A Taxing Blog

Friday, August 01, 2003
Terrorism Insurance: It does exist, sort of  
A reader pointed me to This article on CFO.com describing the current options. Lloyd's, AIG and Berkshire Hathaway will write you a stand alone policy, but it might cost a company $5 or $10 million a year to get a policy with a cap of $50 to $150 million.

As another reader pointed out to me, the market sometimes does recognize the "fat tail" phenomenon, and when it does, it prices accordingly. The example given to me (if I understood it correctly) was in the options market, where far-out-of-the-money options are priced higher than Black/Scholes would dictate.

Here, the fat tail means that the 10:1 odds that the insurance market is laying out in stand alone terrorism policies does not mean that they think there is a 10% chance of another Sept 11 happening in any given place in the U.S. Rather, it reflects the inherent difficulty in assessing the risk.

I'm usually reluctant to blame risk aversion for corporate decisions, but here it sure looks like a cognitive bias is in play. A company's decision to buy terrorism insurance seems economically irrational to me given the caps on the policies and the price. But surely no corporate manager would be fired if she bought insurance, while she might if she didn't and something bad happened. I suppose we could characterize this as an agency costs problem, but risk aversion or loss aversion seems the more accurate term here.

Defending the Terrorism Futures Market  
Slate has an article claiming that the Pentagon's decision to cancel the futures market is cowardly and dumb. Thanks to Tyler Cowen at the Conspiracy for the pointer.

I find the program interesting to consider, as it ties in to research that I've been doing on just how far we can push the efficient market hypothesis.

To begin, I don't find the concept of betting on terrorism morally reprehensible. Suppose the owner of the Landmark Building in the City of Metropolis takes out an insurance policy against terrorism. It pays a premium of $10 million a year, and if a terrorist act occurs that shuts down the building for a month or more, the insurance company pays out $500 million. The Landmark owners have essentially made a bet that a terrorist act will occur, and they are getting 50:1 odds. The insurance company would presumably write similar policies for dozens of other buildings, thus spreading the risk. (N.B. if a futures market existed, it would make it easier for the insurance companies to hedge.)

The problem with the terrorism futures idea is that markets are not very good at predicting highly unusual events. For example, the capital markets went into a major spin in about five years ago when Russia defaulted on its sovereign debt. For many years the conventional wisdom was that nuclear powers simply do not default -- it was unthinkable. Argentina might default, but the US, UK, France, Russia, never. Markets like to think that the likelihood of events looks like a bell curve, but often the oddity of human behavior makes the impossible (or highly, highly unlikely) possible. Life is not a bell curve. Or, as the economist Eugene Fama stated, "Life always has a fat tail."

I'm not sure if insurance companies are writing policies on terrorist acts. I suspect they are not, but rather are careful to carve out terrorism in force majeure clauses. (If anyone knows the answer to this, let me know.)

If insurance companies are indeed refusing to insure against terrorism, then this is evidence that the market believes that the risk is impossible to quantify. Returning to the example above, suppose the insurance company writes policies on fifty major landmark buildings in the City of Metropolis. If a terrorist act destroys one building, the insurance company is okay, as all the accumulated premiums from other buildings spread the risk. But if a major terrorist act shuts downs all of Metropolis, and all fifty buildings file legitimate claims, the insurance company will not have the money make good on the claims.

Perhaps a terrorism futures market would help this situation by creating a reinsurance market so the insurance company could hedge. But to make such a market work, we would need private sector experts who believed they could properly quantify the risk of highly unlikely events --- and yet those private sector actors would not access to the most relevant information (classified intelligence and national security information). This is like asking M&A arbitrageurs to bet on the likelihood of a merger without allowing them to read the relevant SEC filings or talk to anyone on Wall Street.

As it is now, we each have a very small incentive to sniff out terrorism. (If anything, we are probably spending too much resources on casual anti-terrorism measures, like reporting mysterious bags on the subway platform.) A terrorism futures market would have the benefit of concentrating financial incentives on a handful of private reinsurance actors, who then would have the financial incentive to become experts in assessing the risk. And the government would be able to free ride on this expertise.

In the end, I don't think the terrorism insurance market is deep or liquid enough for the program to make sense. Markets are good at predicting somewhat unlikely events (e.g. frost in Florida in March), not highly unlikely events (e.g. Russia defaulting on its debt). And the "positive externalities" here are slim, as the market participants would have a hard time developing intelligence and monitoring behavior in the way that insurance companies normally do.

Besides, if you think the Ashcroft-led DOJ has little regard for privacy and nondiscrimination principles, just imagine how private sector investigators (not subject to the civil rights laws) would act.

Wednesday, July 30, 2003
Math and the Stock Market  
Driving to work today, I heard a great interview on NPR with John Allen Paulos, author of A Mathematician Plays the Stock Market. I haven't read the book yet, but it sounds like a good intro to understanding how our cognitive biases blind us to the randomness of the market. For example, Paulos talked about our natural tendency to seek out information that confirms decisions that we've already made and ignore contrary data, even if our original decision was tentative.

Of course, one shouldn't place too much faith in the power of math, either. I just finished reading Roger Lowenstein's When Genius Failed, about how Merton and Scholes and the rest of the gang at Long Term Capital Management bet a little bit too much on the efficient market hypothesis.

Tuesday, July 29, 2003
Enron Settlement  
NYT story on the Enron-related settlement between the SEC and JP Morgan Chase and Citigroup. The banks acknowledged that they were wrong to rely on technical compliance with the accounting rules that allow special purpose vehicles to disguise what are effectively loans, but are not treated as such on the balance sheet.

I really have to wonder if the banks are just saying what the SEC wants to hear. It would take a major change in corporate culture for investment banks to assume the role of watchdog.

There is a parallel here in tax, of course. Will investment banks now shy away from participation in tax shelters that meet the technical requirements of the Code but might fail under the economic substance doctrine? I doubt it.

And perhaps they should not. Bankers generally have no particular expertise about whether a carefully structured transaction satisfies the relevant tax and accounting rules. Enron may be a special case, as the number and amount of the transactions were staggering. But in general, the lawyers and accountants should be the ones feeling the heat. And, at least as far as I am aware, the SEC has done little to put pressure on the lawyers to think harder about whether they should easily offer up opinions on SPVs.

One final note: pressure from the SEC on law firms might be more effective than pressure from the IRS. In these borderline cases the worst the IRS can do, I think, is impose a small penalty (and usually on the client, at that). If the SEC can make out a case for fraud, there is the potential to eliminate a law firm's ability to practice in front of the SEC --- an enormous penalty for most large firms that center their practice around the capital markets.

Monday, July 28, 2003
Pension crisis?  
NYT story on the possibility of a looming pension crisis. In short, it seems more and more likely that we might need to bail out the PBGC, the quasi-public agency that guarantees the retirees from private companies will get the pensions they were promised.

The story draws an apt comparison to the S&L crisis of the late 80s. The PBGC is funded by contributions by private companies. But like S&Ls, we have what academics like to call a "moral hazard" problem. If the PBGC fund runs out of cash, the government will kick in general revenue money to make up the difference. And if pensions are effectively guaranteed by the government, then companies have an incentive to gamble a bit with their pension plans by investing in higher risk securities (equities and certain hedge funds). After all, if things go well, the company retains the residual upside, but if things go badly, the government will still pay the pensions. Thus, aggressive pension fund managers are acting exactly as we would expect "rational" economic actors to act.

Urban Institute / Brookings on the AMT  
The Tax Policy Center has posted a new paper on the AMT. One sentence in the intro best summarizes why we should care: "This article examines how a tax that was originally aimed at 155 taxpayers could grow under current law to target 33 million."

Thanks to Brad DeLong for the pointer.

Search WWW Search taxpolicy.blogspot.com