Financial commentators use metaphors* all the time. Derivatives are “financial weapons of mass destruction,” for example. Actually, people use metaphors all the time. But what is the substantive content of a metaphor? More technically, if A is (like) B, then why should we believe that A has some attribute that B has?
I’ve been meaning to write about this for a while, but then Robert Shiller handed me a perfect example in the Financial Times, in his “defense of financial innovation:”
“The advance of civilisation has brought immense new complexity to the devices we use every day. A century ago, homes were little more than roofs, walls and floors. Now they have a variety of complex electronic devices, including automatic on-off lighting, communications and data processing devices. People do not need to understand the complexity of these devices, which have been engineered to be simple to operate.
“Financial markets have in some ways shared in this growth in complexity, with electronic databases and trading systems. But the actual financial products have not advanced as much. We are still mostly investing in plain vanilla products such as shares in corporations or ordinary nominal bonds, products that have not changed fundamentally in centuries.
“Why have financial products remained mostly so simple? I believe the problem is trust. People are much more likely to buy some new electronic device such as a laptop than a sophisticated new financial product. People are more worried about hazards of financial products or the integrity of those who offer them.”
The point of this metaphor is to convince you that, since houses and consumer devices like laptops have become immensely more complex in the last hundred years, financial products should as well – and the fact that they have not is a problem.
This is a perfect example of a misleading metaphor. No one today would want to live in a house from 100 years ago (not the house itself, but all the stuff in it, is what Shiller means), nor do we want to give up our laptops. Because most of our financial products were around 100 years ago, we must be missing out on all sorts of potential improvements. But nowhere does Shiller show – or even argue – that there is some underlying feature of financial products that makes them like technological products in this respect. In technology, for example, we have Moore’s Law – the observation that every 18 months (originally two years) the achievable density of transistors doubles – which implies that products can get smaller and cheaper. Shiller makes no equivalent claim for financial products. The point of our Democracy article was to argue that there is in fact no equivalent for financial products, because financial innovation is fundamentally different from technological innovation. You may not agree with us, but at least we argued the case, instead of relying on a metaphor.
I think metaphors are a great way to illustrate abstract concepts, especially to beginner audiences. I use them all the time. But their value is solely illustrative; they don’t ever prove a point. If you say A is like B, and you want to show that A has some attribute that B also has, you have to prove it without reference to B. A good example is Paddy Hirsch’s video comparing a CDO to a pyramid of champagne glasses; there, he walks you through why a CDO has the properties of a pyramid of champagne glasses, instead of simply asserting it.
Getting back to Shiller’s article, this is where he goes after the passage quoted above. (Sorry for the long quotes, but I want to be clear that I’m not leaving things out to help my case.)
“The problem is that financial breakdowns come with low frequency. Since flaws in the financial system may appear decades apart, it is hard to figure out how some new financial device will behave. Moreover, because of the low frequency of crises, people who use financial instruments often have little or no personal experience with the crises and so trust is harder to establish.
“When people invest for their children’s education or their retirement, they are concerned about risks that will not become visible for years. They may not be able to rebound from mistaken purchases of faulty financial devices and they may suffer if circumstances develop that create risks that could have been protected against.
“Thus, to facilitate financial progress, we need regulators who ensure trust in sophisticated products.”
Note the rhetorical flow. Shiller has deflected our attention away from the real question – is financial “progress” good? He has taken care of that with his metaphor, and now he can deal with a different question – why we haven’t had as much financial progress as he would like.
Then he goes from there to argue (this time it is an argument) that people are overly cautious about financial products – something that would come as a surprise to the millions of people facing foreclosure because they didn’t understand the mortgages they purchased. This is another clever device: use a good example (I think Shiller may be right about excess conservatism in long-term investments) to try to prove a broader point (but he is wrong about other financial products like Option ARM mortgages or reverse convertibles).
These clever rhetorical moves take him to his conclusion: regulation should promote the use of sophisticated products.
From there, the article actually gets better, because Shiller gives us examples of areas where he thinks financial innovation would be good. And here I don’t really disagree with him that much. I agree with him that reliance on housing as an investment vehicle is bad. I don’t really agree that target-date mutual funds are such a good idea (since as far as I can tell the conventional wisdom about switching from stocks to bonds as you age is the equivalent of an old wives’ tale), but they are probably better than many of the things people are currently invested in. Retirement annuities, another thing he recommends, would definitely be useful if you could get them at a decent price. (I believe now they suffer from a significant adverse selection problem.)
For the sake of argument, I am willing to concede that these are useful innovations that would make people better off. But you cannot use three useful innovations to argue that, in general, financial innovation should be promoted. You have to do one of two things. You could make a conceptual argument that shows that, in general, financial innovation is a on balance good thing (benefits exceed costs). Or you could make a list of all the financial innovations that have taken place in some domain in some period of time and show, empirically, that their actual benefits exceed their actual costs. But Shiller does neither; he uses a metaphor instead of a general argument, and he cherry-picks his examples without mentioning those toxic innovations that exist solely for investment banks to separate money from their “high net worth” clients. I’m disappointed, because Shiller is (a) brilliant, (b) right about a lot of things, and (c) clearly not in the head-in-the-sands, efficient market, financial-markets-are-always-right camp (much of his career has been spent debunking that nonsense).
I’m still looking for a defense of financial innovation that rises to this basic standard.
* Warning for sticklers: I’m going to use “metaphor” and “simile” interchangeably. If that bothers you, stop reading now.