In our next session of office hours (Tue, June 16th, 4pm ET), we will have our second brainstorming session.

For those of you who couldn’t make the first brainstorming session, Bob Lipe led us in a discussion of research ideas related to leasing.  As in any good brainstorming session, Bob wasn’t just giving away research ideas, but I for one came away with a research idea that I’m going to pursue with one of the other session participants.  I’m sure others did as well!

This time, I will lead a discussion on topics related to fair value.  In my opinion, proponents and critics of fair value have had much to say recently about fair values.  In Tuesday’s session, we will consider some recent assertions that have been made (e.g., by Board members, academic researchers, and others).  My intent is to ask whether these assertions have existing empirical support and, if not, discuss how various methodologies might be brought to bear on each topic.

To provide some structure for our discussion, I have put together a list of questions and background reading.  My hope is that each participant will have a special interest in one or two of the questions and so spend more time thinking about those items.  That way, we will (hopefully) have several people willing to jump in with thoughts/ideas on each topic.

Feel free to add a comment if you have additional thoughts or references that might be helpful in preparation for Tuesday.

To attend office hours in Second Life, click here.


Questions for discussion

Q1) What economic attributes is fair value meant to capture?

Market prices are influenced by many factors: expected cash flows, the time value of money, risk, market participants’ attitudes toward TVM and risk, investors’ capital flows and liquidity needs, etc.  In theory, which of these factors should fair value capture?  In practice, what factors are (or do preparers/users believe should be) captured in fair value measurements?  Do these answers differ for level 1, 2, 3 measurements?

Suggested background reading:  SFAC 7; SFAS 157

Q2) Is fair value inherently (and excessively) procyclical?

One of the biggest concerns with fair value measurements is that fair value is procyclical and that it has contributed (or even caused) the current credit crisis.  While academic research has identified various potential feedback mechanisms, each mechanism requires some assumptions about the way in which regulators, investors, and/or managers will behave.  Is there empirical evidence that these (or other) fair value feedback loops have caused any major firms to go under?  How would one go about testing these theories (broadly) or specifically their role in the current credit crisis?

Suggested background reading:  For potential mechanisms, see notes on Haresh Sapra’s presentation in office hours on fair value here and see also Bloomfield, Nelson, and Smith (2006);  For arguments against the role of fair value in the credit crisis, see sections B and C (pgs. 8-14) of Hunt (2009).

Q3) In bear markets, how do investors respond to fair values?

One type of feedback mechanism work like this:  As the market value of a firm’s assets decline, fair values cause firms to report losses.  As these firms see a decline in the strength of their balance sheets, they are forced to sell some of the assets with declining values, which sets off a new round of write-offs.  Related to this, there seems to be some concern that, as fair value losses get racked up, liquidity will dry up and prices will spiral downward.  However, others (including some Board members) have argued that fair values facilitate identification of risk and valuation and so facilitate market liquidity.  Is there evidence on this issue?

Suggested background reading:
Newt Gingrich, “Suspend Mark-to-Market Accounting Now!”
Plantin, Sapra, & Shin (2008)
Comments by Tom Linsmeier in Dec 15, 2008 Board meeting minutes (pars. 5-9)

Q4) What happens when risk aversion shifts in the market place?

When risk aversion changes and investors rebalance their portfolios, the ensuing flight to quality can alter market prices, even though the underlying fundamentals haven’t changed.  Is this what we want captured in financial reports?  Do users understand when this happening?

On a related note, the FASB recently issued an FSP requiring firms to separate their losses on non-trading, debt securities into their credit and non-credit portions.  Can preparers separate these two risks?  How do users interpret the separation? (e.g., do they differentially price the credit losses? is the differential stronger for HTM securities?)

Suggested background reading:  See my blog post on the FASB’s recent FSP.

Q5) Is there role for what Larry Smith calls “activity based measurement”, meaning differential accounting based on how management uses (or intends to use) an asset?  What changes when management plans to hold a security to maturity?  In a similar vein, Leslie Seidman has talked about using a three-dimensional matrix to categorize financial instruments when thinking about whether fair value is an appropriate measurement attribute.

Extending a little more along this line of activity based measurement, should accounting standards anticipate how their standards will be used?  And should this inform standard setter’s choices (such as when deciding between fair value and amortized historical cost)?  To what extent should standard setters be neutral or conservative when issuing rules?

Suggested background reading:  Leslie Seidman in the Cautionary Camp on Fair Value

For thoughts on the conceptual framework and anticipatory standards, see this post.