Overconfidence or (Ir-)Rationality (April 2018)

Many of us make their daily living by seeking to make accurate predictions or judgements under uncertainty, and to make them with confidence vis-à-vis our respective clients.  What good is an evasive lawyer when asked for an opinion, what good a fidgety banker asked to provide a fairness opinion or deal-related judgment, what good a doctor providing five potential hypotheses when asked to diagnose a potential disease, what good a stock picker that cannot make us his mind on Buy or Sell?  Nobody would say that confidence beats judgement, but in our professions, by and large, professional success is not just a function of the quality of judgement, but also of the conviction with which it is delivered (and, over a longer period time of course, the track record of good judgement calls vastly outnumbering bad ones; hindsight being the best advisor of all).  If you cannot give a straight and confident answer, best to stay in academia!  Or politics, where posterior probabilities or appraisal of a statement/policy by voters/pollsters will more often than not determine an answer – unless you are called Macron.

Most of us have also been trained in rational, structured thinking, either by training to be lawyers, or economists, or doctors.  Not many of us have the benefit of being mathematicians or physicists, worlds in which outcomes can mostly be calculated without the shadow of much doubt.  We have therefore come to rely on our rationality as the ordering system for a highly confusing, variable, changing and unpredictable world.  And our human need for certainty, for assuredness, for comfort, for order, has given us the confidence to rely on our rational faculties to resist the call of chaos, the ubiquity of hubris, the unavoidability of surprises, the randomness of outcomes. 

Until the stock market fell 10% earlier this year, most market participants predicted continued stability, even if the markets are supposed to be a random walk. 

Until Trump got elected, not one serious analyst would have predicted that he might win.  Nobody can quite make out why he has not yet been impeached, either.

Until the Brexit referendum went the way of the leavers, not many punters would have bet that the famously pragmatic, European people of the British Isles, age-old explorers and world conquerors, creators of the Commonwealth, would vote to seek to lift anchor and sail off into the unknown because of one too many Polish plumbers buying Polish beer in the local corner shop run by a Pakistani. 

Until Kaczynski won presidential and parliamentary elections in the most successful post-communist country in central Europe in 2015, which by that stage had experienced 20 years of growth, only being surpassed by Australia in that indicator of success, nobody would have thought it possible that the proud and freedom-loving Polish people that rebelled and defeated Soviet hegemony and totalitarianism, and who propelled democracy in all of the satellite states of the Warsaw Pact countries, indirectly demolished the Berlin Wall and paved the way for German unification, would provide such strong support to a regime that has repeatedly violated the constitution, neutered public media, seeks to rewrite history and inscribe certain historic interpretations into law as truisms, is picking fights with nearly all of Poland’s large neighbours, and plainly has no long-term concept that will ensure sustainable Polish freedom and independence in a secure international setting. 

So maybe I am not alone in trying to understand how so many good and sensible people have been led to make so many questionable judgements, which in most cases will not be aiding their prior concerns (such as the underemployed blue collar workers in the north of England, the localized victims in American rust-belt towns such as Janesville (Wisconsin), as described in ‘Janesville:  An American Story ‘ (Amy Goldstein, 2017), or the Polish people who would like old certainties to hold – such as the Russians are to blame for the Smolensk plane crash rather than human error under faulty supervision, to name a few examples).

Well – you may be as relieved as I am that there is at least a good start of an explanation why we can err so widely, so systematically, and (unfortunately) so predictably in coming to errors of judgement under uncertainty, notwithstanding our training and our confidence in our predictive capabilities.

The book had been lying on my shelf for more than half a decade, too many deals taking priority over gaining some wisdom to question some judgements, but when one of the most respected contemporary US writers wrote a loving and brilliant biography of the academic friendship that begot that theory, Michael Lewis in ‘The Undoing Project’ (2017), and that author was immediately recognizable to those of us in the financial markets that laughed about his ‘Liar’s Poker’ (1989) and admired the balls and analytical strength of the very few people betting in 2007 against Subprime as retold in ‘The Big Short’ (2010), when the herd was ‘still dancing’ in the words of the soon-to-be-fired Citibank CEO before that bank lost billions of USD --  well then the penny dropped even for me. 

Economists are said to be a nerdy, arrogant and highly confident tribe, and don’t look kindly onto outsiders, most of which they don’t consider being sufficiently good at maths, for a start.  So not many non-economists have won the Noble Prize in Economics, and certainty not many academics that don’t accept the primacy of mathematics over intuition and creativity.  Well, in 2002 that all changed when Daniel Kahneman received the Nobel Prize in economics for the work that I have found fascinating and enlightening over the last couple of weeks and months in seeking to throw some light on our errors of judgement, and affording some explanation as to why we might be getting it wrong so much in recent months and years.  Daniel is a psychologist, and one of his outstanding characteristics is that he is never sure of himself nor his opinions.  How refreshing!

Source: grawemeyer.org

A survivor of the Holocaust in his native France, by the age of 21, having taught himself psychology for two years at Hebrew University in Jerusalem, he was thrown in the deep end by becoming head of the young Israeli army’s conscript vetting and approval programme.  He was asked to devise a method to appraise which conscript was best suited to which army unit, who was ‘officer material’ and who was not.  And the survival of Israel might depend on it to a not insufficient degree.

He noted that the interviewers of conscripts were exceedingly confident in their judgement.  Tall strapping lads must make good officers, to overstate the point.  Kahneman did not trust any of the subjective judgements involved in that, and devised an algorithm based on questions he devised.  He took subjective judgement out of the process (and the recruiters did not like it one bit), and let his algorithm decide who was getting which posting and unit.  The Israeli Army uses his system to this day (and was often asked by US Army personnel how it is that they have the same equipment as their US army colleagues but have never lost a war).

He also noted a feedback process that involved chastising pilots who had flown bad sorties (negative feedback enticing an improvement in performance in the next sortie), and not providing positive feedback to pilots who had done well, since they were anyhow likely to perform worse on the next sortie.  The lesson drawn by the Israeli Air Force had been that it was better to punish bad performance and not to laud good performance (a generalized pattern in humanity).  Kahneman, however, recognized a simple statistical truism – what economists call ‘reversion to the mean’.  The quality of piloting averages out over time, and thus good sorties are followed by bad ones and vice versa – again the illusion of the superiors in thinking their feedback altered the course of future performance of their pilots belied overconfidence and ignorance of some basic statistical trends.  Kahneman concludes that ‘it is part of the human condition that we are statistically punished for rewarding others and rewarded for punishing them’ (cited in Michael Smith p.126).

Kahneman and Tversky, his brilliant co-author throughout 12 years of productive questioning of established beliefs and theories, also successfully challenged Utility Theory, on which most of modern economics is based.  They showed that people have a strong affinity to the status quo, that ‘regret’ is a key motivation in economic affairs, and that it is not sufficient to look at the level of someone’s well-being but at the change in such well-being as the determinant of preferences and utility.  It gets complicated but rewarding (see further reading).

If you think this is all just academic and without bearing on real life, consider the examples below.  An intuitive application of their analyses are the concepts of (i) insurance and (ii) lottery.  This is best illustrated by looking at the 2x2 matrix below.

Low probability gain (“the lottery”) or loss (‘insurance’)

The bottom left corner explains why people bet on potentially large wins with very small probabilities (lottery effect).  Although they know that their chances of winning are miniscule, and that ‘the bank always wins’ (which is why most lottery systems are national monopolies), people still go out every week and buy lottery tickets from their meagre wages.  It is not rational, either on an individual level or across groups, yet the hope of winning (plus overconfidence of the type explained above and ‘bad intuitive statistics’) lets people make systematically irrational choices with monetary losses.  Not exactly what standard rational economic theory would predict …

In the bottom right corner, the risk aversity to a large loss, even if it is very unlikely to occur, lets people go and take out insurance.  Of course most insurance companies make money most years, proof that here also, ‘the bank always wins’.  But their insipient ‘theory of regret’, which builds on human’s intrinsic instinct to protect their status quo and the strong psychological urge to avoid losses and negative surprises, explains why we each individually and as society systematically purchase too much insurance against very low-probability events.

I did not buy any of it … until I took some of their tests, and could not help pinch myself when I got it wrong many times, despite having been warned about the biases that play tricks with my heuristics …

You have to try it out, I append one such test, there are many more on the internet, it will open your eyes about YOUR OWN systematic wrong thinking.  (They tried these on statistics professors and they got it wrong as a cohort, so don’t feel too bad about yourself 😉).

https://www.vanityfair.com/news/2011/12/kahneman-quiz-201112

Thanks to Daniel and Amos, a whole new branch of economics has developed, called Behavioural Economics, also generating a Noble Prize (Richard Thaler, 2017) – see further reading for some introductions.

For those of you with too little time, my fellow Kennedy School of Government graduate student, former BBC economics editor, FT columnist and nowadays head of Bloomberg Economics, Stephanie Flanders, interviewed Michael Lewis in November 2017 about his book; the 1.5-hour discussion on Youtube is a great stimulant, too:

https://www.youtube.com/watch?v=nbblhD8K4WA

And you can get a glimpse of the very special relationship between Daniel and Amos from this interview with Daniel:

https://www.youtube.com/watch?v=w20FiXzzyOM

Further reading for those intrigued:

1.       Daniel Kahneman:  Thinking Fast and Slow (2011)

2.       Michael Lewis:  The Undoing Project (2017)

3.       Judgment under Uncertainty:  Heuristics and Biases, by Amos Tversky and Daniel Kahneman (Science 1974)

4.       Prospect Theory:  An Analysis of Decision under Risk, by Daniel Kahneman and Amos Tversky (Econometrica 1979)

5.       The Framing of Decisions and the Psychology of Choice, by Amos Tversky and Daniel Kahneman (Science 1981)

6.       Nudge.  Improving decisions, about health, wealth and happiness, by Richard H. Thaler & Cass R. Sunstein (2008)

7.       Predictably Irrational.  The Hidden Forces That Shape Our Decisions, by Dan Ariely (2008)

8.       Misbehaving.  The Making of Behavioural Economics, by Richard H. Thaler (2015)

*****