A CONKERS3 FEATURE : The Analytical Surveyor
Analytical Surveyor has worked for over 40 years in the City and West End of London as a real estate investment manager, an equity analyst, a multi-manager, a fund manager and as a researcher and strategist. In his area of expertise, he is highly regarded and currently works as a consultant.
There is an enormous amount of academic research that provides insights into the functioning of investment markets. Rarely will this help with stock selection, but much of it will help in understand how and why the market functions the way it does. The Analytical Surveyor trawls through the work that is available in the public domain and summarises the work into usable notes for investors.
There’s never a recession coming:
Compounding is wonderful. It has many admirers, including Albert Einstein, who uttered the famous words “Compound interest is the eighth wonder of the world”.
The assumption is that compounding is something positive for investors, and the essential ingredients are a good long-term asset and enough time to become the long-term. Hence investors are typically encouraged to ‘stay in’ and avoid trying to time the market. Much is made of analyses that show that if you were historically out of the market for just a limited number of the ‘best days’, then your performance would have been dramatically poorer than if you had remained invested for the whole period. Rather less emphasis is placed on the mirror-image argument that being out of the market for some of the worst-performing days would have dramatically enhanced your performance.
And that is the problem. Good steady compounding of positive growth certainly leads to wealth generation, but it only takes the occasional big set-back to completely destroy the accumulation. Just consider where the FTSE100 has gone over, say the last 10 years, and it is clear that its phenomenal growth since the 2008/9 recession has only really repaired the damage caused by the recession itself.
It is therefore hardly surprising that investors are wary of such downturns, even if the effects are not permanent (i.e are only cyclical) and rush for the exits when one is apparently imminent. The received wisdom is that the markets anticipate such downturns about a year ahead and, if that is any guide, 2019 may be the year of a European/US recession. Increasingly, the markets are becoming very nervous, and some commentators are arguing that the next recession may come in the next six months.
Economists, however, struggle to provide any support for that possibility. They point to generally good economic growth, albeit one or two countries are struggling, high levels of employment, contained inflation, and low costs of capital (including debt). As a result, it is difficult to find any economist, even from the independent houses, producing such a forecast.
It is therefore worth turning to the work done by Zidong An, João Tovar Jalles, and Prakash Loungani[1], in response to the question ‘How Well Do Economists Forecast Recessions?’ As they point out in the introduction: recessions are not rare: economies are in a state of recession 10-12 percent of the time. What is rare is a recession that is forecast in advance.
To make the point, they calculate the average forecasts for 63 countries from 1992 to 2014 for GDP growth made in April of the year before the recession, in October of the same year, and in April of the year of recession. The first set of forecasts point to 3% GDP growth (strong growth indeed), the second to 2% while, by April of the year of recession, it is about -0.8%, and by October about-2.5%, but still underestimating the outcome of -2.8%.
To obtain a more comprehensive understanding of what happens in the forecasting processes, the data is analysed at a country and time level. In total, 148 of 153 recessions are missed in the prior April, and this declines over the subsequent months. But even by October in the year of the recession, 35 recessions are missed, a truly grave failure.
A comparison of the performance pre- and post- 2008/9 recession indicates that a larger proportion of recessions were successfully forecasted and the mean forecast error was smaller (except for the year-ahead forecast for emerging economies) in the post period than in the pre period.
In addition to identifying type 1 errors (a recession happened but was not forecast), type 2 errors (a recession did not happen but was falsely forecasted) are considered. The number of type 2 errors is small relative to the type 1 errors. Furthermore, many of these are cases where the forecast may be for growth just below zero while the realisation ends up just above zero. Hence, despite their being an increase in the number of falsely forecasted recessions, the mean forecast error of these also decreased with time.
The results suggest that forecasters either do not have the information or the incentives to forecast recessions. Lack of information could be due to data not being timely or of not being good enough quality, or the models may not be good enough to be able to predict outlier events. It is also possible that recessions may occur because of events which are themselves difficult to predict or there may be a lack of incentives to predict a recession, such as reputational risk (if wrong) or the fear of disaffecting clients.
The paper does not set itself the task of explain which of these causes are responsible but it does identify a number of important and interesting factors or non-factors.
Recessions that follow a crisis (either a currency or a debt crisis) are, apparently, no easier to predict or recognise than recessions where there is no crisis. Triggers may be significant with hindsight but they are not perceived to be a pre-cursor at the time. Perhaps financial events only become crises when they are combined with recessions. Other research indicates, however, that recessions that are combined with financial crises are of much greater depth.
The authors also find that the economic models are not capable of generating big swings in outcomes away from some steady state level; to the extent that some forecasters rely on such models, they too will tend to have difficulty when outcomes depart strongly from normal. Other research indicates that forecasters tend to smooth their forecasts too much.
The conclusion is that while forecasts may be useful in identifying trends and maybe even upturns, they are not very good at recognising recessions. One particular problem would seem to be that recessions are generally not preceded by economic weakness. On the contrary, they are more typically preceded by strong growth which suddenly falls away. So, just because an economy is doing well today provides little assurance about how it will be doing in six or 12 months’ time.
For recessions, investors are ‘on their own’. But recessions are fundamentally caused by a fall in confidence or sentiment. Tracking that, not necessarily through formal indicators (which would be picked up by the forecasters’ models anyway) is, in my experience, of some help.
[1] How Well Do Economists Forecast Recessions?, IMF Working Paper, Zidong An, João Tovar Jalles, and Prakash Loungani, March 2018
