When Citigroup said in its first-quarter results that it booked a sizeable fair value gain from the dismal performance of its own debt — it proved something of a talking point.
The idea that Citi could book a $30m profit on its own debt* — essentially profiting from a deterioration in the value of its bonds — seemed ludicrous if not downright misleading. But to clarify things, UBS’s excellent accounting expert Dennis Jullens has issued a bit of research.
Here’s what he says:
Fair value option is an accounting choice under IFRS and US GAAP allowing companies to account for financial liabilities at fair value and to recognize the difference between fair value and carrying value of the liabilities as gains or losses in earnings. Companies can choose which financial liabilities to elect fair value option for, and once elected, the decision is irreversible. . . .
We illustrate the effects of fair value option on earnings with an example by using a corporate A-rated bond index for maturities of 3 to 5 years for financial institutions in the Euro area as a proxy for an average corporate bond. We compare this with the Eurozone sovereign debt index for similar maturities to approximate the yield movements of a government bond.
Suppose in May 2007 a financial institution issues a €500 million bond with a 5- year maturity and an annual coupon of 5%. The government bond yields 4.3% on the same day. The company has elected the fair value option for this debt instrument. By the end of 2007, the corporate bond yield increases by 1.2%, while the government bond yield decreases by 0.1%. We estimate the fair value of the bond to have fallen to €475 million. The bank will report a fair value gain of €25 million on its own debt in FY 07 earnings. Corporate yields continue to rise in 2008 while government bond yields, in contrast, fall, causing further widening corporate credit spread. As a result, the company recognizes fair value of gain of €69 million on own debt in FY 08.
Now — the reverse of that is that if the corporate yield decreases and the government bond yield is stable — or even rises — that gain could quickly turn into a loss. In UBS’s example, if the bond yield peaks in the first-quarter of 2009 and then falls to 6 per cent in the second-quarter, and the government bond yield still hovers around 3 per cent — the company would report for the first-half of 2009 a fair value loss against its own debt of €77m — reversing its 2008 gains.
So the flip-side of profiting from your own crummy creditworthiness, is that you lose from your own improving creditworthiness. Sad but true.
Here’s Jullens’ commentary:
The example above clearly demonstrates the absurdity of an entity booking gains when its credit quality worsens and reporting losses when creditworthiness improves. We feel that the financial sector earnings boosted by large fair value gains on own debt may not truthfully represent the actual performance of the sector. Looking forward, if the credit markets recover significantly during 2009, falling corporate bond yields could translate into material losses for banks, and even more than reverse gains from previous periods.
It also raises a valid concern that sizeable fair value gains in one period being wiped out by equally large losses in the next period would introduce substantial and unwarranted volatility in earnings. . . .
Also of note though are the variations in the banks’ own use of fair value for own debt. Jullens has the below table, showing fair value gains on own debt for 10 financial institutions over the past two years. You can see from the table that the absolute size of the gains varies substantially. On a relative basis, Jullens also says while the gains average about 10 to 20 per cent of pretax earnings, there are extremes of over 100 per cent and even 200 per cent.
In other words, if corporate credit improves this year — as it appears to be doing — some banks will be booking bigger fair value own debt losses than others. Ironically however, if the improvement in corporate credit isn’t uniform — if the debt profiles of some banks (ahem, Citi) don’t improve — those institutions could still post fair value gains on their own debt while banks with improving credit profiles suffer. It truly is a mad, mad accounting world.
* The $30m fair value gain on own debt is separate to the $2.5bn positive credit value adjustment booked on its derivatives positions, we think. Though the rules for valuing the two appear to be pretty similar.
Related links:
A Citi-fied Catch-22 – FT Alphaville
Banking credit catch-22 in action? – FT Alphaville
CVAs or ‘the magic of your own credit profits” - Holding to account