What industries have the largest percentage of their assets measured under fair value standards?  Whisper your answer to the person sitting next to you before clicking to see the answer, courtesy of this paper by Gartenberg and Serafeim of HBS.


Click on the graph to enlarge it.  I was surprised to see the high proportions in the hi-tech firms.  I think the graph is a bit misleading, because so many tech firms have very high market-to-book ratios; a graph of the fair value assets as a percent of market values would probably tell a very different story.

I am trying to figure out if I am surprised by what the authors deem the central contribution of their paper.

We calculate the lagged (disclosed at the end of the third quarter) ratios of firms’ assets measured at historical cost and at fair value levels 1, 2 and 3, in accordance with recent reporting changes, to the total assets of a firm. The central contribution of our paper is to relate these lagged ratios to subsequent stock market performance of firms in the 4th quarter. In broad terms, if the arguments are valid that fair valuation contributed to depressed equity prices, we would expect, all else equal, that firms holding relatively higher proportions of assets measured at fair value should experience worse abnormal returns in the 4th quarter. Conversely, if arguments hold that asset opacity contributed to the sell-off, we would expect firms holding relatively higher proportions of assets valued at historical cost to perform worse. We test these opposite predictions and find results that are inconsistent with critics of fair valuation: namely, that firms with higher proportions of fair-valued assets experienced higher abnormal returns. This finding supports the assertion that firm opacity – as measured by assets held at historical cost rather than at fair value – contributed more to the sell-off in the 4th quarter of 2008 that did fair valuation.

They do a reasonable job of showing that the result isn’t specific to particular industries, or driven by small firms.  But I still don’t quite know what to make of the result.   That firms with fewer fair-valued assets experienced negative returns as they were forced to take 4Q 2008 impairments that were already recognized by their peers as downward FV remeasurements?  Perhaps firms with lots of fair value assets used their discretion to biased FV estimates upward, and garnered positive returns the old-fashioned way — through earnings management.