"We never sell at a loss" policy of accounting Friday, March 14, 2008
This influence of this blog is astounding. Having only written about mark to market accounting yesterday with this piece, Steven Bensinger, AIG’s chief financial officer, has stepped into the limelight with AIG being the first company publicly to air a proposal to move away from “fair value” accounting [mark to market].
Joking aside, under AIG’s proposal, which has been presented to regulators and policymakers, companies and their auditors would estimate the maximum losses they were likely to incur over time and only recognise these in their profits. All other unrealised losses would be recorded on the balance sheet but would not affect profits. In AIG’s case, this method would have reduced the impact of the $11bn writedown on fourth-quarter results to $900m.
Besinger is reported to have said AIG’s proposed change could bolster financial institutions’ confidence and help thaw the current freeze in credit markets.
So the plan is to write down the assets on the balance sheet but tuck the away the adjustment in a balance sheet "revaluation reserve", out of the spotlight. Suddenly the apparent "health" of companies would be transformed and confidence restored to markets.
As I described here the mark to market policy forces companies to recognise losses even when they have no intention of selling assets at the current prices. However, decisions can change just as easily as prices, hence at the balance sheet date there may be no intention to sell but, a month later, a new decision may be reached. Hence the scope for manipulation increases.
Imagine the following scenario:
- Management buy a bond trading at discount of 20% to par. They declare at the outset that the bond is to be held to maturity in 10 years time.
- The bond initially falls in value to a discount of 25%. At the balance sheet date Management re-affirm that [having fallen in value] they have no intention to sell and no P&L adjustment is made
- The bond then sharply rises in value to par, despite having 10 years to mature. Management then decide to sell the bond to realise the cash early and recognise the triumph of their investment skills in the P&L.
Certainly, many private investors take the view that you don't lose money until you sell because your cash balance isn't affected until then and thereby end up kidding themselves that they are no worse off when prices fall. Yet is staggering for the CFO of a major financial institution to believe that sophisticated investors will not re-introduce such accounting adjustments when evaluating the company [I concede that some accounting stunts do appear to fool investors]. However, it wouldn't be the first time that policy makers adopt an idea allowing an artificial and brighter view of reality to be presented.
I think there is a simpler test to apply. Looked at from the viewpoint of AIG insurance [general or life] customers, would they prefer to know that the assets that were covering their policies and which might have to be sold to pay out claims, had dramatically fallen in value and as such may impair AIG's ability to payout?