Mean machine - Focus on regression to the mean rather than extrapolating from the present
Regular visitors to The Value Perspective will know we have never been as confident as some other investors about the chances of China’s economic growth figures continuing to defy gravity – just as they will know we are rarely slow in flagging up any academic research from Harvard economics professors when the conclusions chime with our own views.
In this instance, the Harvard economics professors – including former US Treasury Secretary Lawrence Summers – are admittedly not as blunt as we were in, say, Moving Target, where we observed: “For some, it is now inconceivable Chinese growth could ever reach the depths of a rally bad number such as, oh, 6% or 5% – when a really bad number would actually be one preceded by a minus sign.”
Nevertheless, Asia-phoria meets regression to the mean – the most recent version of which was published by Summers and international development expert Lant Pritchett towards the end of 2014 – is a fascinating examination of why anyone who think China will continue posting growth of 7%-plus for decades to come is likely to be gravely disappointed.
Expensively disappointed too since, calculate Pritchett and Summers, the difference between China continuing on something like its current growth trajectory between now and 2033 and – as they suggest is more likely – it following the example of other economies that have developed at a similar rate and regressing to the mean is $42 trillion (£27 trillion, give or take a few billion pounds).
According to the paper, there are “two common failures in economic forecasting” – and what better time to restate those than the turn of a new year when economic forecasts are so thick on the ground? One failure, say Pritchett and Summers is “excessive extrapolation of the (recent) past into the (distant) future, particularly susceptibility to ‘irrational exuberance’”.
The other is “excessive subjective certainty that relies on confidence in continuity and hence consistently under-predicts discontinuities – even relative to their known past probabilities”. How is that relevant to China and indeed India? Well, while the International Monetary Fund (IMF) may have softened its bullish outlook on those two economies, it still remains very positive about them.
In fact, observe Pritchett and Summers, “the IMF World Economic Outlook and the general discussions about the future of the global economy are often premised on the continuation of perhaps slower growth but growth that is super-rapid (more than two standard deviations above the cross-national average) in both China and India”.
They go on: “The single most robust empirical finding about economic growth is low persistence of growth rates. Extrapolation of current growth rates into the future is at odds with all empirical evidence about the strength of regression to the mean in growth rates.” As you may imagine, we particularly like the star billing given to that great value investing staple, regression to the mean, but it gets better still.
An interesting addition to the latest version of the paper might almost have been written for The Value Perspective, noting: “Many of the great economic forecasting errors of the past half century came from excessive extrapolation of performance in the recent past and treating a country’s growth rate as a permanent characteristic rather than a transient condition.”
As examples, Pritchett and Summers highlight a textbook from the early 1960s that predicted there was a substantial chance the USSR would overtake the US economically by the 1980s as well as noting a view that was widely held until the end of the 1980s that Japan would continue to grow and outcompete the world – rather than seeing its GDP grow by just 12% over the next two decades.
Coming at the issue from the opposite direction, the pair invite us to “consider the pervasive pessimism of even a decade ago regarding Africa. Since then, African countries have emerged as a majority of the world’s most rapidly growing nations.” When it comes to forecasting future growth, they suggest, past growth performance is of very little value.
Pritchett and Summers demonstrate this by examining the growth history of different developing economies and comparing the growth rate they enjoyed as they developed, as it were, with the growth rate that followed this period. They found that, in most instances, there was more variation in growth rates over time within the same country than there was between different countries.
In other words, as the years pass, the growth rates of developing nations look less like themselves and more like the average growth rate. The paper identifies 28 countries that have enjoyed “super rapid growth rates” similar to China and, unfortunately for those forecasters who are looking to China to continue in a similar vein, the median growth rate of the periods that followed for those 28 is just 2.1%
So it seems reasonable to conclude that, sooner or later, China will revert to the mean and proceed to grow in line with the rest of the world. That begs the question of what form this reversal will take and here Pritchett and Summers highlight a big difference between developed and developing nations – not how fast they grow when they are growing but what happens when that growth slows.
According to the paper, when they are growing, countries with a per capita income greater than $20,000 a year see average annual growth of 3.8%, while those with a per capita income below that level average annual growth above 5%. When richer countries decline, however, they fall by about 2% a year whereas poorer countries decline by well over 4%.
That would suggest that, when the Chinese economy does ultimately slow, it could suffer a relatively hard landing and the crucial questions for investors – and of course for forecasters – will be the extent to which the powers that be are able to put a brake on this decline and how quickly they are able to return the country to growth.
Interestingly, the paper is not simply seeking to bash faceless academics – Pritchett and Summers note that, over the years, the IMF, to which they are both connected, has consistently over-relied on historic growth rates to predict future ones. Either way, one moral of this story has to be that investors would do well not to be too bullish about the outlook for China or indeed India.
Pritchett and Summers conclude: “We are not arguing one can predict with any degree of accuracy or confidence a slowdown but certainly policymakers need to be prepared for a wider range of extended slow-growth outcomes in these Asian giants than those that currently dominate the discourse. Hitching the cart of the future global economy to the horse of the Asian giants carries substantial risks.”
For our part, we would merely add that regression to the mean is a bigger and stronger phenomenon than most people tend to recognise and one that captures many of the subtleties about the world that extrapolation does not – competition, say, or a whole host of other factors no forecaster (or investor) can hope to understand fully about the difficulties inherent in growing a country or a business.
As we never tire of saying here on The Value Perspective, the future is uncertain and very hard to forecast. Investors would do well to resist the temptation to extrapolate the future from how things look in the present and instead recognise the possibility that things can change materially going forward.
Fund Manager, Equity Value
I joined Schroders European equity research team in 2007 as an analyst specialising in automobiles. After two years I added the insurance sector to my coverage. In early 2010 I moved into a fund management role, and then took over management of two offshore funds investing in European and Global companies seeking to offer income and capital growth.
The views and opinions displayed are those of Nick Kirrage, Andrew Lyddon, Kevin Murphy, Andrew Williams, Andrew Evans, Simon Adler, Juan Torres Rodriguez, Liam Nunn, Vera German and Roberta Barr, members of the Schroder Global Value Equity Team (the Value Perspective Team), and other independent commentators where stated.
They do not necessarily represent views expressed or reflected in other Schroders' communications, strategies or funds. The Team has expressed its own views and opinions on this website and these may change.
This article is intended to be for information purposes only and it is not intended as promotional material in any respect. Reliance should not be placed on the views and information on the website when taking individual investment and/or strategic decisions. Nothing in this article should be construed as advice. The sectors/securities shown above are for illustrative purposes only and are not to be considered a recommendation to buy/sell.
Past performance is not a guide to future performance and may not be repeated. The value of investments and the income from them may go down as well as up and investors may not get back the amounts originally invested.