In business, risk is measured in dollars. This seems a simple statement, but behind it is a vastly complex and difficult concept: taking a nebulous, uncertain and wholly conceptual event and turning it into numbers.
It is like measuring the volume of the color yellow. How do you do that?
Whether the quantification is simply snatched from thin air, or run though a computer, it is done with models.
For those going with a “gut feeling” or relying on experience, the model being employed is called a heuristic. Heuristics are the mental programs that we develop to handle complex problems quickly. They have served us well for millennia; it was far better, after all, to simply run when you heard something roar inside of your cave than to calculate the odds that it was a cave bear and consult actuarial tables to guage the probablility of losing an arm vs. becoming bear chow.
Another type of model commonly used is the more formal computer model. This is what the astrophysicists and PhD’s working on Wall Street often use to assess market risk and make investments.
While these are often maligned, they are very powerful tools with distinct advantages over the gut-feeling approach. Trying to measure the returns on 45 different funds in never before seen market scenarios is not something that our minds ever evolved to handle.
In between these two approaches there are an infinite variety of models: spreadsheets, looking for patterns in charts, reading tea leaves, etc.
No matter what the model, they are terribly dangerous. The problem is that the person who creates the model, tends to believe it. And the more official, formal and credible the model is, the more compelling it is.
Models create their own siren song. We might know that they are luring us into folley, but we put a lot of time into them and they seem so right.
Even the smartest people fall into this trance. Take for example Nobel Prize (economics) winners Fischer Black and Myron Scholes. They created a model to value options that was so good, they build a multi billion dollar investment company around it. And it worked great…..until it didn’t. When it stopped working, it caused a colossal market crash that almost took down the US economy.
But that is not the most interesting thing about their model. Rather it is that in spite of this horrific, public failure, the Black-Scholes model is still the one most commonly used on Wall Street and elsewhere.
It is as if your best friend drank a glass of wine and immediately fell dead, and everyone else in the restaurant ordered the same wine.
This is not to say that quant models are any more dangerous than any other kind of model. Whichever one you use, you must not fall not the trap of believing that the model is reality.
In putting dollar figures to a risk, you have to make sure that you are not increasing the level of risk by believing that your assessment is anything but a sophisticated, well thought out guess.
It is like measuring the volume of the color yellow. How do you do that?
Whether the quantification is simply snatched from thin air, or run though a computer, it is done with models.
For those going with a “gut feeling” or relying on experience, the model being employed is called a heuristic. Heuristics are the mental programs that we develop to handle complex problems quickly. They have served us well for millennia; it was far better, after all, to simply run when you heard something roar inside of your cave than to calculate the odds that it was a cave bear and consult actuarial tables to guage the probablility of losing an arm vs. becoming bear chow.
Another type of model commonly used is the more formal computer model. This is what the astrophysicists and PhD’s working on Wall Street often use to assess market risk and make investments.
While these are often maligned, they are very powerful tools with distinct advantages over the gut-feeling approach. Trying to measure the returns on 45 different funds in never before seen market scenarios is not something that our minds ever evolved to handle.
In between these two approaches there are an infinite variety of models: spreadsheets, looking for patterns in charts, reading tea leaves, etc.
No matter what the model, they are terribly dangerous. The problem is that the person who creates the model, tends to believe it. And the more official, formal and credible the model is, the more compelling it is.
Models create their own siren song. We might know that they are luring us into folley, but we put a lot of time into them and they seem so right.
Even the smartest people fall into this trance. Take for example Nobel Prize (economics) winners Fischer Black and Myron Scholes. They created a model to value options that was so good, they build a multi billion dollar investment company around it. And it worked great…..until it didn’t. When it stopped working, it caused a colossal market crash that almost took down the US economy.
But that is not the most interesting thing about their model. Rather it is that in spite of this horrific, public failure, the Black-Scholes model is still the one most commonly used on Wall Street and elsewhere.
It is as if your best friend drank a glass of wine and immediately fell dead, and everyone else in the restaurant ordered the same wine.
This is not to say that quant models are any more dangerous than any other kind of model. Whichever one you use, you must not fall not the trap of believing that the model is reality.
In putting dollar figures to a risk, you have to make sure that you are not increasing the level of risk by believing that your assessment is anything but a sophisticated, well thought out guess.