- “US companies sound inflation alarm. From home builders to toymakers, businesses say costs are rising.” – Financial Times Headline, March 30
- “Costs are going up everywhere. It’s DefCon 4 right now.” – CEO of a major American company
- “The 10-year U.S. Treasury yields are at a post-shutdown high of 1.63%. …These levels remain historically low, and …actual measures of consumer prices are still unremarkable, and well below 2%. But the shift in market psychology has been very swift, with the prospect of a return to secular inflation discussed seemingly everywhere. Why?” – Bloomberg
After years, decades, of quiescence… inflation is set to explode. Or so we are told. Bond investors have been “routed.” Treasury yields are rising, reflecting the sell-off. The stock market is on edge. Ken Griffin, of Citadel, the mogul-du-jour
- “…warned of a doomsday scenario in which accelerating inflation deepens a bond market sell-off, and sends stocks tumbling.”
Doomsday? Michael Burry (the “Big Short” star) “compared Germany’s path to hyperinflation in the 1920s to America’s current trajectory” in a tweet labeled #doomedtorepeat.
Is the Threat Real?
There is nothing yet in the numbers to indicate significant inflationary pressure. The trend, if anything, has been downwards, deflationary.
MORE FOR YOU
Even the 50 basis point “surge” in the bond yields in recent weeks – which triggered the latest inflation panic – had not breached the trend lines.
In fact, looking back, it would seem that deflation has dominated the credit markets over the past several years, with negative rates smothering Europe and Japan bravely setting a (technically) non-negative 0% interest rate target for its long bonds. Even the U.S is just crawling out of a situation of negative real interest rates.
The Inflation question is all about what might happen in the future. The “doomsday” scenario hasn’t happened yet — but is it about to? There has been no significant inflation for several decades. What is it in the current situation that portends a sudden change?
This leads to murky questions of definition and causality. If we knew what caused inflation to start or to accelerate, we could look for the precursors associated with the causal factors.
Unfortunately, we don’t know what to look for because the cause of inflation is unknown, or at least uncertain (and don’t let any economist tell you otherwise). The standard answers have been canceled out by the events of the last decade. We have entered the “puzzle” phase – as Janet Yellen described it famously in 2019.
Let’s review some of the broken theories.
Inflation Is About Rising Commodity Prices – Not!
We’ll start with a definitional question. What is Inflation?
Isn’t the answer obvious — rising prices? As a core economic problem, Inflation was first conceptualized by economists in terms of commodity prices. (The analysis of its monetary aspects came later.) It is still the starting point for the conversation. The Wall Street Journal recently cited rising commodity prices as evidence for the awakening of the dragon. They mentioned, for example, the recent rise in the prices of copper, and crude oil. The Financial Times added its concern about the prices of lumber, polyethylene, raw materials in general…
The problem is — it is not true.
The Federal Reserve Chairman knows this —
- “Jay Powell has so far dismissed rising prices… He [sees] the resulting impact on inflation as “neither particularly large nor persistent.”
The numbers show that he is right.
Relying on Federal Reserve data, the correlations of all commodity prices over the last decade with the rate of inflation are negative.
And what of “the most important price in the global economy” (as the WSJ refers to the yield on the 10-year Treasury Bond)? – it is also negative since 2009.
In other words, observing the data over this extended time period an analyst could have concluded that rising bond yields were generally associated with lower inflation – not higher inflation as the WSJ warns. In detail, however, the picture is much more confused. Here are the year-by-year correlations of bond yields (10-year Treasurys) with the Consumer Price Index.
In shirt, there is practically no meaningful, stable, persistent relationship between bond yields and inflation over the past decade. There is no meaningful interpretation of the recent small rise in Treasury yields in the context of this overall pattern.
Causality: The Broken Theories
But if rising prices of commodities or credit do not drive inflation, what does?
There are three classical answers.
This is the “rising prices” argument, in so many words. When the Saudis and others jacked up the price of oil in the 1970s, the “shock” created high inflation for a while as the cost increases wormed through the economy. This reinforced the idea that inflation is at least sometimes caused by commodity price increases. The charts shown above disproves that.
But – there is a bigger problem with the cost-push theory: technology. It is now clear – as it may not have been in the 1970s – that digital technology is driving an enormous, across-the-board decline in prices of… almost everything, on an ongoing basis. Moore’s Law has morphed into a tech-driven deflation process that economists struggle to understand. In fact, it is obvious. (The point has been argued sufficiently in many places, so we’ll let that statement stand on its own for now without quantitative elaboration.)
Tight Labor Markets
Lower unemployment leads to higher inflation. This is the theoretical landscape described by the famous Phillips Curve – the trade-off that supposedly holds between unemployment and inflation. The causal concept is that as the labor market tightens, workers gain bargaining leverage and extract wage increases from employers, who pass along the costs in the form of higher prices for finished goods, the empowered workers have the wherewithal to pay the higher prices, etc.
Well, the Phillips curve — which actually never worked as a precise description of reality, is certainly not working now. As Janet Yellen has admitted:
- “The slope of the Phillips Curve—a measure of the responsiveness of inflation to a decline in labor market slack—has diminished very significantly since the 1960s. In other words, the Phillips Curve appears to have become quite flat.”
In fact, unemployment had fallen (prior to the pandemic) to the lowest levels in half a century – plunging right through the several thresholds where inflation should have kicked in. It didn’t happen. The tight labor market explanation is null.
Loose Monetary Policies
“Inflation is always and everywhere a monetary phenomenon” – said economist Milton Friedman, once upon a time– but he was not the father of the monetary approach to the problem of inflation. The idea that governments with printing presses can destroy the economy by creating too much money is a fear that predates modern economics. The spectre of German hyper-inflation in the 1920’s still haunts some of us in the 2020’s.
Again, however, this once-plausible view of the matter has been blown to pieces by the central banking policies of quantitative easing in the last dozen years or so. Trillions of dollars have been pumped into the global economy, by the Fed, the European Central Bank, the Bank of Japan – and inflation has not budged.
Once again, the anti-monetarist argument has been made extensively in the media – usually portrayed as a recent truism, and a deep mystery. It can stand here without additional quantitative adornment. Suffice it to say that no one can sustain any longer a simplistic “always-and-everywhere” monetarist viewpoint of inflation.
“Expectations” as a Causal Factor
The failure of the classic explanations has led to the emergence and popularization of a new and innovative “theory.” Many economists have adopted what is essentially a psychological explanation of inflation: inflation is caused (they say) by peoples’ expectations of inflation.
- “Inflation is a function of two forces: the inflation expectations of the public and the amount of slack in the economy. Inflation expectations are an important consideration in wage and price decisions, and slack influences the pricing power of firms and workers.” [Hanging on to the older theory along with the newer.]
This is a bizarre argument, even if it contains a grain of truth. It is close to the same thing as saying that inflation causes itself, that it arises somehow spontaneously from public sentiment. It says that if we expect something to happen, it will happen. Imagine if a similar reasoning were applied to the question of what causes recessions. Would we be comfortable saying that if people expect a recession to occur, that is enough to bring it about?
There is much to say about this. The idea is not completely absurd, because the financial and economic system does exhibit what has been called reflexivity.
- “Reflexivity refers to circular relationships between cause and effect, especially as embedded in human belief structures. A reflexive relationship is bidirectional with both the cause and the effect affecting one another in a relationship in which neither can be assigned as causes or effects…the self-reinforcing effect of market sentiment.” – Wikipedia
But connecting expectations to inflation has not turned out to be any easier than the older models, and has not really gained any predictive power.
Measuring Inflation Expectations
The Federal Reserve tracks several measures of inflation expectations. All of them are weak when it comes to forecasting .
The inflation metric that has the financial press worked up lately is the “5-Year Breakeven Inflation Rate” (based on the difference between the 5-year Treasury Bond yield and the yield on the inflation-indexed 5-year Treasury Bond). Many economists believe that this comparison ought to expose the “expectations” component.
The sharp upward trajectory since the pandemic crash in March 2020 has convinced a lot of people that inflation must be about to explode.
The “surge” shown in this chart is visually striking. But Breakeven is a poor predictor of actual inflation. The correlation of the 5-year Breakeven forecast with the actual inflation rate (CPI) measured five years later is negative 20% since 2003. Breakeven is better correlated with current inflation levels (25.3% positive). In other words, like many measures of market sentiment, it is more sensitive to the immediate context – it reflects the current inflation fear-talk – and less accurate, even contrarian (backwards), when interpreted as a forecasting tool.
The Fed’s “T5YIFR” is an alternative metric which measures the expected inflation rate over five years in five years time. The trend is down-sloping.
Again, it proves a poor predictor of actual inflation. The correlation of the T5YIFR forecast with the inflation five years out is just negative 8%. Basically, it has no predictive power, contrarian or otherwise, with respect to actual future inflation 5 years out.
Interestingly, when mapped against current inflation, the movements in the T5YIFR seem more markedly contrarian. Up-moves in the T5YIFR are followed soon by down-moves in actual inflation, and vice versa. (Apologies for the busy-ness of this chart.)
- “The Sticky Price Consumer Price Index (CPI) is calculated from a subset of goods and services included in the CPI that change price relatively infrequently. Because these goods and services change price relatively infrequently, they are thought to incorporate expectations about future inflation to a greater degree than prices that change on a more frequent basis.” – Federal Reserve
The “sticky CPI” shows a sharp downward trend, over the past year – predicting lower inflation?
Alas, the correlation of today’s “sticky CPI” with actual inflation 12 months later is also negative – perhaps contrarian – at negative 26.2%.
The Michigan Survey
The University of Michigan’s well-known series of consumer surveys includes a pol of inflation expectations.
The Michigan survey numbers always seem to run hot — consumers expect higher inflation than what actually occurs. They are also quite off the mark. Consumer’s forecasts have a negative 38% correlation with the actual numbers. A 6-month lag (that is, correlating the consumer expectations today with actual inflation 2 quarters from now, treating it as a forecast) results in a negative 41% correlation.
This is not surprising in light of the contrarian character of most measures of consumer sentiment. In fact a negative 40% correlation is almost strong enough to use as a forecast of what is likely not to happen. As an investor, the advice would be to expect the opposite of what the survey respondents project. The“wisdom of the crowd” is a fallacy here (as in most situations).
Finally, we might ask whether there is momentum in inflation that can be used to predict (if not to explain) its future trajectory. Unfortunately, the strength of the trend in the data is weak, and rapidly disappears (for predictive purposes).
In short, the current measures of expectations – including a simple extrapolation of the current state – do not forecast inflation successfully. To a limited degree, they have a contrarian value — they might be used negatively, or in reverse. But even that relationship is weak an unstable.
Still, the power of expectations is something to be reckoned with — especially when the public’s understanding of the issues, and expectations for the outcomes, are subject to the strong negative skew supplied by the major media – as described in my previous column.
In the next column, we will consider how the negativity of the media may be creating what might be called an expectations bubble – which could have a serious impact on the design of economic policy, leading to real economic damage. It has happened before.