Sunday, February 1, 2009

Whatcha Gonna Do When 1+1 No Longer Equals 2?

Several widely respected experts have recently said the same thing: all US/European banks are insolvent if they mark everything to market.

How could this be, after so much write-downs and bailouts? I have no evidence to support or refute them. So my only logical choice is to join them.

My guess is, forget about CDS/CDOs. They're past problems, known problems. The hidden toxic dump remaining, the next bomb that keeps blowing up, may be the highly customized, highly complex structured deals they've been accumulating over the years. There's no wholesale market for any of them. Decomposing them carries substantial risk of mismatching due to the various disparities in the market today. Hedging? If Merrill got into a $15B trap doing the simplest CDS/bond basis trade, how can you have any confidence of any hedge/arbitrage/trade working as expected?

Here lies the biggest surprise to the financial world so far throughout this crisis. 1+1 no longer equals to 2.

If the industry is still struggling to explain the CDS/bond basis and determine whether and how to trade it, then good luck with the structured deals. I've seen some of them. It could easily take a highly specialized expert days to digest it, break it down to pieces, figure out how it'd behave under different scenarios, and calculate risk based on existing standard models. Except, of course, the assumptions made by many standard models have been proven way off-base by the market over the past year. Now, on top of this, take away 1+1=2.

Portfolio decomposition is THE foundation for synthetics and much of structured finance. If you take this away, you take away a big part of the foundation of financial pricing. But the world should not be surprised. It happened before for Long Term Capital. Calling the market stupid is as productive as calling reality stupid, even though you could very well be correct. The market is just pricing in some factors omitted by standard models. I have a model that can explain and quantify these factors but it's beyond this article and beside my point.

My point is,
1. it's futile to expect anybody to price/hedge lots of the structured trades meaningfully, even with the best/purest intentions, and
2. even if there is a liquid market, the pricing mechanism is so different now that many tried-and-true, fundamental assumptions in finance are no longer valid.

It's a wild new world. It may be rational still, we have to assume it so. But it's so fundamentally different that it'd take some time (at least months, quite possibly years) for the industry to make sense of it. If you think I'm exaggerating, think about the impact of abandoning Libor and the US treasury curve having a credit spread of 50 bps embedded.

What we're going through is wholesale, across-the-board, fundamental repricing of every financial instrument in existence.

So why are we still debating about which banks are good and which are bad, what their valuation should be, whether to take away bad asset and how to value them, etc etc?

Forget about valuation and risk management! It's not possible! It's a new world that we don't understand!

There, feels better already. Now we can calm down and think rationally.

Now that we admit we don't know how to price them and cannot possibly know for a long time, the solution becomes apparent: don't price them.

1. Ask banks to do a one-time categorization of assets they deem "hard to price". This hard-to-price pool cannot change in the future.
2. Sweep these hard-to-price assets aside. Get rid of all hedging and stop all trading of this pool. It will be held to maturity except, for perpetuals (e.g., real estate), the bank can decide when to sell (but never buy back into the pool) subject to some hard deadline (e.g. 30 years).
3. Provide a total cost, not including operational/financing costs/hedging costs so far since initial trade, future collateral/margin costs, and any realized P&L due to position changes so far since initial trade.
4. The total cost becomes a nominal addition to the bank's Tier-2 capital base for regulatory and accounting purposes.
5. As assets in the frozen pool mature, the realized P&L with respect to the reported cost is accounted for in Tier-2 capital.
6. The frozen asset can be used as collateral, at reported cost, at the Fed window prior to maturity.
7. Everything else will be marked to market.

The final settlement at maturity is the only sure answer to the perennial question of "what's its worth".

Currently banks do have some flexibility in which assets to mark to cost. But there're too many restrictions in some regards while too much flexibility in others. By doing it across the board (regardless of whether the asset is owned by the mortgage division, a trading desk, or the financing department) and at the same time, one time only, we eliminate the regulatory arbitrage and uncertainty.

Under this system, banks will have to prove some illiquidity threshold for assets they put in the pool. Such threshold will be determined by expert panels set up by the government. Other than the illiquidity criterion, banks are free to choose which ones to keep frozen. If they choose assets already marked down, they'd get a one-time boost, which they must disclose in the quarterly report.

The merit of this approach is to provide capital relief to the banks without government subsidy, government guarantee, or some other artificial price intervention. Banks would not be forced to sell assets and/or raise capital in the worst moment. It's the ultimate bailout without spending a penny.

The hope is that, when held to maturity, the macro-economy will recover and most assets will pay off. Nobody is seriously predicting Armageddon after all. In fact, China used essentially the same approach circa 1998 and it worked out beautifully.

This is not the best solution, of course. The best solution is to let all banks fail, use a small fraction of the trillions of bailout money to support deposits and the massive ensuing unemployment, let the good and capable to restart from scratch. We'd have a lean and healthy private partnership Wall Street in no time, in which risk and reward are matched in magnitude and duration, owners actually have control, and stupidity/mediocrity has nowhere to hide.

But that's no going to happen, is it?


Foobarista said...
This comment has been removed by the author.
Agile said...

Peng Bo,
I found your blog through seekingalpha. Nice blog. This post regarding the solution to banks bad asset is quite innovative :-)

Good job !!!


Bo Peng said...

Thanks, Agile. Most of the stuff I write here are published on SeekingAlpha. said...

simply stunning! At the summit site, a modern and beautiful
well done