Wednesday, September 7, 2016

Putting numbers on a prediction of vigor (intro)


We spent most of last month looking at household debt service, the TDSP series at FRED. Since the last recession, debt service has been showing a "bowl" shape, the left part of a bowl shape. I got all hot and bothered about the bowl. So I continued the debt service line upward to the right to complete the bowl, in the process predicting the future path of debt service.

Just so you know I'm not completely crazy, let me remind you that I also said this bowl-shape is similar to the bowl-shape you can see in the debt service numbers between 1991 and 1995. I pointed out that the earlier bowl-shape immediately preceded the good years of the latter 1990s, years the economist Robert J. Gordon called a macroeconomic miracle and that others have called "the Goldilocks years" and "the new economy". I said I expect the economy to get good again soon, as it got good during and after the "upward to the right" years of the 1990s bowl. I even said I expect labor productivity to get good again.

Maybe I am completely crazy.

Anyway, I didn't just eyeball the debt service trend. I used Excel and got a prediction in numbers. More recently, just looking at graphs, I happened to look at household debt relative to debt service. It shows a big, sweeping, upward curve. But the increase does appear to be slowing: the curve is concave-down since about 2002. I thought Oh, I can see where this line is going. I could make a guess about the future path of household debt relative to debt service.

I already have a prediction for debt service. If I can also get a prediction for the "household debt to debt service" ratio, then I can come up with a prediction for household debt in billions.

That's big. Because if I can predict the path of household debt, I think I can predict GDP and we can talk about vigor.

Well... it would be a prediction based on other predictions that essentially emerged from extending two or three lines out into the future. That doesn't seem to merit being called a prediction of GDP. Okay. But I can guess where GDP will go over the next few years, and I can put some numbers on it.

After that, instead of simply saying "I expect vigor" I will be able to tell you how much vigor I expect and when I expect it. It's the kind of prediction we can look at later and say "Art was wrong" or "Art was right". And either way, that would be a satisfying conclusion to this month-long story.


I have to break this thing up into several posts. I'd love to do it all at once. But I think each piece of it could take a couple days or more to get right.

And you should know, there is no GDP prediction till the end. I can't tell you now how things are going to turn out, because I have not done the work yet. I'd love to say there is a big prize at the end, but we won't know till we get there.

You'd think being able to say "Art was right" would be a big prize. Yeah, it would be. But we're not that far along. The last number I'll get is the GDP number. And when we look at it, we will be able to say nothing more than whether the number is reasonable.

After that, we get to wait and see how good my guess turns out to be.

No comments: