If you took a look at the House Bill that went down in flames today, you might have caught Congress fiddling with GAAP as Rome burned. See Section 132 of the bill proposed to empower the SEC to suspend application of FAS 157, Fair Value Measurements at any time, and for any company. Doesn't Congress know that the SEC already has the power to make and break GAAP?
See also Section 133, which directs the SEC to submit a study to Congress on FAS 157 as it applies to financial institutions, and preferably to make FAS 157 and the FASB a scapegoat for bank failures. If Congress were serious about studying accounting's role in how our economy's "fundamentals" have been heading due south, they shouldn't be looking at FAS 157, because it only addresses "how," and not "when" fair value is to be measured. "When" is, of course, the real question, which means that Congress should be questioning instead the accounting standards that the members' banker friends actually do favor: FAS 115 (marketable securities), FAS 140 (securitizations), and FAS 114 (loans). You know, the ones with the loopholes. Such is politics.
The Blog Must Go On
I feel somewhat abashed to be discussing the niceties of financial reporting in the wake of current events, but if Congress can do it, why can't I? My topic for today is to follow up my previous post on current value accounting, discussing whether the current values to be measured should be replacement costs or fair values. Specifically, I want to share with you my reactions to Federal Reserve Chairman Ben Bernanke's view of "hold-to-maturity" versus "market value." Two other comments on my advocating for replacement cost accounting (and a history lesson from Bob Jensen) also got me thinking a little bit more about what I wrote.
First, Bernanke gave congressional testimony arguing that two prices existed for troubled loans: a "hold-to-maturity" price and a current market price. At first blush, this sounds something like the difference between replacement cost and fair value, but it isn't. This is important to me, because I think that some people (for example, see the comment of George Weintraub on my previous post) may erroneously associate replacement cost measurements with 'value in use,' which I believe would be the FASB's nomenclature for what Bernanke has called "held-to-maturity." There can be no question that value in use invariably presents one with a reliability problem due to the subjectivity of estimating future cash flows and discount rates (i.e., 'mark-to-model'), but that is not necessarily the case with replacement costs.
Moreover, I'm not sure that Bernanke, whose intellectual capacity I respect and admire, is making much sense here (and I suspect he knows that). If we are talking exclusively about financial assets and a global marketplace, how could it be that the market price understates the present value of holding a loan until maturity? If the subject were changed to producing or consumption goods, one could expect a difference between value-in-use and value–in-sale. However in the case of financial assets, such a presumption is perhaps a step too far. Teaming up with Treasury Secretary Paulson to sell such a dubious notion to the public may in the long run affect Bernanke's reputation for independence, just as Greenspan's reputation was damaged by his flip-flop on Bush's tax cuts.
Moving on to replacement costs and financial assets, the 'bid-ask' spread captures the difference between replacement cost and fair value. The 'bid' price is fair value, because it is the amount that someone is willing to pay to purchase the asset from you. The "ask" price (a higher price) is the replacement cost, because it is the amount that a market participant who already has the asset is willing to sell it to people who want it. But, even if there is no formal market maker to provide both bid and ask prices, the concept still applies for the purpose of comparing replacement cost and fair value accounting. For example, and moving from financial to real assets (and I owe this one to Bob Jensen), bluebook prices for vehicles are the functional equivalent of average ask prices by used-car dealers.
Second, Jim Noel, my longtime friend and sounding board, expressed his doubts after reading my previous post that replacement costs were an appropriate basis of measurement for bank balance sheets. That's because the single most important question to ask about a bank, as the financial crisis painfully illustrates, is liquidity. Indeed, the strongest argument for fair value is that it is the gold standard for measuring liquidity.
My response to Jim is to revert to core principles (at least mine) of financial reporting: investors primarily seek information about wealth and changes in wealth. In normal circumstances, liquidity is of secondary importance. As a case in point, maximum liquidity leaves one with just cash and no wealth producing assets.
Granted, there are circumstances when liquidity is paramount, but that should not mean that principles-based accounting needs to be either replacement costs all the time, or fair value all the time. For example, when ARB 43 requires a departure from historic cost to measure inventory, the general basis of measurement is replacement cost. However, replacement cost valuations are capped by 'net realizable value' and floored by 'net realizable value less incremental costs to sell.' A second, and more general example, is the required switch to liquidation values when an entity's ability to continue as a going concern becomes doubtful to a sufficient degree.
The Fair Value Debate Needs Some Rules
My general sense is that some order is needed to make sense of the challenge to fair value. Whatever your position is on the issues, I would like to suggest that you think about providing answers to the following three questions—in the order listed:
- Which assets and liabilities should be measured at current value?
- Which 'attribute' of an asset or liability should be measured—i.e., replacement cost, fair value, historic cost, present value of future cash flows?
- What rules would be needed to assure that the attribute chosen is reliably measured?
Which assets and liabilities should be measured at current value — So much of the balance sheet and earnings volatility that the fair value naysayers are complaining about is brought on by our 'multi-attribute model', where some assets and liabilities are measured in terms of actual past cash outflows and others are measured in terms of estimated future cash inflows. The status quo is simply not stable, so we have to decide whether to move backwards to historic cost or forward to a single set of current values.
Of course, I believe that means we have to finally bring ourselves to measure liabilities as well as assets at their current values. This shouldn't be so hard. A simple economic principle that I pointed out in a previous post is that all of the liabilities of an entity are perforce assets of other entities. If we can measure these financial relationships when they are on the lender's balance sheet at current values, then why in the name of symmetry can't we use that valuation for borrowers? Oy, it makes me crazy.
Which attribute should be measured – Some say fair value, some say present value of future cash flows, some are sticking with historic cost, and I say replacement cost. Whatever your druthers, reported earnings will not pass a reasonableness test until we stop being content with adding apples and oranges on the balance sheet.
How to reliably measure the current value attribute chosen? — I hope George W. (the Skeptical Texas CPA) and others may now see that the difference between replacement cost and fair value is really not a difference between market quotes and mark-to-model. Rather, it's how best to measure what it is you think ought to be measured. Maybe fair value and replacement cost measurements will require mark-to-model every now and then, but for reasons I pointed out in a previous post, replacement cost measurement will require less of it.