in that scenario. Doing this will guard against the tendency to over-rely on diversification, over-lever or put too much exposure into the markets that are particularly sensitive to a crisis.
The problem with this advice is that it is exactly the opposite of what will make sense when the crisis has yet to emerge. More to the point, it is exactly the opposite of what makes money as the market is building into a crisis.
Before a crisis (I won’t say “during a bubble”):
- Equities are rising
- Volatility is low
- Credit spreads are narrow
- Correlations are low
- The opportunities are in the hinterland markets of low liquidity
- The yield curve is steep
Correlations are low because, in an attempt to find value in when every portfolio manager, trader and dentist is spending his time combing the hills for value, the slightest difference between otherwise similar instruments is worth mining. And with the languid pace set by the low volatility and money sloshing over the sides, people have the luxury of spending time in fine-tuning.
With so much money flooding into the market (and so much money means so much leverage), people start to scan the landscape for the less known – and less liquid – markets to find value.
Regime-switching models of market crisis
It might be reasonable to consider crises as hundred year flood events if we mistakenly treated them as being drawn from the same distribution as those normal market days.But they are not . It is following a different dynamic, a dynamic that we have seen enough to become familiar with. People sometimes look at periods of market crisis in the context of a regime switching model, and this gets more to the point. There are the normal times and then there are the crisis times. But what I am suggesting above is that there are (at least) three regimes. There are the crisis times, the normal times, and the pre-crisis times. The transition generally is not from normal to crisis, but rather from pre-crisis to crisis. And the move from the pre-crisis to the crisis regime is more gut-wrenching because in almost every dimension things are moving from one extreme to another.
The killer is that what protects you in a crisis is also what leaves money on the table pre-crisis. The best trades and market positions in the pre-crisis regime are the ones that cause the greatest losses in the crisis.
No comments:
Post a Comment
Note: Only a member of this blog may post a comment.