The economy in 2022

“There are two kinds of forecasters: those who don’t know, and those who don’t know they don’t know.” You might think that the emergence of the Omicron variant of coronavirus reinforces the truth of JK Galbraith’s quip: what happens to the economy next year will depend in large part on the course of the pandemic, and nobody can predict what this will be.

In fact, though, Galbraith’s remark is true pandemic or no, because economic forecasts have often been wrong. Since 1997, the average error in UK economists’ forecasts made in December for real GDP growth the following year, as surveyed by the Treasury, has been 1.2 percentage points. (This ignores the direction of the error; on average, growth has been 0.2 percentage points lower than expected.) If we ignore 2020, when GDP fell 10.8 per cent after economists had forecast 1.1 per cent growth, the average error has been 0.8 percentage points.

And these errors haven’t been random. One mistake forecasters have made is to exaggerate the effect that their personal concerns would have on the economy. In 1998 they forecast weak growth, fearing that the Asian financial crises and collapse of Long-Term Capital Management would hit confidence. But in fact the economy expanded nicely. And in 2016 they though the Brexit referendum result would cause a significant economic slowdown, but it led to only a mild one.

A bigger mistake has been to fail to foresee recessions. 2020 was not unusual in this regard. In December 2007 economists predicted GDP growth of 1.9 per cent for 2008, but in fact the economy contracted then. The error here is neither a uniquely British one nor a recent one. Back in 2000 IMF economist Prakash Loungani studied forecasters’ performance around the world and concluded that “the record of failure to predict recessions is virtually unblemished”. And in the US just before the 2008 financial crisis Glenn Rudebusch and John Williams wrote that “professional forecasters have essentially no ability to predict future recessions” – a conclusion which applied equally well to the recession a few months later.

It’s in this context that we must read forecasts for 2022. The average forecast, as surveyed by the Treasury, is for real GDP growth of next year of 5 per cent. That would be down from this year’s expected 7 per cent, but still the third best year since 1968. This does not of course mean the economy will actually expand by this much. It just means that economists are confident of decent growth.

And there are reasons for them to be. One is simple maths. Because the economy has grown so much since last winter’s lockdown, year-on-year growth will look good in 2022 even if there is no further expansion. If GDP merely remains at September’s level it will be 2.3 per cent higher in 2022 than in 2021.

But we’ve reason to expect better than that. Uncertainty about policy – not just about that focused on the pandemic but also about Brexit – has fallen sharply in recent months. This means that companies that had been delaying expansion plans until the fog lifted will finally start to implement them. That should lead to a rise in capital spending and in employment – with the latter helping consumer spending to grow.

Equally, though, there are risks to this – and not just from the possibility that fears about new coronavirus variants (of which Omicron might not be the last) will keep people out of shops and pubs.

One comes from fiscal policy. Higher National Insurance Contributions in April, cuts to Universal Credit and a squeeze on public spending mean this will tighten next year. The OBR forecasts that cyclically-adjusted next borrowing (a measure of fiscal policy) will fall from 8.3 per cent of GDP in 2021-22 to 3.9 per cent in 2022-23.

Another risk comes from consumer spending. Retail sales volumes have fallen 4 per cent since their post-lockdown bounce in April, during which time households’ bank deposits – which had soared during 2020 – rose 3.2 per cent. This suggests we’ve not seen the post-lockdown release of pent-up demand which some had hoped for. Instead, we might have fallen into more frugal habits.

On top of all this is the fact that trend growth was slowing down long before the pandemic. In the 15 years to 2019 real GDP per person grew by just 0.8 per cent a year. That compares to 2.4 per cent a year in the 15 years to 2007 and over 3 per cent during the long boom of 1946-73. This slowdown is of course not confined to the UK: we’ve seen a similar thing in the US and eurozone. Economists disagree upon why this has happened. But they agree upon one thing – that whatever the causes, they have not gone away. Even if we do see a good bounce in economic activity next year, we must not mistake this for a reversal of the long-term slowdown.

What will happen to inflation?

Another question for 2022 is: what will happen to inflation? Economists expect it to fall, from 4.2 per cent in October to 3 per cent by the end of next year.

Just as GDP forecasts are wrong, however, so too are inflation forecasts. Since the Treasury started surveying CPI forecasts in 2004 the average forecast made in December for the following 12 months has been wrong by a percentage point, if we ignore the direction of that error.

As with GDP forecasts, these errors have not been random.

For one thing, forecasters have tended to under-predict inflation, by an average of 0.3 percentage points.

Such under-predictions have been especially large early in upturns. In December 2009 economists predicted inflation in 2010 of 1.8 per cent, but it turned out to be 3.4 per cent. And inflation is now over 4 per cent even though forecasters this time last year predicted just 1.9 per cent.

Such errors tell us that economists over-estimate the extent to which recessions and spare capacity reduce inflation and under-appreciate the power of another effect. This is that recessions scramble the patterns of supply and demand, thus creating mismatches between the two. That leads to localised shortages, and hence inflation to the extent that such shortages do more to raise prices than excess supply does to reduce them.

Which is perhaps a reason to suspect that inflation will fall next year. These mismatches should gradually diminish and as they do inflation should fall. A precedent here is the recovery from the 2009 recession. During this, inflation rose at first, peaking at over 5 per cent in 2011 – but it then fell back to zero by 2015.

This is not the only reason to expect inflation to fall. Another is a simple base effect. Next spring will see this year’s rises in oil prices start to fall out of the annual inflation data. Even if the level of prices stays high, therefore, the annual rate of change (which is what inflation is) will fall. And prices might not stay high. In China the annual growth rate of the M1 measure of the money stock has fallen from 14.7 per cent in January to 2.8 per cent now. In the past, such slowdowns have led to falls in the prices of oil and other commodities.

What’s more, despite shortages of lorry drivers and care workers, the labour market overall is not tight. Total hours worked – a measure of labour demand – are still 2.8 per cent lower than they were at their 2019 peak. Thanks in part to this average weekly wages in the private sector are only 2.3 per cent higher than they were last December, which means they have fallen in real terms. So far, then, there are no signs of the sort of wage-price spiral that causes serious inflation.

This isn’t to say there are no inflationary risks. There are, and one comes from the virus. If a new wave of it were to threaten China the country’s zero-Covid policy could lead to lockdowns and port closures there, thus disrupting already-stretched supply chains. Neil Shearing at Capital Economics says: “Compared to previous waves of the virus, which were on balance disinflationary, a major new wave could now be inflationary.”

Even without this risk, economists believe that central banks will finally raise interest rates next year. Futures markets are pricing in a half-point rise in the US fed funds rate and a full point rise in UK bank rate, with rises also in the eurozone. This doesn’t mean rises are imminent. Uncertainties created by the Omicron variant and downside risks to economic activity are, says Holger Schmieding at Berenberg Bank, “a clear reason to hold back” from raising rates soon.

Again, though, these forecasts could be wrong. A big reason why bond yields have consistently trended down even since the financial crisis is that investors have consistently expected rates to rise and been surprised by their failure to do so. Of course, this doesn’t mean 2022 will see another such surprise. But it does warn us to take such forecasts with a pinch of salt.

Which poses the question: if forecasts are so unreliable, what’s the point of them? Former Bank of England governor Mervyn King likened them to guide-posts. Their purpose isn’t to tell us what will happen, but to tell us how surprised we should be by what actually does happen.

As investors, though, we shouldn’t pay much attention to point forecasts. Think of golf. When you are on the tee it’s not always wise to try to hit the ball as far as possible: you might shank it or go into a bunker. Instead, you should think about all the possible places the ball might end up. It’s often better to avoid the worst places rather than aim for being as close to the hole as possible. What you should do, says James White at Elm Partners, is to “choose the target that gives us the highest expected strokes gained”.

The same, he says, applies to investing. We must think about the range of outcomes, not just the single likeliest one. Which is why diversification across assets is so important. And why cash should be part of most portfolios. The point forecast for returns on cash looks grim; real interest rates are negative. But the worst-case outcome for cash is better than that for other assets.

For investors, probability distributions matter more than forecasts. But still, forecasting keeps economists out of mischief – which is not nothing.