BY JOHN MAULDIN
This essay is going to insult a bunch of smart, maybe even brilliant, people. It is not polite nor is it politically correct. I will try to be better. But right now, I am pretty pissed.
Here’s the thing.
No serious scientist would run a two-variable experiment. By that I mean, you run an experiment with one variable to see what happens.
If you have two variables and something happens—either good or bad—you don’t know which variable caused it.
You first run the experiment with one variable, then do it again with the second one. After that, you have the knowledge to run an experiment with both.
And yet, the Fed is running a two-variable experiment. It raises rates and reduces the balance sheet at the same time.
It is decidedly the stupidest monetary policy mistake in a long line of Fed mistakes.
You Can’t Model the Future
Powell and the Federal Open Market Committee listen to very smart PhDs from all the best schools. They all use fabulous multi-algorithmic models.
These models apparently proved that you could raise rates and reduce the balance sheet at the same time. With no consequences.
I’m sure they are smart and nice people—and their kids and dogs love them.
My problem with them is that they mistakenly trust models based on past performance. Or even worse, models based on monetary theory that is clearly, evidently, badly, manifestly wrong.
They have been using these models to forecast the economy for decades. And they are about 0 for 300 in being right.
It is statistically impossible to be that bad unless your models are fundamentally flawed, which they are. Their underlying economic theories simply don’t work.
Because they have no politically and academically acceptable theories to substitute. They are slaves to their own mal-education.
They think this makes them smarter than the markets. I can’t say it any stronger than that.
I have actually been in the room when someone slammed a Federal Reserve economist about said models. He went so far as to say that the best thing that Powell could do would be to fire all those PhDs and ignore their models.
As you might imagine, the Fed economist was not happy with that analysis. The veins in his neck were popping, he was red faced, and his voice was raised. This clearly got his goat.
Now, here’s the lesson I learned, which was burned into my brain. The assaulted economist asked a very simple question, (neck veins popping): “You can’t take away a model without replacing it with another model. What model will you replace it with?”
The critic, who is perhaps the best observer of the bond markets I know, stammered a little bit and then forcefully said, “You can’t actually model the future.”
Messing Up the Economy with Worthless Models
When I say the words “past performance is not indicative of future results,” I damn skippy mean them.
All past performance models were built in a particular macroeconomic environment.Unless you can find a macroeconomic environment that is very similar to today’s, every model deserves a tad bit of skepticism.
Maybe it will work and maybe it won’t. It is up to the macro analyst to try and figure out which one will work well enough to confidently invest your money.
I can’t tell you how hard and difficult and truly daunting that is. Especially after you have done it for many years and have the scars to prove it.
I’ve looked at a lot of macroeconomic models. I can’t describe how much I would love to find a macroeconomic forecasting model that was actually reliable.
To have such a crystal ball would not only be soul soothing. It would also be extremely profitable for my clients and, admittedly, me. It would be the Holy Grail.
All those PhDs at the Fed still haven’t found the Holy Grail after 40 or 50 years. Hell, they haven’t even found a decent cup of coffee. But they think they have.
So their bosses confidently run a two-variable experiment with our economic system.