Book Notes: Fooled by Randomness by Nassim Nicholas Taleb

Book Notes: Fooled by Randomness by Nassim Nicholas Taleb

Fooled_by_Randomness_Paperback

Maxims

  • Embrace novelty, unpredictability, and randomness in your life
  • Know what you don’t know, and don’t be afraid to admit it

 

Notes

Rare events shape more and more of the world around us, but our minds cannot grasp them.

It’s almost like there are two planets–one in which we live, and one in which people think we both live in, that is considerably more deterministic.

Past events will always look less random than they were (hindsight bias).

People seem to have followed Descartes’s method of formal thinking. Reality is more of the Renée Montagne variety—open and vague, requiring insecure, introspective thinking.

It takes extreme courage to be introspective, and to accept one’s own limitations.

Probability is a branch of applied skepticism, not an engineered approach.

Recognizing that the world could have been different is the core idea behind probability.

Many people tend to mistake causality and necessity. For example, all millionaires may be hard workers, but not all hard workers are millionaires.

It doesn’t matter how often people succeed if failure is too costly to bear.

Solon’s warning to Croesus—good fortune gained through randomness will not last.

Nero and John.  Nero is a proprietary trader does not make risky trades or participate in naked options.  He lives within his means, and spends much of his time as he chooses.

John is a high-yield bond trader. He and his wife live in ostentatious lifestyle, with a large house, luxury cars, and exotic vacations.

John makes multiples of what Nero makes, but he is surprised when the market crashes and he loses everything (and then some).

Alternative histories: $10 million earned by being a dentist is not the same as $10 million earned by playing Russian roulette. The outcomes that did not transpire must be considered just as much as the outcome that actually transpired.  The world must be viewed as a scientist (i.e. Liebnitz’s parallel worlds), not as an accountant.

The author has had two managers in the author’s life; one conservative family man who rewarded trades based on their outcomes (regardless of randomness), and one flamboyant Frenchman who thought about risk and probability constantly—even making his employees plan for a plane crashing into the building.

The corporate risk managers job is to minimize the risk of a blowout.  This is a difficult task when the roulette wheel is invisible, so the solution is for the manager to make vague recommendations on minimizing risk (but not all risk, for fear of losing his job).

Looking back, the past always looks deterministic. This is because only one of the infinite possibilities or realities actually occurred.  A mistake is not something to be determined after the fact, but only with the information available at the time.

An idea’s ability to withstand the test of time is correlated to its value and relative fitness.  Noise gets filtered out over time.  Journalists are paid to grab your attention, and therefore reading the news is a futile way to gain meaningful knowledge.

Over a short time period, one observes noise in a portfolio, not the actual returns.  As you increase the time period between observations, the signal-to-noise ratio increases.  Furthermore, losses are perceived as being more painful than gains by up to 2.5 times.

The cross sectional problem: at any given time in the markets, the richest trader is often the most foolish.

Evolution implies fitness to only one time series, not all of the potential ones.

The more fit to a given time series that a species becomes, the more susceptible it becomes to rare, unpredictable events. The short-term success that it experiences will be magnified by the hormonal impact of additional serotonin in the bloodstream, making it appear to have a superior genetic makeup.  This same evolutionary effect is also observed for successful traders that make risky bets and experience big short-term payouts.

The problem of induction–i.e. all swans I have seen are white, therefore all swans are white–is taking many particular pieces of information and forming generalizations from the synthesis of them.

Survivorship bias: in a random environment, the successful cases will always be highlighted, however, the failures will often be forgotten. This applies to the financial markets, book reviews, and many other Arenas.

Adverse selection bias: buyers and sellers have access to different levels of information (asymmetric information)–e.g. funds that advertise as being successful only advertise because they have been successful.

The birthday paradox: things that seem like coincidence are often just a poor understanding of statistics and probabilities.  The birthday paradox is a great example–the probability of sharing a birthday with a random person is 1/365.25, but the probability of two people sharing a birthday in a group is significantly higher.

Data mining fallacy: people will seemingly find meaning in random data just by searching for it, provided there is sufficient amounts of data (e.g. the Bible Code).  Humans have the tendency to see patterns and try to explain them using the benefit of hindsight.

Reference case problem: there is no such thing as a truly random process in practice.  Pure randomness derives from the precise mechanics of the process it belongs to (e.g. the balance of the roulette wheel and the eccentricities of the ball).

A path dependent outcome is one where the result is decided on previous events, rather than randomness.  Examples include the QWERTY keyboard and acting roles.

The information age has accentuated the effect of the majority of results being transferred to one or a small handful of parties (winner-take-all situations).  This is known as a Pareto or power law relationship.

People misjudge probabilities due to lack of considering the conditions upon which a probability should be judged. This is called conditional probability.

The brain is capable of distinguishing signal from noise, but the heart is not.

Wittgenstein’s ruler: the measurement of an item with a ruler can say as much about the ruler as the item being measured if the credibility of the ruler is in question. This applies to critiques and book reviews, among other things.

Mathematics is a tool upon which to meditate, not to compute, in the realm of probabilities.

Attribution bias: people attribute their successes to skill, and their failures to randomness. This explains why most people think they are above-average drivers, for example.

Focus on (and reward) the process, not the results, in areas characterized by randomness.

Satisficing: the intersection of satisfying and maximizing.  In contrast to optimizing, which requires a mindset of insatiability and constantly chasing better and nicer things.  More money should lead to more freedom, not becoming a slave to obtaining more of it.

Some degree of randomness in our lives saves us from being optimizers of the wrong things, and instead requires one to ask what is most effective to focus on.
Satisficing helps us avoid the internal treadmill effect of trying to maximize every experience.

 

My Thoughts

Fooled by Randomness presents a new lens through which to look at the world.  It discusses not only how to think about randomness, but also how to live with it and how to prevent it from destroying you.  I appreciated the anecdotes, literary and historical references, and subtle wry humor spread throughout.  The book is especially helpful for those who work in industries with high levels of risk and unpredictability, such as finance and oil & gas.

 

Review: 95/100

Advertisements

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

w

Connecting to %s