Friday, January 16, 2009

VAR numbers as road signs

In a previous post there was mention of the notion that seeing a VAR number too often can cloud the judgement of the user and lead him to accept it as truth. One can make the same analogy to road signs - from this article ("Distracting Miss Daisy") by John Staddon.

...the overabundance of stop signs teaches drivers to be less observant of cross traffic and to exercise less judgment when driving—instead, they look for signs and drive according to what the signs tell them to do. ... But this is emblematic of the sort of signage arms race that has become necessary in the U.S. When you’ve trained people to drive according to the signs, you need to keep adding more signs to tell them exactly when and in what fashion they need to adjust their behavior. Otherwise, drivers may see no reason why they should slow down on a curve in the rain.

Do more road signs make roads safer or prevent more accidents?

In 1949, a British statistician named R. J. Smeed, who would go on to become the first professor of traffic studies at University College London, proposed a now-eponymous law. Smeed had looked at data on traffic fatalities in many different countries, over many years. He found that deaths per year could be predicted fairly accurately by a formula that involved just two factors: the number of people and the number of cars. The physicist Freeman Dyson, who during World War II had worked for Smeed in the Operational Research Section of the Royal Air Force’s Bomber Command, noted the marvelous simplicity of Smeed’s formula, writing in Technology Review in November 2006: “It is remarkable that the number of deaths does not depend strongly on the size of the country, the quality of the roads, the rules and regulations governing traffic, or the safety equipment installed in cars.” As a result of his research, Smeed developed a fatalistic view of traffic safety, Dyson wrote.

So should we abolish VAR numbers?

A few European towns and neighborhoods—Drachten in Holland, fashionable Kensington High Street in London, Prince Charles’s village of Poundbury, and a few others—have even gone ahead and tried it. They’ve taken the apparently drastic step of eliminating traffic control more or less completely in a few high-traffic and pedestrian-dense areas. The intention is to create environments in which everyone is more focused, more cautious, and more considerate. Stop signs, stoplights, even sidewalks are mostly gone. The results, by all accounts, have been excellent: pedestrian accidents have been reduced by 40 percent or more in some places, and traffic flows no more slowly than before.

What I propose is more modest: the adoption of something like the British traffic system, which is free of many of the problems that plague American roads. One British alternative to the stop sign is just a dashed line on the pavement, right in front of the driver. It actually means “yield,” not “stop”; it tells the driver which road has the right of way. Another alternative is the roundabout. ... A “mini-roundabout” in the U.K. is essentially just a large white dot in the middle of the intersection. In this form, it amounts to no more than an instruction to give way to traffic coming from the right (that would be the left over here, of course, since the Brits drive on the left). ... most right-of-way signs are informational: there are almost no mandatory stops in the U.K. (The dominant motive in the U.S. traffic-control community seems to be distrust, and policies are usually designed to control drivers and reduce their discretion. The British system puts more responsibility on the drivers themselves.)

The above quote leads to the suggestion that banks use their own VAR numbers to set their own capital requirements which was indeed what they did and did not prevent the crisis. A 99 percent chance that Bank A will not lose more than 50 million itoday without accounting for the overall economic condition (or correlated risk) is essentially meaningless. Perhaps financial authorities should spend more time providing direction on whether daily VARs need to be adjusted to account for economic conditions than to the numbers themselves. For instance, if an economy shows signs of some stress then authorities need to provide direction as to how VAR numbers should be adjusted.

However, this suggestion is subject to the same criticisms of whether the Fed should pop asset price bubbles. But rather than use interest rates, setting VAR limits or making VAR adjustments can trigger an overall audit of the financial sector where participants can disagree on whether the VAR limit makes sense in the same way that BCS ratings are dissected. This way the authorities are in some sense ahead of the game without causing the disruption of an increase in interest rates.

No comments: