Review of David Merkel’s Analysis of ROE During the GFC

Link here.

Abstract:

From 2005-2010, the change in public company returns on book equity [ROE] was wrenching during the financial crisis. The results were uneven by sectors, and even by geography, for stocks traded in US equity markets. This paper looks at the differences, and attempts to explain why there was so much variation by sector and geography. After that, the paper attempts to explain the correlation between changes in ROE and stock returns, by year, sector, and geography.

In a world in which I didn’t have only 20 minutes to read, analyze and write about this paper, I’d like to think through his model choices. I would feel much more comfortable on this point if he accepted the Russ Roberts Science challenge and have a section discussing the process by which he arrived at the process by which he arrived at his conclusions.

Aaaanyway, the paper is interesting in that it identifies some interesting countries (Mexico? Israel?) that had companies that did very well during the crisis. Another interesting thing is that he decomposes the performance of individual US States but immediately discounts the conclusion by saying that the location of these corporates are due to historical accident:

To some degree, historical accidents help explain why some states have high contributions to returns on equity, and others low contributions. Washington State has Microsoft, Amazon, and Costco, all of which started out there. Michigan has General Motors, Ford, and Chrysler; the automobile industry has long been a big part of the state economy.

The contribution to ROE of Arkansas can be entirely attributed to Wal-Mart. Washington, DC can largely be attributed to Danaher, though Fannie Mae pulled the contribution to ROE down considerably as it failed in 2008.

The results of Kansas are dominated by Sprint Nextel, which has been a weak competitor in wireless telephony, though YRC Worldwide also had some impact on the low contribution to ROE as it was too acquisitive heading into a major recession. Virginia has many strong companies, but Freddie Mac pulled the contribution to ROE down with it failure in 2008.

Companies don’t move often, so attributing the differing contributions to ROE to state policies is unlikely. In the extreme cases listed above, all of the companies listed had been headquartered in their respective states for a long time, and most had been started there

I’d have two comments:

1. What’s the point of decomposing them, then?

2. Can’t you just attribute ALL variance of corporates to ‘historical accident’? Can there be no policy implications?

On point #2, I’d defend Merkel by saying that policy implications need a big enough sample that you can reasonably hold other factors constant. You’d need a dataset of every industry in every state over every conceivable macro-economic environment, then control for those other factors. Same applies for analyzing different countries.

But, you might say, every industry isn’t in every state! Yep, that’s why this kind of analysis is probably better classified as ‘interesting’ than ‘science’.

He probably should have left the geographical component out if he (rightly) concluded that there aren’t any policy implications. Or at least chose a different basis than political geography: how about companies on coasts vs inland? High vs low altitudes? Near vs far geographically from ‘bad’ industries (like financial services)?

Anyway, none of the criticism is a knock on Merkel who is a first class analyst with a first class blog.