top of page

Prediction Machines

Updated: Jul 28, 2022

Chimp Investor is ranked #3 of 90 by Blog Overview and #10 of 100 by FeedSpot in their relevant categories. If you enjoy reading this blog, please post a review on WealthTender.

It's easy to make predictions but making them useful is another thing all together

All living things make predictions by filtering information from the world around them. We humans constantly make predictions, whether in relation to how far you need to move your finger to scratch your nose, or avoiding getting hit by a car.

We are prediction machines, and often find shortcuts. How often do you find yourself predicting when you can cross a busy road, not by waiting for the little green man but by watching whether the traffic light has turned orange? Or, if the lights are not visible, whether traffic is slowing down in response to an an orange light? And, yes, we also know that a car speeding up may be the first indication of the light having just turned from green to orange.

We predict continuously, without thinking, and in clever ways.

Making predictions is at the heart of the investment business. Active fund managers make predictions about future prices of markets, stocks etc, whether directly - technical analysis - or indirectly - fundamental analysis. If you as an individual investor buy a passive fund, you are predicting, whether you know it or not, that it will perform in a certain way.

Incidentally, in relation to this last point, I have been writing for a number of years now about how the performance of passive balanced funds was likely to deteriorate as a result of bonds, one of the two major components, having become very expensive.

Rising inflation over the last 15 or so months has just made things worse for balanced funds. It has pushed up yields of equities and bonds, causing nominal prices to fall, but it has also caused these poor nominal returns to be even worse in real terms, a doubly whammy. If inflation stays high or rises further in the coming years, this poor performance will continue.

But I digress.

Some predictions are as good as useless. I'm not going to get much kudos predicting that the sun is going to rise tomorrow. And predicting the opposite isn't much use either - in the unlikely event that I'm right, kudos will be in short supply. Making predictions where probabilities are effectively either 0 or 100pct is churlish - you need to stick to the middle ground where there is uncertainty.

Uncertainty alone however is not sufficient to endow a prediction with utility. I'm not going to make any money over time predicting which side an unbiased coin will land. Why? Because the odds are 50pct, will always be 50pct, and everyone knows they are 50pct. To make money from a prediction, you need uncertainty in combination with a sound reason to think the odds are wrong, an edge.

In skill betting and investment markets, market odds are expressed either explicitly in terms of odds - for example, four to one, equivalent to a 20pct probability - or implicitly as prices - the current price of Unilever tells you what is generally expected about its future. Indeed, it has been mathematically proven that to maximise your winnings over time, your bet/position size as a percentage of your purse/portfolio should be equal to edge divided by odds, edge being a measure of the difference between the market's probability of some outcome and your own probability of it.

An edge can either be some sort of specialist expertise/ability or it can be inside information. Examples of the former might be an algorithm written by maths geniuses and run on a very fast computer, or the ability to accurately appraise emotions that drive markets rather than be subject to them. As for using inside information, your return might include a spell at Wandsworth nick.

Predictions to which nobody listens may be at worst useless, at best wasted. According to UCLA professor Albert Mehrabian, only 7pct of meaning is communicated through what you say. Most is through how you say it - tone of voice, body language, etc. A fund manager's or research analyst's predictions are only useful to their employer if they are helping to generate revenue. Predictions that are not supported with credible evidence or intellectual support also tend to receive short shrift from people, conspiracy theorists excepted.

As someone whose job it has been, and is, to make predictions, I feel I have a responsibility, both to myself and to my audience, to review past predictions as well as to repeat current ones. There is little accountability in the investment industry in relation to making predictions; the term "two cent's worth" is a reference to an opinion not being worth very much. The financial media knows that viewers and readers are far more interested in what people think will happen not what they thought would happen.

Regardless, here is a summary of my current predictions, all of which have been set out in previous blog posts - clarity of communication is important. In a later post I will also appraise my past predictions - accountability and humility are important too.

Over the next couple of years I think inflation will fall but then rise again and remain high for several years. Inflation will fall in the medium term as a result of a recession that is caused by real incomes squeezed by high inflation no longer being able to support economies. In other words, as others have argued, the best cure for high prices is...high prices. Increased interest rates will also dampen demand.

Some newsletters have pointed out that if there is going to be a recession, it will be the most widely anticipated one ever. The implication of this is that since a recession has not happened yet, predictions of it must be wrong. I'm not sure I entirely agree with this. In 2007 it was very apparent that difficulties in financial markets - subprime, etc. - could well deteriorate further and spill into the real economy. They did. A year and a bit later.

It also presupposes that on the economic front things happen linearly i.e. that if things are deteriorating only slightly, as they currently are, they won't get too bad and at some point will reverse.

That isn't always the case. Economies are examples of complex non-linear adaptive systems. They hit tipping points at which things that have been changing only slowly suddenly deteriorate quickly. Consumer spending on non-essential items in many countries has started to fall, but only gradually. It is not hard to see that at some point credit cards get maxed out, consumer spending starts to fall faster, prompting wider fears among both consumers and companies, and further spiralling downwards. A tipping point, and a recession.

While it would no doubt be painful for many, a recession would be no bad thing, as the lower demand would help to lower inflation. Indeed, a recession from a purely economic perspective should be seen as a treatment for high inflation, which itself is a symptom of some underlying problem or sickness - Zimbabwe, Turkey, Argentina etc are examples of extreme sickness.

A recession however only treats the symptoms, it does not necessarily cure the underlying disease. There are reasons to think that there are now underlying problems in the world economy such that following a recession and a period of falling inflation, inflation will rise again and remain persistently high for several years. Indeed, this is what I expect.

The current high inflation is undoubtedly the result of covid- and Ukraine-related supply issues, but also of excessive fiscal policy in the US having pushed up prices everywhere - we live in an interconnected world. Furthermore, the effect of companies seeking to make supply chains more secure by repatriating manufacturing from lower cost emerging economies is to push prices higher still. We have also started to see so-called second round effects, namely accelerating wages.

Moreover, the Russian invasion of Ukraine may have heralded a new era of geopolitical risk which could see persistently high/rising food, energy and metals prices feeding through to consumer prices more broadly. I also fear that central banks' bloated balance sheets render them less able to manage inflation as they have done quite well over the last forty years.

My big prediction in recent years was that safe haven bonds would prove to be unsafe. Perhaps I was a little early, given that the early stages of the pandemic saw them perform well. However, this was short lived, and safe haven bond prices in real terms are now lower than they were before the pandemic and indeed over most periods to date during the last ten years.

The good news is that for investors whose portfolios are still significantly exposed to so-called safe haven bonds, a recession in the next year or two would see them perform well. This however should be considered an opportunity to sell them - in other words, a second bite at the cherry, not cause for celebration.

The views expressed in this communication are those of Peter Elston at the time of writing and are subject to change without notice. They do not constitute investment advice and whilst all reasonable efforts have been used to ensure the accuracy of the information contained in this communication, the reliability, completeness or accuracy of the content cannot be guaranteed. This communication provides information for professional use only and should not be relied upon by retail investors as the sole basis for investment.

© Chimp Investor Ltd

48 views0 comments

Related Posts

See All

Bình luận

bottom of page