To start this contribution on inflation, let’s open with some facts. First, central banks around the world tend to target inflation rates. In developed countries, an inflation target of 2% is not unusual. Second, since major central banks around the globe started to target inflation in the early 1980s, inflation has fallen from double-digit figures to low single digits. So, it is not that difficult to conclude that central banks’ targets are realistic and that there is hardly any reason for investors to worry about inflation in the medium term.
As a follow up, let’s undermine this view with two intriguing facts. First, if we take the median of all 19 inflation rates by country over the 111-year period 1900-2010 from the Dimson, Marsh and Staunton database, a staggering 3.9% inflation rate emerges. Things would be even worse if we averaged these inflation rates, with a 5.9% inflation rate due to hyperinflation in Germany in the 1920s. Second, the ECB targets an inflation rate below but close to 2%. Since the start of the euro in 1999 we have seen two major crashes on stock markets, an average compounded economic growth rate of just 1.6% and an inflation average of 2.1%. So, we have a relatively hawkish central bank that did not even reach its own inflation target in a period of economic weakness*.
And now comes the interesting part. Suppose history repeats itself exactly as it did in the past 111 years and we meet again in 2023 we would see a typical central banker of a typical country who would argue that they have done an excellent job as inflation has been close to 2% for the majority of years. However, actual inflation has more often been above 2% than below, making compounded average inflation 3.9%. Please note that inflation targeting usually takes place without exactly specifying what central banks target. So while central bankers might be interested in the number of years the inflation was close to 2%, a typical investor would also experience inflation spikes that results in an average inflation rate of 3.9%.
We think that investors should also pay attention to the past when making long-term predictions about inflation. We believe long-term inflation to be around 3%, referring to a compounded average. That is above the seemingly specified but actually unspecified inflation target of a typical central bank. It comes in at the compounded average of the GDP weighted average global inflation data series of all nineteen countries in the Dimson, Marsh and Staunton database. Still, it looks a conservative estimate as it is below the median (3.9% as indicated above) or average (5.9%) of the 19 individual compounded inflation rates. As we have shown, even if one is willing to believe central banks will be able to fulfil their inflation targets, this gives leeway for compounded rates that are significantly above 2%.
In the medium term, we might very well surpass a 3% inflation rate. We do not anticipate such a risk for 2012 or 2013, but for the years beyond. We identify three reasons. First, as Reinhart and Rogoff have shown (‘This Time is Different’), a banking crisis enlarges inflation risks. It is a helpful factor in reducing a sovereign debt pile. We do not give too much credence to promised price stability as we think central banks do not operate in a vacuum. If the global electorate prefers a pro-growth monetary policy with elevated inflation above defaults with price stability, that is what they will deliver.
Second, there is a change within the global population that is spurring demand for commodities. It is not the rise of the global population itself that increases the risk of commodity price inflation. The world’s population has almost tripled over the period 1950-2012 from 2.5bn to around 7bn people, so why worry about an additional 2-3bn more inhabitants on planet earth in the next 40 years? If anything, growth of the worldwide population is falling sharply. But, the number of people in the middle-classes is rising quickly; between 2000 and 2020 the middle-class has accelerated from roughly 1.5bn to close to 4bn people. The emergence of a global middle-class and the increase in consumption that results can actually increase commodity prices if supply does not immediately catch up with demand. Even if supply would immediately catch up with demand, it still can induce higher prices as the marginal costs of production rise.
Third, there is the problem of ageing which results in labour scarcity. It could be that ageing results in a struggle between the generations, with the young unwilling to contribute an increasing proportion of their income towards paying for the elderly in their retirement. Then, the young would ask their employers to compensate for the rising pension costs, which could result in cost-push inflation. Clearly, in the end, it depends on the monetary authorities whether labour costs push inflation or whether only the relative price of labour increases. In other words, ageing brings the risk of inflation but does not cause it. We also note that another way of solving the ageing issue would be a significant rise in the retirement age, a significant increase in labour immigration or a cut in social security. In these scenarios the inflationary pressures from ageing will be limited.
We conclude by discussing inflation hedging and protection. Here, once again, one should be precise about the definition. For example, if one refers to long-term protection against inflation, real estate is often mentioned as a good hedge. That is not surprising as it is a real asset. Therefore, it should offer better protection against inflation than nominal assets. But, what about changes in inflation in the short to medium term? Or, what about hedging against a quick rise in inflation, let’s say an acceleration from 2% to 6% in a two-year period?
To answer such questions it does not help to examine the effects of inflation on the returns of a range of asset classes over 40 years. Over that period we have only seen two inflation spikes, one in the 1970s and two minor hiccups afterwards - the rest has been noise.
Based first on theory, second on long-term analyses for the main asset classes equities, bonds and cash, and third on the analysis of the four inflation spikes since 1970 for a wider range of asset classes, we believe that the ex-ante inflation-protection properties of asset (sub) categories indicatively could be characterised as in the figure. Here, we assume the increase in inflation comes at least partly from rising commodity prices. Over the past 40 years, core inflation and headline inflation peaked at the same time, suggesting a strong relationship between commodity prices and other prices. Next, we suppose that the acceleration in inflation is limited to around two percentage points per year. Hence, commodities and commodity-related equities appear to be the best hedges against inflation risk. Runaway inflation would hurt equities more than is suggested in the figure.
We once again stress that the past has shown that changes in inflation only explain a part of the variance in asset returns, so these indications should be interpreted with caution.
* For this period, in the US inflation and economic growth averaged 2.5% and 1.9% respectively. For the UK these figures are 2.2% and 1.9%. For the bears among us, Japan experienced 0.7% growth and -0.3% inflation