By James Kwak
I’ve had Robert Litan’s recent paper defending most financial innovation (the web page doesn’t tell you much; you need to grab the PDF) on my to-do list for a while now. I wasn’t looking forward to writing about it, since I’m a bit tired of the subject, and I don’t think I have much more to say. So thankfully Mike Konczal beat me to it, in a two-part series. Part I is really brilliant, and has not one but two insights. The first (to simplify) is that we generally think of innovation in products as making them simply better on all dimensions. We don’t realize that, with most new financial products, we are just getting to a new point on the risk-reward spectrum that wasn’t there before. Now, it might be good for the economy as a whole for that new point to exist. But as consumers, we don’t realize that the good properties of a new financial product are almost invariably counterbalanced by some bad properties.
The second insight is that real, good financial innovation does not look like a new product; it looks like a new way of dealing with an existing product. Konczal’s example is TRACE, a recent system for increasing transparency in the market for corporate bonds (you’ll have to read his post for a more complete description). The effect has largely been to make pricing more transparent and reduce spreads, which is good for investors. More broadly, as Felix Salmon said somewhere (probably many times), financial innovation should show up as lower prices for all the bread-and-butter financial products–equity and debt underwriting, interest rate swaps, etc.–not has higher profit margins for dealers.
Konczal’s Part II asks some more general questions about Litan’s results. I have some different questions.
Litan grades each innovation on three dimensions:
“The table illustrates that financial innovations are appropriately measured or ‘scored’ on three dimensions, of which the net impact on productivity or total output is only one. Financial innovations also have distributional impacts – for example, by expanding access to certain products (loans and investments) – and can affect convenience of the users of financial products and services.”
Litan calls these access, convenience, and productivity/GDP. And here are his scores:
But I don’t really understand these categories, especially the first two. Take “access.” On its face, it seems hard to imagine how a new financial product could reduce “access,” whatever that word means here. So you would expect the kind of grade inflation you see in Litan’s chart. Sure, I see how ATMs make it easier to access your money. I see how index funds make it easier for retail investors to “access” smart investment strategies. But what about CDOs? Litan says that they made it easier for people to access mortgages:
“While it worked, the CDO thus was instrumental in expanding access to mortgage financing to a large class of people who previously could not buy a home. We know now, of course, that many of these subprime borrowers never should have been offered these loans, especially with little or no down payment and with little or no documentation of their incomes (or lack thereof). But some portion of those who received subprime or Alt-A mortgage financing – we won’t know the exact share until the foreclosure crisis has run its course – clearly benefited from it. Thus, on my access and convenience measure, I am inclined to give the CDO a qualified, temporary ++.”
I agree with Litan that CDOs increased access in this sense. But I don’t think it was a good thing to give people access to mortgages they couldn’t afford. Litan tries to capture this criticism by giving CDOs a — for productivity/GDP. But I don’t think that captures it. I think this increase in access was bad. So if the Access column is supposed to carry a normative judgment, then it should be negative here. Alternatively, I am willing to accept the ++ in the Access column (to mean that CDOs increased access) only if we agree that access itself is neither good nor bad; sometimes it’s good, sometimes it’s bad. But then all those plus signs under Access should no longer be read as benefits of financial innovation, but simply as descriptions (and near-tautological descriptions at that) of one effect of financial innovation.
Something similar goes for “convenience.” What financial innovation would reduce convenience? That’s why you see lots of plus signs. But is convenience unequivocally good? The convenience provided by credit cards, home equity loans, and (indirectly) asset-backed securities and CDOs sure made it easier to overborrow and lose your house. Is convenience always good? I think it’s generally good, but there’s a second edge to that sword that isn’t captured simply in the productivity column.
The weirdness of this chart is underlined by SIVs. Here’s Litan:
“Clearly, given this history, it is impossible to score the SIVs’ contribution to GDP as anything but –. Indeed, by helping banks to circumvent their capital adequacy requirements, SIVs enabled the banks to excessively leverage what capital they did have, which greatly magnified the economic impact of subprime mortgage losses after housing prices quit climbing and began to fall.
“On a positive note, by adding liquidity to the mortgage securities market, the SIVs temporarily enhanced access to mortgages by subprime borrowers and made it more convenient for them to do so. But even a temporary ++ score on these dimensions cannot come close to making up for the financial damage they helped cause.”
So Litan thinks SIVs were bad. I agree. But a casual reader looking across the SIV row will see four plus signs and only two negative signs. Maybe that’s the casual reader’s fault; but I think it’s partly Litan’s fault for creating a misleading chart.
That said, I think most of the details in the paper are good. (Although I don’t agree that venture capital is a financial innovation–for exactly the reasons Litan puts in his paper.) My main concern on that level is that Litan doesn’t always place much weight on the negative side effects. For example:
“As for their contribution to GDP, interest rate and currency swaps reduce the transactions costs of having to sell and buy the underlying loans or bonds. In addition, foreign currency swaps, in particular, facilitate cross-border financial flows, and thus, like futures and options, accelerate globalization and the benefits it brings. My judgment, therefore, is that these arrangements should be scored somewhere between a + and ++ for their contribution to GDP.”
As we now know, currency swaps also help governments hide their debt. And we’ve known for a long time that companies use both interest rate and currency swaps to massage their earnings. That’s a bad thing. Admittedly, it’s hard to quantify. But it should be in there somewhere.