James Kwak, discussing a recent Bernanke speech defending financial innovation and a Ryan Avent post parsing it, underscored Avent’s observation that Bernanke had trouble coming up with an example of the sort that the financial services had in mind these days (ie, novel products making use of derivatives and other risk slicing, dicing, and distribution tools). His examples were pedestrian but important consumer products like credit cards and well seasoned advances, like securitization (dating from the 1970s).
The lone relatively current example? Subprime lending. Bill Black (via e-mail) has trouble seeing that as virtuous, at least in combination with securitization (which presumably was the innovative bit):
The very last thing any sentient being should want to do is restore a secondary market in nonprime loans.
Kwak made a useful distinction that innovation that is serves to increase access to credit may not necessarily be a good thing. Note that this runs contrary to the pre-bust mantra, that more credit availability was a virtue:
Where I come from (career-wise at least), innovation meant that you invented something that people wanted, or you figured out a cheaper way to make something that people wanted, or you figured out a way to improve something that people wanted….
Financial innovation, however, comes in two forms. There are financial innovations that make our lives easier. One is the automated teller machine (ATM)…
The other kind of financial innovation has to do with extending access to credit. Here I think it’s less clear that innovation is unequivocally good.
It is certainly possible for a society to be below the optimal level of access to credit. Consider the idyllic banking paradise that gets mentioned a lot these days, in which people deposited their savings with local banks which, in turn, lent money out to trustworthy local homebuyers and held onto those mortgages to maturity. The good thing about this model is it encouraged responsible underwriting. The bad thing is that it isn’t very good at moving capital (money) from one part of the country to the other. Imagine in Iowa no one needs a mortgage, so the banks have no place to lend and can only pay their depositors 0.1% interest. In Florida lots of people need mortgages, so the banks offer 4% on savings accounts, but they still can’t attract enough cash and people who would buy houses can’t. (Or, alternatively, people who would take out loans to expand their businesses can’t.)
Yves here. Matching deposits with loans demand across regions has been a non-trivial and sometimes contentious problem in American history. The impetus for the old state banking restrictions was the emergence of national banks that hoovered up enough local deposits in some areas as to be perceived to be depriving businesses (then usually farms) of needed credit. Back to Kwak:
The effect of securitization should be to moderate differences in interest rates – mortgage rates can come down as money moves into Florida, but they may go up as money leaves Iowa – and perhaps to lower them overall by making more money available to the market as a whole. If we were in a situation where too few people were getting mortgages, this is a good thing. But it is also possible for too many people to be getting mortgages, as we now know. Something similar happened with venture capital and startups over the last fifteen years. After the IPO rush of the late 1990s, billions of dollars of new money piled into the VC industry; that money flowed to thousands of companies that should never have gotten funded, resulting in lost money for investors and lost time and effort for thousands of generally bright and well-meaning entrepreneurs…
In short, financial innovations whose sole function is to increase access to credit do not in and of themselves make the world a better place.
While this is a useful distinction, and helps advance the discussion, I think we can go a good deal further.
One of the reasons economics have been lulled into viewing the widespread proliferation of new derivatives and risk transfer products is that it helps advance a fantasy put forth in a seminal paper 1954 paper by Nobel prize winners Kenneth Arrow and Gerard Debreu. The paper gave a rigorous proof (at an economist wet dream level of elegance) of the existence of a multimarket equilibrium in a decentralized economy. However, the model assumed that there were forward markets not simply for every commodity but also for all conceivable contingencies and that no one holds money for longer than a single time period.
This paper shifted attention in the discipline away from how economies and competition actually works, that is, the processes by which prices are set and adjusted, a dynamic process, to proving that a specified set of prices could clear all markets simultaneously.
It also had the effect of dignifying the development of more risk transfer products as “completing the markets”, when markets as complete as Arrow and Debreu envisioned could never be obtained. Individuals can’t anticipate how reality will unfold far out enough to ever specify the needed contingencies (say hedging the risk that you might be ten minutes late to a client meeting could cause you to fail to win a possible piece of business). As economic historian Mark Blaug put it:
Following the methodological standard learned sitting at the feet of Arrow and Debreu, a modem economist would rather say little precisely than much vaguely. If there is such a thing as “original sin” in economic methodology, it is the worship of the idol of the mathematical rigor more or less invented by Arrow and Debreu in 1954 and then canonized by Debreu in his Theory of Value five years later, probably the most arid and pointless book in the entire literature of economics.
Note how a prototypical discussion of the merits of financial innovation, from a speech by Timothy Geithner in March 2007, assumes the virtue of more advanced hedging tools:
One of the widely presumed benefits of the last go-round of what passed for innovation in the money world was that it allowed for more sophisticated hedging and risk sharing.
Over a long period we have seen innovations ranging from the syndication of bank loans and the direct provision of credit through the capital markets, to the spread of asset-backed securities and products that separate different parts of the payments stream and different dimensions of the risk in a credit obligation into different instruments.
These changes have contributed to a substantial reduction in the share of total credit held by banks. They have produced a greater separation or distance between the entity that first arranges a loan and those who end up holding the risk, and more intermediaries in that chain. And they have contributed to a dramatic increase in the number and diversity of creditors to any individual borrower, and a greater capacity to actively trade credit risk…
But there is a mundane and more subtle reason to be skeptical of these risk transfer arrangements. The fact that some is good does not necessarily lead to the conclusion, contrary to Arrow and Debreu, that more is better.
The first is that these new products were sufficiently complex and opaque that risks were too often dumped on the hapless who didn’t know what they were buying. We’ve seen everything from German Landesbanken, Norwegian villages, Australian pension funds, Jefferson County and a host of other municipalities buying products or entering into swaps they didn’t understand. Even the supposedly sophisticated Harvard endowment pushed way out on the risk curve with lots of illiquid investments and derivatives, to its undoing. For the first set, they were given assurances by salesmen and lacked the sophistication to make an independent assessment of the risks. In the case of Harvard and many of the other risk seeking endowments, true belief in the brave new world of finance, compounded by years of apparent success, led to overconfidence, a surprisingly naive pursuit of return with insufficient attention to the downside potential.
The less obvious concern is that the view that more hedging tools (which presupposes more hedging) is ever and always a god thing assumes the users can make smart decisions about to how to use them. The classic argument, offered by Myron Scholes among others, is that business people should worry about making their business run better and lay off the risks they can shed. The argument is that hedging is cheaper than equity.
But that presupposes a God’s eye view of when to pass those risks and how far out to hedge them along. Look at the (comparatively) simple case of airlines and fuel costs. Airlines watch the energy markets intently, yet regularly hedge (as in lock in, since my impression is that most use futures rather than the far more costly long dated options would be) fuel prices, often at the worst moment. How many panicked when oil went north of $110 a barrel?
In a competitive business, getting it wrong doesn’t merely mean a quarter or two of lower earnings than you’d have otherwise (and remember, if you muff it badly, you can show net losses). No matter whether the cash flow shortfall relative to competitors that did better is small or large, wrong footing it puts the company at a comparative disadvantage. It has less in the way of funds to invest or use as cash buffers (although with the until this year of companies running lean, many would have paid it out in some combination of dividends, share repurchases, and goodies for top management, so the competitive implications in reality were probably less than in theory).
If a company has exposures of any complexity and wants to hedge, assessing the alternatives of how to execute the hedges is not trivial, particularly one decides to manage the hedges actively. Thus the idea that you can have a tidy specialization, with businesspeople simply laying off risk onto more savvy risk management types, is spurious. To do an adequate job, you need to have at least a base level of competence. Given the propensity of financial firms to take advantage of widows and orphans, a caveat emptor posture and investment in attaining a higher level of expertise seems prudent.
When does the frictional cost of having these all these savvy and highly paid risk professionals (who may overengineer matters to justify their own existence) become counterproductive relative to the real performance gains? Funny that that the cutting edge quants haven’t turned their modeling skills loose on that question.