The Pretence of Knowledge

This post is based on Friedrich August von Hayek's lecture titled 'The Pretence of Knowledge' to the memory of Alfred Nobel, on December 11, 1974. Adapted quotations may be used without quotation marks.

Key Learnings

Investing, like economics, is essentially complex. This is because outcomes do not depend only on the relative frequency of individual actions or occurrences, but also on the manner in which the individual actions are connected with each other. For this reason we cannot replace information about every individual actor (human being) with statistical information, and require full information about each actor if we wish to derive specific predictions about individual events. 

All of the particular information possessed by every one of the participants in the market drives the value of assets (or goods). This means the price is determined by a sum of facts which in their totality cannot be known to the scientific observer, or to any other single brain. This is the source of the superiority of the market order, and the reason why free markets are the most efficient allocators of resources using information which only exists in dispersed form. 

Our inability to assimilate all of this data means that investing and economics must not be treated as Physical Sciences. In fact, Hayek says that a scientistic attitude is decidedly unscientific in the true sense of the word, since it involves a mechanical and uncritical application of habits of thought to fields different from those in which they have been formed. As mentioned above, we do not have the ability to access (or even measure) all of the necessary data. This is why economists and investors with scientistic attitudes frequently treat the data which happens to be accessible to measurement as important that – not necessarily the best data. We know, for example, that good quality management is essential for a good business to grow sustainably. However, there are shades of grey in measuring the quality of management. This is why a purely numbers-based/scientific approach to investing would only take measurable quantities into account, proceeding on the fiction that the factors which can be measured are the only ones that are relevant.

Without access to all the necessary information, we are confined to making directional predictions – predictions of general attributes, but not containing specific statements. It is possible to say, for example, that a company will be worth substantially more over a long time horizon; however, we must not – or more precisely, cannot – predict the exact value and the time at which this value will be attained.

Implications for Investing

Hayek's paper does not criticise the use of numerical evidence to help put the magnitude and implications of forecasts into a context. However, he cautions against the dangers posed by the false sense of comfort that numbers can give us, and critically aware of the arrogance that (successful) precise predictions can engender. This is why he refers to the "danger [posed] by the exuberant feeling of ever growing power which the advance of the physical sciences has engendered and which tempts man to try, 'dizzy with success', to use a characteristic phrase of early communism, to subject not only our natural but also our human environment to the control of a human will." His speech culminates with a reflection on the importance of humility when dealing with essentially complex phenomena: "The recognition of the insuperable limits to his knowledge ought indeed to teach the student of society a lesson of humility which should guard him against becoming an accomplice in men’s fatal striving to control society – a striving which makes him not only a tyrant over his fellows, but which may well make him the destroyer of a civilisation which no brain has designed but which has grown from the free efforts of millions of individuals." All of this means there are a few key things we must remember as investors.

• The market is typically efficient because it is the sum of all of our viewpoints. This is why it is important for investors to ask themselves: Who doesn't know that?
• Trying to be overly scientific, particularly with insufficient data and knowledge, is dangerous.
• Rely on first principles thinking as opposed to blindly using correlations. These correlations are directional – not deterministic – and only work when the ceteris paribus condition holds, till an unexpected event leads people to stop believing in their determinism. In the real world, there are many variables at play, and simple relationships are only useful oversimplifications.
• A great many (important) facts cannot be measured and so, are disregarded. The idea that only facts which can be measured are relevant, is fantastical.
• You must have an understanding of the Narrative and Numbers of the business.
• We must use numbers directionally (as a reality check), and try to avoid being seduced by the false sense of comfort that numbers may offer. 

Is this changing?

To an extent. We are becoming more able to collect data. Not only do firms such as Facebook and Google target advertisements based on our online interactions and search history, but Amazon has a record of our purchases, Netflix a record of what we watch, and Spotify a record of what we listen to. Investors can (or might eventually be able to) analyse traffic in parking lots outside malls, the weight of bags being carried by consumers, lexical choices in transcript, and even the expressions and tones of voice of CEOs to determine their confidence and honesty levels. This is both, good and bad; we will have more data to test hypotheses, but we will also risk becoming more comfortable and more arrogant as we make misleadingly precise predictions.

Whistle while you work (The Economist)

Great article from The Economist about the importance of having happy employees for business performance. Go read my post on Good Corporations Benefit ALL Stakeholders!!!

__________________________________

Article:

Research suggests that happy employees are good for firms and investors. 

THERE IS an old joke about a new arrival in Hell, who is given the choice by Satan of two different working environments. In the first, frazzled workers shovel huge piles of coal into a fiery furnace. In the second, a group of workers stand, waist-deep in sewage, sipping cups of tea. The condemned man opts, on balance, for the second room. As soon as the door closes, the foreman shouts “Right lads, tea break over. Time to stand on your heads again.” 

Terrible working conditions have a long tradition. Early industry was marked by its dirty, dangerous factories (dark, satanic mills) and in the early 20th century, workers were forced into dull, repetitive tasks by the needs of the production line. However, in a service-based economy, it makes sense that focusing on worker morale might be a much more fruitful approach. 

Proving the thesis is more difficult. But that is the aim of a new study which examines the relationship between happiness and productivity for workers at British Telecom. Three academics—Clement Bellet of Erasmus University, Rotterdam, Jan-Emmanuel de Neve of the Saïd Business School, Oxford, and George Ward of MIT—surveyed 1,800 sales workers at 11 British call centres. All each employee had to do was to click on a simple emoji each week to indicate their state of happiness. Those workers were charged with selling customers broadband, telephone and television deals. In total, the authors had adequate responses from 1,161 people over a six-month period. 

The results were striking. Workers made 13% more sales in weeks when they were happy than when they were unhappy. This was not because they were working longer hours; in happy weeks, they made more calls per hour and were more efficient at converting those calls into sales. The tricky part, however, is determining the direction of causation. Workers may be happier when they are selling more because they anticipate a bigger bonus, or because successful sales pitches are less stressful to make than unsuccessful ones. 

The academics tried an ingenious way to get round this causation problem by examining a very British issue—the weather. Workers turned out to be less happy on days when the weather in their local area was bad and this unhappiness converted into lower sales. Since they were making national calls, not local ones, it is unlikely that customer unhappiness with the weather was driving the sales numbers. So it was worker mood driving sales, not the other way round. 

Even if this reasoning proves to be correct, businesses may struggle to find it of comfort. Short of locating all their call centres in California or Hawaii, companies cannot control the weather conditions their workers face. The academics point out that “what we are not able to do, given our data and setting, is adjudicate as to whether investing in schemes to enhance employee happiness makes good business sense”. It is possible that the costs of such schemes might outweigh any gains in productivity. 

More research is clearly needed. But there is evidence that happier workers are good news for shareholders, as well as productivity. Analysts at BofA Merrill Lynch Global Research studied the stocks of firms rated on Glassdoor, a website which allows employees to rate the companies they work for. Those with the highest ratings outperformed those with the lowest by nearly five percentage points a year between 2013 and 2019. The analysts also used software that picked over the text of employee reviews and found that incorporating this approach improved the risk-reward trade-off (as measured by the Sharpe ratio) of the strategy. 

The analysts have now applied the same approach to picking stocks based on particular sectors. Again, the sectors where workers gave the best reviews on Glassdoor over the 2013-2019 period easily outperformed those where employees gave a thumbs down. None of this is unequivocal proof. The history of equity investing is littered with strategies that worked well when back-tested but then disintegrated when applied in the real world. But at the very least, it suggests that companies should consider the merits of a contented workforce. And that might mean giving them harps and ambrosia, rather than devilish treatment.

Superforecasting in Investment Decisions

This post is the second of two based on Shane Parrish's Knowledge Project interview with Philip Tetlock, co-principal investigator of The Good Judgement Project, author of 'Superforecasting: The Art and Science of Prediction', and a professor at the Wharton School.

All investment decisions are based on present expectations of the weighted average of the various futures that could unfold with time. That's a mouthful, but the essence of the issue is that in order to arrive at a future value for a company, we must make assumptions about multiple factors, each of which can materially alter the 'intrinsic' (expected) value of a business. So, investors deal in forecasts. In this post I will discuss briefly three ways in which the quality of our forecasts can be improved, and suggest applications for investment committees. 

Guesstimating
A good place to begin forecasting is guesstimating. Making guesstimates allows us to identify gaps in our knowledge, clearing up the zone of ignorance. Making guesstimates also enables us to review our assumptions when evaluating the quality of a forecast.

Forecasting teams
Some investors swear by a committee-based approach to asset allocation. The idea is this will result in ideas being challenged and hypotheses being questioned, reducing the impact of biases and the potential for errors. There is some merit to this approach, but often organisational hierarchy and homogenous information limit its benefits. Forecasting teams can be made more effective if individuals' success rates are tracked, and a weighted average of forecasts, which gives more weight to individuals with stronger track records in the relevant industry, is used. Forecasting teams can be made even more effective if they have access to distinct pools of information because decisions made when different pools of data point in the same direction, are likely to be more accurate. In the investing world, this is very difficult to achieve, but could still be partially achieved by splitting sources (annual reports, sell-side reports, company management meetings, and data analytics) encouraging team members to research each company independently and arrive at a conclusion before a discussion.

History rarely repeats itself...
...but it does rhyme. Mark Twain's eloquent quotation is often used to emphasise the importance of using past data and patterns in investment decision making. Tetlock's work cautions against over-learning patterns in history. Perhaps we should pay equal attention to both parts of Twain's wise words, as we continue to rely on the only truly concrete information available to us.

Superforecasting

This post is based on Shane Parrish's Knowledge Project interview with Philip Tetlock, co-principal investigator of The Good Judgement Project, author of 'Superforecasting: The Art and Science of Prediction', and a professor at the Wharton School.

7 Billion Forecasters
We make implicit forecasts on a daily basis. Whenever we choose to cross the road, whenever we buy a good, whenever we say something, we make an implicit forecast; we forecast the likelihood of being hit by a vehicle as we cross the road, we forecast our future use of the good and ascribe a present value to it, and we forecast the response we expect to elicit.

Implicit forecasts are convenient and serve us well when we are crossing roads, buying certain goods, and talking to people. Of course, you wouldn't want to spend hours by the roadside calculating the probability of death should you choose to cross the road at a certain moment. However, implicit forecasts are less useful when we are dealing with complex, nuanced situations (or making important investment decisions). 

Explicit Forecasts
Implicit forecasts made with words are vague and can't be falsified. Explicit forecasts (with specific probabilities attached), by contrast, are easier to fault. This is why they enable us to hold ourselves accountable and evaluate past investment decisions. This evaluation is important because both luck and skill are involved in forecasting, and it is important to be able to consider alternative histories and attempt to distinguish lucky winners from good decisions, and unlucky losses from mistakes.

Good Forecasting
Forecasting is roughly 70% skill and 30% luck according to Tetlock's research. And, good forecasters are made not born. Good forecasters have specific knowledge in their field, are open-minded, and believe that it is worth estimating probabilities of real world events. We can make good forecasts by reading about our subject (in the case of investing, this may be the relevant company), and cultivate the necessary qualities for consistent outperformance by encouraging debate in our organisations and adding probabilistic thinking to our decision-making process. Most importantly, by second-guessing successes and evaluating failures, we can become better forecasters.