The A/S and A/D Curves Are Not A Wassily Kandinsky Painting
Practitioners of economics have funny ideas about inflation these days. Inflation data amongst economists have developed into a veritable Rorschach Test for schools of economic thought. There’s an inflation metric that describes nearly any view on inflation you subscribe to. If you want high inflation, you point to rent at 3.77%, only ever briefly higher since the 1980s. If you want low inflation, you need merely to look at headline PCE or CPI — both around, or below, 1%. If you take the middle road, and think inflation is just about normal, you pick Core CPI, which is humming along at 2.3% y/y, which is close to the upper bound of the past twenty years, or Median CPI, which as running at 2.6% (and accelerating). A particularly motivated untruth presented is, “inflation in the things we need, and deflation in the things we want”. In reality, some of the biggest deflation has occurred in in food, energy, and clothing, which no-one sane person would argue are less needs than wants.
What gives? Is inflation hot, cold, or just right? The overwhelming majority of analysis falls on the fulcrum of political motivation — whether the analyst is predisposed to believing in more or less monetary or fiscal policy intervention.
The answer has been largely conflated by international developments. Part of the misunderstanding about inflation is probably due to the shift in trade intensity of GDP, primarily after NAFTA (1994), and then again after the Chinese entry into the WTO (2001). The new international supply of goods, particularly non-durable, which are labour-intensive to produce, and low value-add, pushed domestic employment into domestic services. This has profoundly linked changes between Real Wages and Real Services Prices.
The consequence of trade is that there has been a bifurcation of price level change, but not upon the stunted pivot of need or want, but rather on domestic or international.
This has served to turn core inflation into following domestic service labour tightness, and headline-only component inflation into following the international ex-US business cycle.
Before NAFTA, the relationship between Headline CPI less Core Services CPI and the dollar is virtually non-existent:
After NAFTA was implemented, Headline CPI less Core Services CPI became an inverse function of the US Dollar:
Economists — yours truly included — have overestimated incoming headline inflation. Instead of re-evaluating the assumptions of the inputs, they’ve thrown out the model entirely, despite the magnitude of the error (so far) being very small. This piece explores why that is probably wrong.
The foundation of all economics rests on the classical supply and demand curves. A normal supply curve is an upwardly sloping theoretical line or curve, whereby it is assumed that, under normal conditions, price and quantity produced are positively correlated. A simple explanation: let’s say there are 3 homes on your block for sale for $200,000. Bids for a home on your block sudden elevate to $500,000. The number of people willing to sell their home will go higher with the price.
Conversely, a normal demand curve is the downward sloping theoretical line or curve, whereby it is assumed that demand for something wanes as the price goes up, as consumers will forgo (if a luxury) or substitute where possible as it becomes more expensive.
There are conditions where other curve shapes are theorised. A Veblen Good is one that is supposed to have an upward sloped demand curve, where, as price of something goes up, so too does its demand.
It’s even been theorised that Money Is A Veblen Good. The idea is that the value of money and liquidity is subject to a feedback loop. In the broadest form, this seems unlikely. Since 1994, Core PCE inflation has been extraordinarily stable, ranging between 1 and 2.5%. Were liquidity premium feedback loops a general case, this kind of remarkable price stability would probably not be possible. A counter-argument to this is that the Fed has done such a spectacular job of managing the inflation rate through interest rates. The Fed, however, has made massive monetary policy impulses in response to huge swings in the output gap, which, at least on the surface, points to a fairly minimal elasticity to inflation.
Ignoring the observed minimal elasticity to inflation is particularly curious with the fixation of inflation expectations as a primary impulse on realised inflation. This view has been gaining momentum in b0th the financial blogosphere and policy making circles. The principal problem with this view is that it’s wrong. The correlation between 5y breakevens and 5y forward realised inflation, for instance, is 0.05. However, the correlation between 5y breakevens and trailing realised inflation is 0.59. Inflation expectations are consequently an extrapolation of the inflation we’ve just experienced, and that extrapolation isn’t meaningful to predict anything.
This failure to explain inflation follows the money supply theory of inflation’s failure. I’ve forwarded my own theory (published by the CFA Institute, 2013), based on empirical work, which theorizes that estimating normally shaped supply and demand curves is possible through demographic life-cycle and fixed capital stock analysis.
What I really want to understand from the proponents of inflation expectations based theories is what their supply and demand curves look like. Is it impossible to express in a linear equation? Is it fractal? Is it a Wassily Kandinsky painting?
The evidence just does not promote the need to throw out classical supply and demand curves. The inflation data we have received to date rather suggests that we should alter the steepness of our estimate of aggregate supply. Further, perhaps we should be separating out the international from the domestic portions of our economy for analysis and policy-making. We should be studying the implications and historical precedents of over-emphasizing foreign sector slack.
At the very least, we should not be making observational errors where we attribute the effects of the foreign sector slack to domestic economic analysis. When you do, you have to twist your estimates aggregate supply and demand curves into knots, and aggregate supply and demand curves are not Wassily Kandinsky paintings.
