Saturday, June 28, 2014

Reflecting on Context

As the recent posts have been considering context when evaluating economic data, and as my Saturday morning post is not yet (Friday evening) writ, it occurs to me to "reblog" an older post on the topic:

Federal Debt Held By Federal Reserve Banks

F, D, H, B, F, R, B. It's easy to remember: Just look at the title of this post. FDHBFRB is the name of a FRED dataset.

There is another dataset too, with the same letters, plus an N at the end. N is for "new", I suppose, suggesting that the original set of capital letters represents a discontinued series. It does.

Here's what I looked at first:

Graph #1: Discontinued (blue) and New (red) Fed Holdings of Federal Debt, Relative to GDP
Fed Holdings (of Federal debt) relative to GDP. What's wrong with this picture?

Debt went up, that's what's wrong. Debt went up *way* more than GDP since the 1950s. GDP is a small denominator. GDP makes Federal Debt Held By Federal Reserve Banks look bigger than it is.

Here's a better measure:

Graph #2: Discontinued (blue) and New (red) Fed Holdings of Federal Debt, Relative to TCMDO
Fed Holdings relative to Total Credit Market Debt Owed.

By the standard of the 1950s and '60s, Fed Holdings since the 1990s should have been twice as high as they were. By that standard, Fed Holdings are low yet today. Even with that big spike there at the end.

Fed holding should be twice as high, or Total Credit Market Debt Owed should be half what it was since the 1990s. Or some combination of the two.

This is the Arthurian policy recommendation: more Fed holdings, and less credit market debt. Why? Because you can use credit for just about everything these days. But you can't use credit to pay off debt.

Federal Reserve holdings of Federal government debt is a measure of how much money the monetary authority has put into the economy. What should we compare that to? GDP? Why? Because we always compare everything to GDP? I need a better reason.

I compare those holdings to the amount of debt that the banking system has generated from that money. This is not a context at random. It is the most relevant context possible, from my point of view.

1 comment:

Jazzbumpa said...

What should we compare that to? GDP? Why?

It's a good question, really. GDP is easily accessible, people have a fair idea of it's meaning, and it is relevant to a discussion of the national economy. So it's inviting, maybe even just force of habit, and only the occasional skeptic will protest.

But you always have to be careful with denominators. I'm sure you remember me harping on this repeatedly.

Your comparison to total debt is also relevant - and revealing.

It comes down to what kind of a point are you trying to make, and how do you want to make it.

The hazard is there are denominator choices that skew rather than clarify, and dishonest people can use that to obscure the truth.

I think a good approach is to have some coherent rationale for choosing your denominator, always be skeptical of your own conclusions [damn, is that one hard] and try looking at it a different way to see if that suggests a different narrative.

Looking at how YoY % change moves over time can be helpful. The data series provides its own context.

I don't think there's a perfect way to look at data. You just have to be agenda-free, try different approaches, and see what the data suggests.