Robert Shiller talk

Economist Robert Shiller speaks at Google, 3/15/2005What a nice place to work. Monday, I attended a talk by Colin Powell. Today, Yale economist Robert Shiller (author of _Irrational Exuberance_) came to talk about risk. He made the point that in modern society, we have both greater risks, and greater ability to deal with those risks. An example of the former is that we are living longer, and income inequality tends to increase with age. Another is that replication technology leads to more "winner-take-all" professions, ie there are far fewer local singers now, because we listen to the best in the world via cheap, high-fidelity digital copies. This benefits listeners, because they can hear the best, but it does increase the risk of getting involved in the profession. I've discussed earlier how the digital world affects our self-esteem, but this downside was new to me.

On the other hand, financial devices for spreading risk are also easy to copy, and he wants to help bring the world of available methods more in line with the world of potential good methods implied by modern financial theory. He proposed a number of interesting suggestions for (mostly private-market) ways to spread risk. This is a description of and disagreement with parts of an hour long talk, so its a bit long and lacking in detail, but fascinating if you are interested in such things:

21st Century Ways to Spread Risk:

  • Livelihood Insurance: A long-term insurance policy based on an index of occupational salaries. The policy would insure you against a decline in your profession's income. Its price would indicate the risk of taking on that profession (the downside risk). And unlike insurance of your income, there is no individual moral hazard.


    The negative side I see is the concentrating of dispersed interests - a US-only software engineer insurer is going to be very interested in lobbying against offshoring. In a sense, you are trading the employees moral hazard for the insurance companies moral hazard. Yeah, not paying off is going to give you a bad rep and lose you business. But protecting your industry? Many may see that as a positive thing. When I brought this up, Shiller agreed that this might happen, and its an example of the many potential and sometimes hard to imagine downsides to these devices. One might imagine that an insurer would likely insure a wide variety of industries, and thus lobbying for protecting one would hurt all the other ones, so perhaps its not an issue.
  • Markets for new types of risk:
    • Home Equity Insurance: The big monetary risk of a home nowadays is not fire, it is market movements. Shiller has founded a fund to offer securities that can be used to protect against home equity risk.
    • Country Income Indices: Why not sell your country short, and buy the world, in order to reduce your country-associated financial risks?
    • Longevity bonds: (whose payoff depend on the longevity of a large demographic group) to hedge against age-distribution-related risks.
  • Milton Friedman, in _Capitalism and Freedom_ (1962) suggested that individuals sell shares in their future earnings, but was afraid that it would be difficult to track people and enforce. The most obvious way to use this is to pay for education to increase your future income - and its an example of how you can fund education in a libertarian society. Apparently Australia's government is doing this. And there is a private US group, myrichuncle.com, doing the same thing - with different loan terms for different professions. They hope to combine large groups of such loans into securities. Not only does it help fund students, but it gives them information about what the marekt consensus of the future earning power of various careers.
  • Inequality Insurance: Index the progressivity of the tax system to income inequality. It would be mathematically defined, rather than legislated, to help maintain some given level of income inequality. While I think such redistribution is generally a bad idea, I find it more philosophically pure to at least be maintaing some mathematical constant than doing it all by legislative fiat.
  • Intergenerational Risk: Why do we have public Social Security, rather than private? Shiller suggests that its because private insurance has trouble dealing with such long lags (ie unborn children can't sign insurance contracts - yeah, but their parents can). He makes a family analogy, where the family unit acts to spread intergenerational risk (rich people give more to their parents and children), so why shouldn't the government? Given his mention of moral hazard earlier, this seems to ignore the vast gulf in moral hazard between a unit of one or two dozen and a unit of one or two dozen tens of millions.
  • Shiller dropped an inaccurate quote describing Bush's privazation plan, which he doesn't like, it sounded like a memorized sound bite: "Basically, the president wants us to buy more stocks on margin" (because the money diverted from SS to private accounts is borrowed from the government at 3%, with your prospective SS benefits as collateral). But we aren't buying *more*, we are *switching* from buying SS to buying stocks on margin, so what matters is the difference between them. He also (correctly) blasted Bush for picking a specific mix of stocks/bonds and a change of that mix over time ("life-cycle funds"), one which is not particularly well-designed even for the average person, and obviously is going to be a misfit for anyone with unusual circumstances.

  • International Agreements for Risk Control: World leaders, and the World Bank, should be pooling risks across countries. ie GDP swaps. This is being done somewhat in the EU. Argentina proposed something like this at an IMF meeting. They even proposed that rather than selling debt, they sell warrants on their own GDP, but no one accepted because it was too newfangled a concept.

Audience questions:

  • "Isn't selling income, like 80% or 100%, tantamount to slavery?" Obviously this depends on the terms of the contract. Shiller points out that if you are only taking a share of income, w/o the right to force someone to work, then its not slavery at all. (We might comment that a great many people give 40% of their income to the state every year, without seeming to consider it slavery). He points out that govt indices of incomes BS, but these guys are putting their money where their mouth is by choosing loan rates based on their major, SAT scores, and other information.
  • I asked why he claimed that lower costs of capital movement led to increased income risk. Doesn't that smooth worldwide income distributions by moving business to poor countries? He said that income inequality had decreased in the past decade because of China, but that it increased risk within rich countries like ours. But if you care about income inequality, isn't reducing the risk that a new soul will be born poor vastly more important than the risk that a rich person will move down to the world average? It seemed strange to praise income equality while bemoaning its effect on those at the top - can you have it both ways? I guess if you increase wealth for the bottom end, rather than transferring. But globalization does plenty of the former.
  • Shiller is most famous for having correctly called the dot-com bubble. Apparently he has now called a real-estate bubble, and was asked why. It's in the new edition of Irrational Exuberance. He said that there was no bubble in places with land - they build more houses. But places that are supply-limited tend to "leave prices unmoored". And home buyers are amateurs, with an irrational belief in what the future home price trend will be. He has studies confirming this, where people's predictions of trends were absurd (ie mean 22%/year, median 10%/year, in LA, over the next 10 years). I guess it seems much simpler to me: isn't the obvious way that supply-limitation affects prices through the law of supply and demand? I mean, there may well be effects beyond that - I somewhat buy the bubble theory, myself. But you gotta take the basics into account first.

    During this explanation, he gave the foolish explanation that real estate long-term returns are so poor because housing is reproducible. But that's not why - it's because its an unproductive asset. Your house is just sitting there, providing you a home, not transforming stuff into more stuff like a business.

  • Shiller was asked a question which hinged on the Efficient Markets Hypothesis (Ray Kurzweil is starting a fund - will it do well?). He thought that smart money could beat the market - but its an awful lot harder than you think.
  • An audience member asked whether, if we believe we are in a real estate bubble, and there are securities, whether we can just sell houses short to make money and correct the market. Shiller points out that you can't easily short sell, because you have to borrow a house and then sell it, and then buy it back. Because houses aren't identical (unlike shares), the buyer of your particular house could just ask for ten million bucks (? the ultimate short squeeze ?). The obvious solution I see is to phrase the contract in terms of equivalent houses. You don't have to pay off your debt of 1 house with exactly that house, but with an equivalent house. Shiller did say that the existence of more sophisticated financial instruments will let this happen.

    I found the question from the audience profound, as a demonstration of how self-interest, and an understanding of finance (or smart gambling - take your pick, its the same thing), lead to correcting such imbalances.

Hope you all found this interesting - the talk was just as rushed as this post, but full of neat ideas.

Share this