The Impact of the Highly Improbable

Uncertainty - Randomness - Extremistan - Fat Tails - Antifragile - Risk

Hello, I am Nicolas Bustamante. I’m an entrepreneur and I write about building successful startups.

Check out some of my popular posts:

Intelligent CEO

The Impact of the Highly Improbable

Surviving Capitalism with Competitive Moats

Subscribe to receive actionable advice on company building👇


The year 2020 reminded me of how extreme and uncertain life is. Being lockdown for several months while living through a health and economic cataclysm is tough. It was the year to read Nassim Nicholas Taleb's books. I read his five essays on uncertainty and risk (Fooled by RandomnessThe Black SwanAntifragileSkin in the GameThe Bed of Procrustes) and, modestly, try to understand the highly technical Statistical Consequences of Fat Tails). It changed the way I see the world. 

We tend to think that we live in what Taleb called, Mediocristan. It's a world in which large events are rare and inconsequential. In Mediocristan, no single event can influence the total. Even the largest observation will be insignificant to the sum. The distribution is Gaussian. It means that most observations are around the average, and the odds of a deviation decline exponentially as it moves away from the average. The standard deviation, which is a measure of the variation of a set of values, is low. All the values tend to be close to the mean. Our brain loves the average because it’s easy to make predictions in such an environment.

Our world is, however, dominated by the extreme, the unknown, and the wildly improbable. We live in Extremistan. An environment in which one single observation can disproportionately impact the total. Taleb calls this event Black swan. They are rare; nothing in the past can convincingly point to their possibilities. They are extreme; a single event can change the entire distribution. They are post-rationalized; despite their outlier status, humans create explanations after the fact to make it explainable. They are observer-dependent; a Black swan for someone might be a White swan for someone else. Black swans make what we don't know more relevant than what we know. One single observation can invalidate a supposed truth that lasted decades. For instance, in 16th century England, people thought all swans must be white because no records reported black swan existence. Then, in 1697, Willem de Vlamingh saw black swans in Australia, thus blowing up a centuries-old belief. 

We overvalue our knowledge and underestimate the probability of being wrong. We tend to think that everything has an identifiable cause and, we don't withhold our judgment by saying, "I don't know." Life is so random and complex that we can't know most things. It is almost impossible to predict the future. Think about the covid crisis, the 2008 crisis, the 9/11 attack, the Great war. Rare, unpredictable, extreme impact, and post-rationalized. My favorite analogy to understand Black swans is the turkey fallacy. The butcher feeds the turkey every day, leading to a constant increase in the turkey's happiness. And yet, one day before Thanksgiving, something rare, unpredictable, extreme will happen to the turkey. The turkey died when the feeling of safety was the highest. The turkey relied on empirical knowledge, extrapolating the past to predict its future. Alas, such reasoning in extremistan is misleading and irrelevant. The lesson is clear: don't be the turkey. 

Black swan events changed the distribution. In statistical terms, it goes from thin-tailed to fat-tailed distribution. From Mediocristan to Extremistant. Because such events are so unlikely to happen and impossible to predict, many choose to ignore the possibility. Taleb argues that many rule out Black swan' possibility rather than building all risk models around them. Yet, fat tail events explain everything in our world. In a normal distribution, five or more standard deviations are almost impossible. In fat-tailed distributions, however, standard deviations are undefined. The average doesn't matter; the most important is the magnitude of the extreme event. Most people, with first and foremost economists and bankers, use statistical tools designed for mediocristan. Measures such as Beta, Sharpe Ratio, Black-Scholes are useless and dangerous in Extremistan. These are huge consequences. It's striking in the investment world, where returns are good until a single rare event ruins most investors. Taleb recalls several anecdotes of investors making money every month for a long time and then losing everything in a few hours. How to survive in a Black swan-prone environment? 

First, one has to identify if the situation takes place in Mediocristan or Extremistan. The difficulty is that confirming a Gaussian distribution requires data over decades. Remember that one single event can change the distribution from thin-tailed to fat-tailed. Moreover, with fat-tails, the previous minimum or maximum isn't a good indicator of future extremes. Most stockbrokers got hurt in 2020 when the oil price went below zero - way under its precedent minimum. Such a situation only exists in Extremistan. Hopefully, we know some things are from mediocristan - for instance, humans' height and weight. Unless we discover a human as tall and heavy as a big building, we know it's a Gaussian distribution. From previous Blacks Swan, we know that stocks, financial products, wealth distribution, or the banking system live in Extremistan. Once one knows about the distribution, the key is to prepare. 

It's better to spend time thinking about avoiding being the turkey rather than trying to predict the future. In most situations, predictions don't matter. It's more useful to prevent the risk of getting hurt. If one is fragile, large events hurt exponentially more than small events. If one is robust large events don't affect you that much. You can build robustness by using diversification, preserving optionality, avoiding leverage, and staying away from negative expected situations. Taleb, for instance, doesn't smoke or use bicycles. Because all fat-tailed risks add up, one ideally aspires to zero harmful tail risks. Fragile gets hurt, robust resists, and the best is antifragile because it benefits from extreme events. Antifragile loves volatility, randomness, disorder, and stressors. For instance, Wolff's law states that if loading on a bone increases, the bone won't break but will remodel itself over time to become stronger to resist that sort of loading. Nature is an excellent example of antifragility. 

The best situation is to have an asymmetry of risk in which the favorable consequences are much larger than unfavorable ones. I understood the depth of the concept of margin-of-safety, dear to value investors, after reading Taleb. Investment wise, by buying cheap, investors reduce their downside risk while increasing the upside. Often benefiting from Black swan's events that crushed the market, investors can profit from low-risk and high-return. I also realize why the best investors are so conservative. They stick to their circle of competence, thus reducing the zone of uncertainty. They often rely on time-tested principles because time is the best stressor. If an idea, a technology, an object, or a company withstands the effect of time, it means it's not fragile to extreme events. 

If one evolves in a fat-tailed environment, it’s better to be prudent and conservative while preserving the resources to be hyper-aggressive after a Black Swan hits. It's OK if work doesn't deliver steady results because it's the variance of the outcome that matters, not the frequency. One can be wrong every day if the cost is low, so long as one makes large gains when one is right. It is particularly striking with creative professions such as artists, entrepreneurs, writers. Payoffs are infrequent but massive. As it is typical in Extremistan, winners often take it all. Obviously, what we don't see is ruined creatives who lost a lot by being on the wrong side of the distribution. Taleb argues for the creation of an Entrepreneur Day to honor ruined entrepreneurs. He writes: "Most of you will fail, disrespected, impoverished, but we are grateful for the risks you are taking and the sacrifices you are making for the sake of the economic growth of the planet and pulling others out of poverty. You are at the source of our antifragility. Our nation thanks you." When one fails, one makes the overall system more resilient. 

Most organizations, such as big bureaucracies, are fragile. When centralizing functions and growing in size, the fat-tailed risk is scaled too. Most organizations eliminate redundancies to streamline processes over time. It, however, makes them more fragile to stressors. Redundancies are inefficient in Mediocristan but highly efficient in Extremistan because it contributes to robustness and antifragility. Building a thriving life or organization in Extremistan is highly counter-intuitive but very necessary. That’s what makes these concepts so fascinating!


Anonymous survey: What would have made this article more helpful?

If you found this article valuable, please consider sharing it 🙌