Against the Gods Book Review

Against the Gods Book Review

Have you ever made a wager in Las Vegas or Atlantic City? Have you made a decision on what insurance premiums you want to pay? If the answer is yes to any of these questions, you have experience with risk management. Since the most recent financial crisis, there has been more emphasis on it. As with individuals or companies, there is a healthy level and an unhealthy level of risk. Of course, a risk versus reward analysis needs to be performed too. Against the Gods is about the history of risk management from its humble beginnings trying to figure out the chances of winning in an unfinished game to its evolution into a complex field. It is a fascinating book if you like to understand how risk works. Below are my thoughts on some of the parts of the book that I found interesting:

 1. Probability of Winning an Unfinished Game

The beginning of risk management started with men trying to figure out the chances of winning an uncompleted game. For example, suppose that a player needs to win six rounds before winning the game. One player has only won five while the other player only won three rounds before the game is finished. How do you divide the stakes? This question was a “brain-teaser” during the sixteenth and seventeenth century. As the story states, the answers for how to divide the stakes is “the beginning of the systematic analysis of probability – the measure of confidence that something is going to happen” which is the quantification of risk. (Bernstein 43)

This example is simple but it is a great analogy for chance and risk. Although the first player has won five games and on the brink of victory, there is still a possibility that the second player can comeback and pull out the victory. Hall of Fame New York Yankee catcher Yogi Berra would say “It ain’t over till it’s over”. Nevertheless, there is clearly a better chance for the first player to win as he only needs one win as opposed to three wins for the second player. There are more permutations that the first player will come out on top. As a result, quantifying these chances into actual odds is the challenge of risk management. In the business world, it is many times more complex than two players trying to win a game as there are many more scenarios to consider and quantify.

2. Cardano’s Liber de Ludo Aleae (Book on Games of Chance)

In Cardano’s Liber de Ludo Aleae, he explores the probability of a roll of the dice. The dice is a simple example of chance as there is an equal chance to land on any of the sides. As a result, there is a 1/6 chance to land on any number and 1/3 chance to land on either pair of numbers. However, what is the probability of throwing a 1 or a 2 with two dice? At first, one would think the probability is 2/3. However, there is a chance that you could land on the same number twice. Consequently, you need to account for that 1/9 chance and subtracting it from 2/3 to arrive at 5/9. (Bernstein 50-51)

This dice is a good example as it is easy to understand probability since the chances are clearly the same for each number unless it is a deformed dice. However, I think it is also a good example of the challenge of risk management because it is difficult to observe the odds. There is no other way or better way to quantify the chances of rolling a dice. However, if you roll the dice 1,000 or 10,000 times, you will not arrive at the calculated odds. In terms of business; although you could assess that an event could have a one in a million chance of occurring, it could happen and devastate your company. For example, the chances of the most recent credit crisis occurring were assessed as low by many but it happened and crippled many organizations. Even after you are able to quantify different situations, preparing for and managing those situations is a daunting task.

3. The Existence of God

Pascal’s Wager deals with whether God exists or does not exist. It is a question that reason cannot solve. It is basically a coin toss if you want to say heads is “God is” or tails for “God is not”. This analysis was the beginning of “Decision Theory”. Some decisions are made based on past experiences. However, there are no past experiments that can prove God. As a result, the decision is made on “future consequences”. If God does exist and one does not believe in him, the individual risks eternal damnation. If God does not exist, it does not matter if you are a good person or not. Based on this analysis, one would believe in God as the risk of eternal damnation dwarfs the notion of not mattering. (Bernstein 69-70)

Clearly, a devote religious person would place the odds of God’s existence closer to 100% while an atheist would put the percentage closer to 0% but not entirely rule out the possibility as it is impossible to completely prove non-existence. Nevertheless, the example introduces decision making influenced by the magnitude of the consequences. Weighing the opportunity costs of lost pleasures (which is something that religion clearly limits in practice) against the infinite loss that would be incurred from eternal damnation is a decision that every individual makes. Most people in the world believe in some form of higher being. As a result, even if some believe the risk of damnation is unlikely, they have decided to protect themselves by believing as the fear of the consequences is so great that it overwhelms any benefit that could be gained by not believing.

This type of analysis was clearly missing leading up to the credit crisis. The equivalent of eternal damnation is bankrupting firms, even large brand name institutions, by investing a ridiculous amount of a firm in risky positions. However, the pleasure or opportunity cost of not investing as heavily would have been lost profits and thus smaller bonuses. Nonetheless, the risk of eternal damnation was not considered. While no one could have predicted the exact effects of the credit bubble bursting, each firm should have analyzed what would happen if it did and realized that too much of its assets were exposed to it. While it is unrealistic to expect businesses to not invest in those high yield assets during its height, firms should have limited the exposure to something it could have survived. It was not a simple toss of the coin as in Pascal’s Wager. Firms had more than the two options of investing or not investing. They could have chose to concentrate less of their portfolio in risky assets, had smaller profits, and survived if the very unlikely did happen. As it turns out, the credit God exists and he was looking for vengeance.

4. The Importance of Statistical Sampling

“Sampling is essential to risk-taking”. Samples of the past and present are used to “guess about the future”. It is inefficient or even impossible to get a 100% sample size. In 1662, John Graunt released his book, Natural and Political Observations made upon the Bills of Mortality. The book recorded births and deaths in London from 1604 to 1661. The key technique used in this book was statistical inference. For example, Graunt used the basis of an average of 8 people per household and multiplied that average by the number of households, 48,000, to predict a total population in London of 384,000. (Berstein 73-82)

Statistical sampling is a very important tool. From the point of an economist, the cost of obtaining a 100% sample to make every decision would outweigh the benefits. As a result, having a statistical sample to project the total population is a very powerful tool to make general observations and conclusions at an acceptable cost. Consequently, it is important to risk management as sampling 100% of past observations to make predictions about the future would be an incredibly expensive exercise if not an impossible one.

5. Lloyd’s of London

Insurance is dependent on “sampling, averages, independence of observations, and the notion of normal”.  Lloyd’s of London was born in a coffee house. Edward Lloyd opened a coffee house where many seamen gathered. Based on the information provided by those men, Lloyd compiled “Lloyd’s List” with information about arrivals and departures or ships, conditions abroad, and conditions at sea. Consequently, the coffee house became a hub for information. At first, individual underwriters would insure individuals for a plethora of events such as robbery, fire, death, etc. Eventually, the underwriters banded together to form one, large insurance company. (88-91)

As one could see from the origins of Lloyd’s of London, information is the key to insurance. Without that information, underwriters cannot properly assess the risks. The success of an insurance company is contingent on the premiums from a pool of policyholders being significantly greater than any expected losses that the insurance company would need to pay out. If an insurance company has poor information and makes a poor assessment of risk to calculate insufficient premiums, the losses that it will be insuring will bankrupt the company.

Another key concept related to insurance companies is the transfer of risk. Individual policyholders are transferring their risk to an underwriter or insurance company for a premium. The insurance company is able to mitigate its risk by pooling the premium of many policyholders in order to have sufficient capital to pay out any potential losses. Moreover, the insurance company makes a significant profit for performing the service and acting as an intermediary for the policyholders.

In theory, credit default swaps work in this manner. However, credit default swaps are different in that “credit events” can cause a firm to pay the buyer of the swap. For example, a car insurance company pays a policyholder if a car sustains damage from an accident. If car accident swaps existed, the company may have to pay out if the speedometer is broken and could potentially lead to an accident without an actual one occurring. Adding onto the problem, the swap increases the magnitude of the payout. Even with higher premiums, there is tremendous danger if you have too large of a concentration in these instruments that act as insurance which is the problem AIG faced.

6. 7 Million human beings or one elephant?

A distinguished, Soviet professor on statistics rationalized not staying in a bomb shelter because there were 7 million people in Moscow so he believed the probability of him getting hit was small. However, once the one elephant in the city was killed by bombing, the professor showed up in the bomb shelter as he realized that if the elephant can get hit, he could get hit as the probability of the elephant getting killed was a lot smaller. (Bernstein 116)

While the professor used the new information to come to a different conclusion, it is an example of how our psychology as human beings clouds our ability to properly gauge risk. The probability of a person being hit did not change after the elephant got hit. However, the vivid image and emotional response to a traumatic event drastically changes our perception of probability even though nothing changed. Even a professor in statistics fell victim to his emotions. An example is the fear of flying after 9/11 that hurt the airlines and tourism. People became afraid of flying for fear of being on a hijacked plane after the traumatic images of the terrorist attacks. However, the response is contradictory to the facts. Terrorist attacks are a rare event as it takes years to plan a significant attack. Moreover, immediately following attacks, the security for airplanes would have been at its peak and thus the chances of an attack immediately after were probably even smaller than it is normally (which is already improbable).

Similarly, this trend could be observed during the most recent recession. During the boom, investors flocked toward high yield investments, even though they were risky, due to the emotional thrill of making money overriding any sense of risk. However, in the downturn, the emotional response led to an overreaction to risk which led to great opportunities for calm investors to make a significant profit on stocks that became undervalued due to the panic. Being able to control one’s emotions and capitalize on others is a great attribute for investing. In the words of Warren Buffett, “Be fearful when others are greedy. Be greedy when others are fearful.”

7. The Art of Risk Management

The notion behind risk management is that “We are never certain; we are always ignorant to some degree. Much of the information we have is either incorrect or incomplete.” In a world of uncertainty, the choice is between rejecting or not-rejecting a hypothesis. If the probability of being wrong is miniscule, you do not reject the hypothesis. However, you cannot accept the hypothesis as the chance of being wrong is not 0. If the probability of being wrong is large, then you clearly reject the hypothesis. (Bernstein 207)

In many situations in the world, it is too complex to accept a hypothesis. Think of throwing a ball. Its trajectory and distance is variable. It is because force from the hand, friction, wind, etc. all affect a throw. However, many of the variables are not readily, observable by us to predict the exact trajectory or distance. Similarly, there are too many variables in most situations, especially in the business world, that affect an outcome. As a result, it is improbable to be 100% correct about something. Consequently, you cannot accept a hypothesis if you are not 100% correct. In essence, you are always wrong. It is just a matter of how wrong you which is the essence of risk management.

 

Works Cited

Bernstein, Peter L. Against the Gods The Remarkable Story of Risk. New York: Wiley, 1998. Print.

Comments