Epsilon is really small

Essential reading from NotSneaky.

My sense is that epsilon (the effect contraction in the financial sector will have on the non-financial sector) is small in the short run and zero in the long run. There are some transaction costs in finding a new bank for “real” economy businesses if their old ones stop providing credit, but in the end we’re talking about a couple strokes on a keyboard and maybe a walk further down the block.

Is there any “micro evidence” that can help us pin down this parameter?

Also, the large dip in financial paper with no corresponding dip in the non-financial sector on the first chart here is evidence of a small epsilon.

As long as there’s at least one bank standing at the end of all this, I don’t think epsilon can get that big. So maybe the bailout is attempting to throw money at this problem hoping to avoid this non-linearity.

In the long run, epsilon is zero as everyone in the non-financial sector has found alternative sources of credit. The fact that a bunch of smart ex-bankers will become smart middle managers in the real sector is gravy. But because someone was stupid-efficient at moving electronic bits shaped as dollars around doesn’t mean they’ll be good at moving electronic bits shaped as goods and services and it probably means they’re just plain bad at moving atoms around (although, Bankers tend to work out a lot).

Using NotSneaky’s handy chart, I predict a recession of about 2% peak to trough.

UPDATE: We’re on our way. GDP down 0.3% last quarter.

2 thoughts on “Epsilon is really small”

  1. Well, I don’t think epsilon’s zero in the long run, except as some local maximum. Otherwise that would mean that the financial sector doesn’t matter at all which I don’t think is true. And certainly the Demirguc-Kunt and Levine papers have shown that the size and efficiency of financial markets has an important impact on the long run growth rate. But I do think the implicit size of epsilon has been exaggerated in some of these panicky stories.
    It’s more like if the share of financial sector in the economy deviates from some optimal value (whatever that is) then epsilon will be non zero and either positive or negative. And much of a commentary on the present crisis suggests that this share is currently too high – which implies a negative epsilon in the long run.
    Basically it’s a bit like lawyers (roughly the same people, too) and the judicial system. You need some of them around to enforce contracts, establish rule of law and guarantee property rights, but you get to many of them and they start chasing ambulances or suing McDonalds for serving its coffee hot (I know, I know). Just like we need fewer lawyers, we need a smaller financial system – but we do need some.

    And thanks.

  2. Well, I guess I don’t mean *finance* is unimportant in the long run. I just mean what we identify as the *finance industry* isn’t important in the long run. Businesses will get financing, but it doesn’t have to be through banks with faux-greek columns out front and smiling tellers. This crisis is in the industry we currently identify with those banks; its not a crisis that affects future entrepreneurs that want to make money being intermediaries.

Comments are closed.