Way back in March, I wrote about stagflation, an economic state with high inflation, unemployment, and slow economic growth. I grade my own articles on a curve, and I give that one a C because I didn’t talk much about stagflation, instead writing about the related Misery Index.
Two or three weeks ago, I saw a thematic ETF launched that is designed to benefit from periods of stagflation which got me thinking about updating my article (short version: this ETF is a bad idea in my view).
At the same time, a few clients asked about the risk of deflation and what we might do about it, so I’m giving it another shot. And, almost immediately, I remembered why I pivoted to the Misery Index: I couldn’t find an index that showed stagflation.
I read an article a few months ago by an investment manager whom I admire who created their own proxy for deflation, and while that one seemed a little complicated to me, it gave me the idea that I could create my own, which I did over the weekend.
It’s a simple model, and I called it ‘Dave’s Stagflation Model’ because I want to highlight that it’s not an Acropolis model and is sort of a ‘quick-and-dirty’ that a Ph.D. economist might scoff at (I’d challenge them to make one since there aren’t any I could find).
First, my model looks at inflation over the last twelve months compared to the previous ten years. It does the same thing with unemployment and the Purchasing Managers Index, which is a common proxy for economic growth.
The model scores when these conditions are present, and a perfect score of three means that all three indicators are worse now than over the previous decade. A perfect score of zero means that everything looks better: lower inflation, lower unemployment, and higher economic growth.
Here’s what it looks like, with recessions shaded in gray:
Broadly speaking, the model looks like what I expected: there are a bunch of high readings in the 1970s and early 80s and not much since then.
I was surprised about the high reading in the 2008 financial crisis because that doesn’t strike me as an inflationary period – we were worried about deflation.
And, it turns out that it highlights one of the ways that Ph.D. might scoff: inflation in Jun of 2008 was high compared to the past ten years, but it was only 2.2 percent compared to the 1.8 percent reading over the decade that ended then.
This model is less about the absolute level of inflation and more about the relative level. And that’s true for the other metrics too.
Satisfied enough with this model, I looked at the current reading and felt good that we aren’t in a period of stagflation (by this definition).
We are clearly tripping the inflation meter and the growth signal – but we’re not even close to hitting the unemployment signal, which is 3.7 percent compared to 5.3 percent over the last decade.
The 2008 experience shows that we could get there quickly – this measure of stagflation was right about where it was right before the recession.
So, I can’t rule out stagflation in the near, relatively near future. At the same time, there are a bunch of periods where the model result headed higher from the current level only to back down shortly thereafter.
Although I had already concluded that there was no predictive value to my simple model (no surprise there), I tried to find market rules for stagflation, messing around with stocks, bonds, cash, and gold. Despite an hour of looking, I didn’t find anything obvious.
I probably could have found something with enough trying, which reminds me of an old joke: if you torture the data long enough, it will tell you what you want to hear.
Even with my spiffy new stagflation model, I don’t have any insights about what we could do differently or better because I didn’t find evidence between periods of stagflation and low returns or vice versa (the r-squared between my model stock returns was -0.01; aka meaningless).
Perhaps I could make a better model. Or, better yet, a Ph.D. economist could make one. But I still doubt that we could find a way to trade stagflation, unlike the ETF provider that I mentioned at the top.
I always feel like it’s worth looking at, and I had much fun building this little model and messing around with it, but I’m not going to put any stock in it, for your account or mine.