- The Obama administration appears to be in the final stages of formulating a plan for a "bad bank" to purchase troubled assets from private sector banks.
- Determining a fair price for these investments is likely to be a point of controversy in the plan.
- The disparity between market prices for these assets and their long-term value evidences the fundamental flaw in applying market-to-market accounting rules to an extremely illiquid market.
Today, US bank shares rallied 20%* on reports the Obama administration is making plans to establish a "bad bank" to buy so-called toxic assets from financial institutions—a financial version of Yucca Mountain, if you will. The hope is once banks' balance sheets are returned to some semblance of normalcy, confidence in banks will be restored, lending will increase, consumers will spend, companies will invest, and the economy will recover.
The only bugaboo in the plan is what price to pay for the assets? Reportedly, the government is considering some sort of discounted cash-flow model rather than relying on market prices since prices in this extremely illiquid market have become significantly disjointed from the held-to-maturity value of the assets. Put another way, many of these assets aren't nearly as "toxic" as the media continually reports. Remember, most collateralized debt obligations (CDOs) and similar vehicles were never designed to be traded frequently, so there's virtually no secondary market for these products. Unfortunately, FAS 157, the mark-to-market accounting rule put in place in November 2007, mandated these instruments be valued as if they are going to be liquidated right away in a non-existent market with few if any buyers.
Consider a collateralized mortgage obligation (CMO) comprised of a geographically diverse mix of subprime mortgages. The default rate on subprime mortgages is high—around 20%. But when a borrower defaults on a mortgage, it's not a total loss. The house is foreclosed upon and sold. The selling price might be significantly lower than the loan amount, but some money is recouped. And the structure of these products means the equity and lower-rated tranches suffer the first losses. Typically, banks own the higher-rated tranches, so they're further insulated from defaults. So even though these assets have been impaired by higher mortgage default rates, most are worth far more than the prices they'd fetch in an immediate open-market sale.
This all leaves us scratching our heads. Isn't this mark-to-model method proposed for the bad bank eerily similar to the models financial institutions used to value their illiquid assets before FAS 157 came to pass? Implicit in the bad bank idea is the acknowledgement applying mark-to-market accounting to such large, illiquid investment vehicles was a terrible idea from the get-go because the market (or lack thereof) drastically undervalues these assets. The result of this unfortunate accounting rule has been a tsunami of write-downs, bankruptcies, shotgun mergers, and the ensuing market volatility.
Herein lays the pricing issue for the government's bad bank. Do they pay something close to the long-term value of these assets to provide private sector banks the maximum amount of relief? Or do they pay something closer to the market value of the assets to give the government-owned bad bank the maximum amount of potential profitability? The former would undoubtedly be seen as an underhanded bailout of an industry that's already received a doozy of a relief package. The latter would make banks' losses permanent, possibly cause more banks to fail, and effectively transfer wealth from private firms to the government thanks to an accounting rule.
A bad bank could work as long at it helps restore confidence in financial institutions and gets credit flowing again. But it's ironic the bad bank plans to do exactly what most banks intended to do with these investments in the first place—hold them for the long term.
*S&P 500 Banks Index