Wednesday, April 29, 2009

Flu risk in perspective

The current swine flu outbreak is frequently compared to the 1918 Spanish flu pandemic. An article (that I found via Boing Boing) by Stephen Hume in the Vancouver Sun puts the individual risk of the Spanish flue in perspective:

It's estimated that about 28 per cent of Canadians and Americans contracted the Spanish flu. Worldwide, an estimated 2.5 per cent of the sick died of complications, which made the pandemic one of the most lethal flu outbreaks in recorded history. Certainly it was one that imprinted itself upon human consciousness for several generations.

But there's another way to look at those statistics. You might observe, for example, that they mean that even during the worst ravages of the 1918 flu, 97.5 per cent of those infected survived and recovered. Or that 72 per cent of the population -- even in the absence of the sophisticated public health planning and infrastructure that Canada and the U.S. have since built -- was not infected during the pandemic.

A pandemic on the scale of the 1918 flu is clearly a public health threat. With the same rates of infection and mortality on the current U.S. population, this would imply 300M * 0.72 * 0.025 = 5.4 million deaths. Yet the risk to the average individual is 0.72 * .025 = .018 or 1.8%. The risk of the current swine flu outbreak is not known, and neither Hume no I mean to imply that the risks are comparable.

In a typical year flu kills about 30,000 people in the United States, mostly the elderly.


Saturday, April 04, 2009

Possible improvements to the Geithner plan

An interesting idea to improve Treasury Secretary Geithner's plan to create a market for toxic asests held by American banks, from an Italian blog by Sandro Brusco.

The big problem is that major banks hold lots of mortgage-backed securities of uncertain value. Geithner's plan essentially puts up a big subsidy from the US government to private entities that purchase these assets from the banks. The banks are reluctant to sell, claiming that the value of the assets is much higher than anyone is willing to pay to buy the assets from the banks. The Financial Accounting Standards Board this week weakened rules that required banks to mark to market (i.e., value assets held on their books at current market prices). Many think that some big banks could become insolvent if they did write down the value of these assets to current market valuations and that the banks therefore are not anxious to sell any part of these assets and establish a clearer market price.

One of many concerns with the Geithner plan is that the subsidy to private buyers of the toxic assets will raise prices well above fair value, and banks will sell only their most distressed assets. This is a version of the lemon problem and creates an incentive for adverse selection. Brusco proposes that Treasury should require managers of the banks to back up the value of the assets they sell by investing a significant part of their salaries in the assets that they sell to government and the government's partners. Then, if the assets are overpriced, the managers will lose money. Since the bank managers are the ones who know the most about the toxic assets, this will reduce the incentive for adverse selection.

I'd like to take Brusco's proposal one step further. Require all the banks to assign a value to their toxic assets and allow them a period in which to sell at those prices, say three months. At the end of that period bank managers must use a large chunk of their pay to buy shares in those assets based on the bank's stated prices. Bruno suggests that Treasury set some income level and require all executive pay above that level be used to buy shares in the assets that are purchased with federal subsidies; i.e., the bank transforms part of their paycheck into shares in these assets. I'd suggest doing the same with my plan, except part of the managers' pay is converted into shares of the assets the bank holds onto, converted at the values for the assets the bank claims on its books. This could provide a disincentive to overpricing the assets.