Further on why inflation

  • Output growth in the second half of 2020 and 2021 is likely to be well above trend
  • Current inflation data is meaningless, but 2021 and 2022 will see high inflation 
  • This assumes that global lockdowns end over the summer

This is a follow-up to The case for higher inflation published on 5th May. It is intended to address some feedback and also to further lay out the case for why the world is likely to see high inflation in 2021 and 2022.

In essence, the arguments are that — assuming of course that global lockdowns end in early summer — first, that there will be a sharp recovery in economic activity in the second half of the year. Second, that inflation in 2021 and 2022 will be well above numbers we have seen for many years.

This is really a case of two stories, but they are interconnected.

To begin with growth, it is important to bear in mind that this recession — broadly speaking — is not one that has destroyed economic capacity. That is not to say that there are no serious consequences. A number of sectors — travel, hospitality, tourism, events — have suffered badly and see little hope for recovery. Other sectors — anything related to medicine and health care, high tech (notably involving distance working and meeting) as well as retail and food delivery — will benefit. Crucially, most of what was in place before the pandemic, including productive capacity, still remains.

Now add to this a number of factors:

  1. Large monetary stimulus programs in most large economies.
  2. Large fiscal stimulus programs in most large economies — at the time of writing amounting to about $8 trillion, that is to say around 10% of global GDP — and there is more still to come.
  3. Suppressed household and corporate demand during the lockdown phase.
  4. Stock building. At the moment, company inventories are generally perceived to be excessive, as the lockdown phase has seen an involuntary rise in inventories. However, this is likely to be reversed in H2 2020, not least as a result of point 3) above. Moreover, companies, burned by the danger of just-in-time inventory levels and dependence on a single supply source (=China) are likely to want future inventories to be higher than in past decades.

Admittedly, all these factors are temporary. While they last, though, their impact should be powerful. Hence, during this phase, we are likely to see growth well above current trend rates of 1½-3½ % (depending on the country). It will probably not regain all the ground lost during the lockdown phase, but it will go a good way towards it, before — most likely — levelling off and growth returning to trend. This means that the recovery is likely to be square root-shaped (√).

Is this too optimistic? One argument sometimes heard is that, in the future and as a result of the pandemic, people will be more careful. They will spend less and want to rebuild buffers and they will in general be more caring and less materialistic. The first half of that supposition (spend less to rebuild buffers) is possible, although again temporary. The second half, however, is highly unlikely. History shows us that this is not how people react. Following a traumatic experience, the instinct of the survivors has generally been to want to forget the past and enjoy themselves. This was the reaction after World War I and the Spanish flu (the Roaring Twenties); as well as after the Black Death of the Middle Ages.

However, temporary above-trend growth need not necessary mean inflation. So why should inflation accelerate? 

There are a number of reasons. First, however, it should be noted that the inflation numbers we are currently seeing and will see for some months, are completely meaningless. The consumer price index is calculated from the prices of a basket of standard commodities that are bought in normal times. But these times are far from normal. If goods and services are not actively bought and sold, their prices do not matter when measuring current inflation. Nor do we know what prices will be when lockdowns are lifted, since we can assume that for at least some categories (clothes? big ticket items?) there will be large-scale discounts to clear excess stock, whereas others (hospitality, travel?) will still not be sold in pre-pandemic quantity for some time.

One key reason to expect higher inflation is the monetary stimulus covered in the previous Comment. To briefly repeat, in contrast to the situation in the aftermath of the Great Recession, the last few months have seen a surge in broad money growth. Tim Congdon of the International Institute of Monetary Research (IIMR) at the University of Buckingham notes that current US broad money growth rates is likely to be the highest ever in US history, with the exception of the Revolutionary War and in the Confederacy during the Civil War. The US broad money measure used in this Comment is growing slightly less rapidly, but still rose by a three-month annualised rate of 50% in April. In other economies, broad money growth has not been quite as rapid, but, as the table shows, there have been very substantial accelerations in recent months. With one exception, broad money growth is at or approaching double-digit rates on an annual basis in all economies shown and considerably higher on a three-month annualised basis. The exception is Japan; but even Japanese broad money growth is the fastest since 1998.

Two or even three months’ worth of rapid broad money growth need not be inflationary. But the monetary surge shows no sign of abating, nor is it likely to. True, growth rates are likely to come off somewhat, if only due to the base effect. US commercial banks’ commercial and industrial loans have been growing at a 13-week (ie, three-month) annualised rate of more than 100% for eight weeks. This number will obviously begin to come down once the initial increase becomes part of the denominator, but the monetary stimulus and the rapid broad money growth it entails is likely to be with us for at least six months and probably to the end of the year as governments and central banks attempt to boost the recovery. Double-digit broad money growth over that length of time is not consistent with inflation near zero, let alone with deflation.

Following the previous Comment, one reader raised two objections to the idea that rapid broad money growth would be inflationary. First, he noted, what if the money is not used productively, but simply used to ‘stay afloat’? Second, what if the desire to hold money balances is infinitely elastic? In other words, the money is not spent but simply accumulates in bank accounts (in technical terms, the velocity of money goes to zero).

The first objection is easy to deal with. As long as money is spent, it doesn’t matter what it is spent on. Yes, productive investments are relevant for future growth, but for current growth, it doesn’t matter if money is spent on shoes and bubblegum or on high-tech. Nor does it matter if it is spent on assets — real or financial — instead of on goods and services. Any spending counts as economic activity.

As for the second point, yes, this is possible. However, as Simon Ward of Janus Henderson Investors shows on 13th May (The quantity theory of wealth), episodes of cash hoarding are both infrequent and temporary in history. For this to happen, you have to assume that current demand levels are the new normal and that demand will never recover from what is, after all, artificially suppressed levels. In an earlier post, Simon also addresses the velocity issue, noting that “Pessimists argue that the monetary surge is being offset by an even larger fall in velocity. This is trivially true currently but the velocity collapse is likely to prove temporary – it mainly reflects, after all, physical limitations on demand and production due to the shutdowns – while the monetary addition will almost certainly prove permanent: does anyone seriously believe that the current crop of “independent” central bankers will be prepared to argue for a reversal of QE, or even its cessation, as the crisis starts to abate?”

However, in addition to the monetary driver of higher inflation, there is another, possibly equally important factor. This is that governments will want to see higher inflation. That is not so surprising. Governments have taken on substantial amounts of new debt to support their economies during the COVID-19 pandemic. Even if government debt need never be repaid, it must still be serviced. Yet no government is going to want to raise taxes or cut back spending, both of which will dampen the economic recovery they need. By contrast, a few years of high inflation, even, say, 5%, will (from a government perspective) be very useful in terms of eroding the real value of the debt.

There is a second strand to this. High inflation shrinks the relative debt burden; but, at least over the next few years, budget deficits will remain wide, so large-scale borrowing will have to continue, increasing the nominal value of the debt. But governments will also need that new borrowing to remain cheap. It is therefore also likely that we will see governments pushing central banks to impose what in Japan is called yield curve control (YCC), that is to say, committing to keep the yield on long-term government bonds at a certain, low, level. In other words, the return of financial repression. In normal times, bond markets would react to higher inflation (actual and anticipated) by falling, pushing up long-term interest rates. Under a system of financial repression, the central bank commits to keeping interest rates at a specified level, usually well below the rate of inflation. This means that inflation can rise unhindered by either central banks or bond markets (in fact, abetted by central bank action).

Nor is that all. When central banks attempt to control interest rates, they give up their ability to control prices. So by agreeing to impose financial repression, central banks would also de facto abandon their attempts to control inflation. True, most major central banks are independent. But, with the significant exception of the European Central Bank, all of them have their independence granted by their respective government and legislature, and what has been given, can be taken away. This can be done openly, or by changing the terms of central bank legislation, e.g. by imposing new targets (as was recently discussed in New Zealand) or eroding the dividing line between monetary and fiscal policy and restricting monetary policy instruments (as the Swedish government’s proposed new Riksbank legislation attempts to do); or simply by loudly calling for changes in policy (as in the case of President Trump).

How long this period of high inflation will be is difficult to estimate. It depends on too many factors, including the future course of the COVID 19 pandemic, the growth outlook and popular and political tolerance of higher inflation. But it is likely to last for at least a couple of years.

Gabriel Stein