Black Swan Events And Economists

Basically the black swan theory is that people underestimate extreme rare events. It’s called the black swan because until austrailia was discovered people made saying that was similar to the “when pigs fly” statement, only it was “if there’s a black swan”. Statisticians understand sampling, and they might say that if you count the number of leaves on one branch, you can bet with 95% accurate within the margin of a number like say 10 leaves how many exist on the entire tree. This is true in lots of instances, however to say there is no black swan, or if there is the sample is under 0.000001% of the swan population would actually have turned out to be a horrible bet. In fact in Austrailia all swans are black. the entire world doesn’t operate the same exact way. You could sample the amount of blackswans in a very vast amount of space and still not find one, but the discovery of Austrailia would be a complete game changer to your way of life.

My take of it based on a handful of interviews and speeches I’ve seen is this. There are several economists with PHDs on the theories of the economy. Some of them are even nobel prize winners for their theories. The problem is they are just that… theories, models, and data based on models that may not neccesarily reflect reality. Even statisticians estimate reality based on theory. It would be like betting a roll of the dice is random when the laws of physics will dictate that if you account for the exact windspeed the exact force, spin, and several other variables, you could in fact “game” the system. Lets say casinos allowed people to use robots, and the robot would throw the dice the exact same way every time, resulting in the dice always coming up the same exact way. Unless they kicked robots out, the casino would be out of business because they operate under the assumption that a dice roll is random. People are often fooled by randomness and the illusion of control, however it’s possible that control is not an illusion, it just requires such presise control that it rarely is actually figured out how to control it, and even if they figure out how, in order to actually get your hand to move in the exact way neccesary to control it would be very improbable. Additionally, casinos would still make enough money in the meantime from all the people who think they can game it when they can’t get the technique and all the people betting against them to make up for the one guy that has perfected the throw. Unfortunately the financial markets are not casinos. Random walk theory and effecient market theory may not be correct. If that’s the case, the financial institutions could theoretically some day be out of business.

In the 2001 when september 11th occured, but with low interest rates combatting it, and increased availability of loans, it appeared the crisis was avoided. Perhaps it’s possible that this just helped increase the sale of more risky assets such as subprime mortgages. In 2008 the subprime mortgage crisis seemed to be evidence of that theory being reality. It’s certainly possible that that is merely people being “fooled by randomness”. It’s certainly possible that we will at some point have a sustainable recovery. But this is all just theory.

Talib himself has said that some prediction is better than nothing and certainly thinks he could be wrong. He says he wakes up every day hoping he is in fact wrong about his thoughts on what could happen in the future.

There is randomness and there are unknowns. Most economic theories do recognize this and make probabilities of events occuring rather than predicting and saying this will happen. The problem is, they fail to take into account the probability that these probabilities may actually be more likely of occuring than they estimate. The danger is that the world doesn’t operate in standard “bell curves” under the assumption that events are independent from each other. I’m not saying that it doesn’t, however economists operate under the assumption that these theories are right.

Investment banks operate under the assumption that these PHD award winning economists are right. So they operate under the assumption that the risk of catestrophic loss might be less than 5% as a result of their total decisions made when it could be much higher. Additionally, they operate under the assumptions that certain bets are independent from each other. This would be like insurance companies insuring a bunch of houses connected together by TNT based on independent risk of a fire. In reality, if one house catches on fire, a chain reaction effect occurs resulting in all of them catching on fire, and insurance companies going broke. Perhaps even data had been gathered on house fires based on sampling and they concluded that not to be the case. However they may not have taken into consideration a rare “black swan” event, such as discovering that TNT had been underground this whole time, and only after several rebuilding of previous housefires had enough dirt been dug up where the fire actually exposes the risk of the TNT getting hot enough where it exploads, setting of a chain reaction. Perhaps an insurance company has actuarial scientists say there is a 1/1000 chance that the house blows up. If the insurance company says, for every $1 you give us, we will give you $1000 in the event that your house gets destroyed, the insurance company would believe to gain based on the rate of interest that they get with the money. In reality they may use consumers fear to even charge them $2. They use their risk management models and make sure they have enough liquid assets so that in the event that a house does catch on fire they can pay for it, and the rare event that 2 houses catch on fire, they can pay for it. However, they dramatically underestimate the risk of 2 houses catching on fire at the same time. In this example, it could almost be about the same as 1/1000 as opposed to the theoretical 1/1,000,000. As you can see as the insurance company continues to insure more and more houses based on these bad assumptions, they are dramatically overexposed to risk.  Imagine that the financial community underestimates risk in the same way.

What if a stock plunging resulted in margin liquidations which resulted in more liquidations and selling which in turn resulted in more. Because of the lack of capital resulting and the insane amount of risk, a domino effect could cascade stocks lower and lower as bank after bank topples over. Of course the interconnectedness of the markets are only one problem. The amout of leverage as well as political ramifications that result in various responses have a cause and effect relationship that may actually change the risk profile of every company entirely, which in turn results by a different response.

It’s not neccesarily a bearish only event, because the amount of liquidation resulting in the opposite effect of massive creation of capital to pay for the derivative metltdown could result in all sorts of capital created that is used for a gamechanger in the positive direction. It could result in deflation, or hyperinflation. The dollar could soar in value as credit unwinds and the banks go under. Actually lessening competition could accelerate the clearing out of ineffeciencies in the market that are actually a burden on the system. Of course maybe the stimulous is good, maybe it is bad. There’s so many variables that you cannot understand and to pretend you can based on a model that’s dependent upon it being a closed loop where things don’t take into account important factors could be a huge mistake.

Lets say you can identify one single cause effect relationship. For example, if you turn up some sort of chemical reaction, it can lead to the heating up of something. But lets say for example, you failed to take into account a certain element and that chemical reaction instead caused the creation of liquid nitrogen which would cool down everything instead. Now you realize you’re getting colder, so you ramp up the experiment. You double the chemical reaction and leverage in this experiment, not realizing there are certain elements like nitrogen in the air that completely change this reaction which worked in theory. So instead you end up freezing yourself to death and the more you try to change it, the worse it gets.

If you want something more mathematical, lets say you believe that the equation to be y=4x and y is set at 1. You believe that X would be equal to 1/4th. Lets suppose X is the value of wealth added to the world by every dollar of stimulous. You notice X decreases and in order to keep things balanced you rapidly increase Y. However, what you may not realize is that the result of a seemingly irrelevent event has instead changed the equation, or perhaps the equation was already wrong. The equation is actually y=-4x. The more you increase y, the worse the situation gets. Even if it’s very rare that our assumptions are wrong, to operate under extreme leverage and continue to ramp up the leverage under this assumption would still be financially irresponsible.

The economy may be lacking in understanding, and this isn’t a specific theory as much as simply some examples showing that it’s possible that everything we think we know could be wrong.

Outside of the lab there are things that are much more connected in ways that can’t be described. Additionally, the extreme rare events occuring could have such a great unknown impact that it becomes a serious “game changer”.

An example from my perspective is that independently you might bet under the assumption that there’s a 5% chance of an extreme event happen, but you make that bet over and over again. A casino bets on their edge of 51% on blackjack knowing that only one hand will be dealt at a time. This is an independent game. No matter how many hands the house bets at once at seperate tables against seperate people, their risk does not increase. So the casino could actually have so many tables and players against them that they have all their money at risk, and it would be the exact same odds of them going broke as it would at a single table betting one after another. However the financial markets SEEM to operate much more dependently. What if instead the casino had only 52 cards only and they delt out 18 different hands at the same table and had all their money at risk. This risk would not be independent. Anyone with a brain could see that if the house had an ace showing and no single player out of 36 cards had a face card that they would not be able to win. They could then buy insurance. If instead the dealer showed a 6, the players could see no face card out there, and they would have 0 reason to hit. Knowing there’s only face cards left in the deck and in the dealer’s hand, the only decision they would make is either to stay if above 11, and double down if 11 or less. If they had even a slight ability to understand the odds and probability the house would actually be in a losing situation and no matter how many tables they set up at once, it’s not going to help them. This is similar to the financial markets if they operate dependently. If someone makes bet after bet assuming that one event is independent from another, and makes highly leveraged bets at a very high frequency, they could be in serious trouble. If they assume that the probability of a rare event occuring is 5% when it’s actually higher, they could be in serious trouble. If each bank is also interconnected and dependent in ways which are unknown, the collapse of one could in fact be a domino that inevitably will knock down the entire system.

Now black swan events are not just negative. The invention of the computer and the internet and any major breakthrough in society is a black swan event. This has huge consequences, but not neccesarily bad ones.

It’s possible that out of the huge domino effect comes something great. That out of something really bad comes something great. Perhaps because of this domino effect and stimulous, we get a major increase in the quality of life and the utility of all individuals increase. Perhaps the large risk that we takes actually ends up paying off in a huge way and results in a positive black swan event. Perhaps because of the domino effect the response of major stimulous results in more credit available which results in people taking out certain loans to create companies that never would have been invented that actually enhances our lifestyle dramatically.

Although it does appear to be much more pessamistic in nature, and people don’t like the fact that life is a greater unknown that we could imagine, it isn’t neccesarily the case.

Tags: black swan, economics, economy, normal distribution, rare events