FYI | Oct 01 2008
By Greg Peel
In November last year I wrote an article entitled “The Perils of FASB Statement 157” (FYI; 09/11/07). This is how it opened:
“Citigroup analysts in the US last night suggested total write-downs of collateralised debt obligations (CDO) forthcoming from all Wall Street investment banks will probably reach US$64 billion. In their third quarter earnings reports banks marked distressed securities down by a total of US$25bn. But since those reports were issued, further adjustments have been forthcoming.
“Citigroup itself sent the market into a spiral last week by announcing the possibility of a further US$8-11bn of write-downs to come, on top of the US$6.5bn reported for the third quarter. Merrill Lynch had previously shocked with a US$8.4bn write-down. Last night Morgan Stanley wrote down another US$3.7bn, which was actually not quite as bad as the US$6bn some had anticipated.
“From the likes of Bear Stearns, Lehman Bros and others, there’s been deafening silence. If Merrills, Citi and Morgan Stanley are extending the write-downs, where are the rest?”
It is quite astounding to read this now, considering total asset write-downs have since passed the US$500bn mark on their way to what many expect will be US$1 trillion. It is also interesting to note which two investment banks, now dearly departed, were slowest to face reality.
The reason why Citigroup moved to make big write-downs in November last year, and why all others were eventually forced to follow, was because the US Financial Accounting Standards Board decided, in the wake of the collapse of two Bear Stearns hedge funds, that qualification was needed with regard to the valuation of complex derivatives such as CDOs. It thus issued Statement 157, to which every US institution had until year-end 2007 to comply. In retrospect, one might suggest that it was Statement 157 that eventually led to the fall of Bear, Lehman and WaMu, the nationalisation of Fannie, Freddie and AIG, and the necessity today for a comprehensive rescue Plan.
The FASB has long established rules regarding valuation of financial assets for accounting purposes. The simple and most realistic way to value an asset at the end of an accounting period is to “mark to market”. This means an asset on the books is valued at the last traded price of such assets on the last day of trade in the period.
If we’re talking about shares, then that value is readily apparent. Shares prices traded on a stock exchange are perfectly transparent and the clearing house ensures there is only one traded price for any one share – that at which total buyers meet total sellers. Such valuations are known to FASB as “level one”.
But what about property, for example? If an institution owns an office block that it has been in for twenty years, and may be in for another twenty more, how should that property be valued? There has not been a transaction in the accounting period to thus visibly set a price. And no two properties are identical, unlike shares.
To overcome this problem, FASB allows for “level two” valuation, which requires only reference to “observable inputs”. If a similar property three doors down just sold for X, then X can be used as a benchmark for valuation. Obviously one can ask an independent valuer or two as well. Accounting problem solved.
But what if levels one and two still do not apply? What if an asset is just too hard to value against any sort of benchmark? For this problem FASB introduced “level three”, which allowed non-homogeneous assets to be valued using “unobservable inputs”. This basically meant “you decide”. As long as an institution wasn’t trying to use some unjustifiably high valuation, the FASB was happy enough.
Collateralised debt obligations fell into the category of “level three”. Such assets were (a) usually issued once and held to maturity, rarely changing hands again; (b) traded “over the counter”, and not on any exchange or through any central clearing house; (c) individually unique – no two CDOs were identical; (d) extraordinarily complex, such that no one at FASB could ever put up much of an argument anyway. Thus it was incumbent upon holders of CDOs to come up with their own valuations, which usually required a computer model to determine. Thus was born the accounting method of “mark to model”.
Let me ask you this: If you’re paid a performance bonus at the end of an accounting year, and you’re able to use your own determination of the value, at that point, of those assets which you control, are you likely to mark your level three valuation high or low?
Exactly.
When the credit crisis hit, few predicted what would eventually transpire. Many expected it to blow over pretty quickly, which is basically why the Dow hit new highs in early November having previously plumbed panicked depths. But with the introduction of FASB Statement 157, everything changed.
Without going into tedious accounting detail, Statement 157 required “level three” valuations to be made with a much more realistic consideration of risk. Institutions had to be able to justify their assumed prices on this basis. At the time, the US housing price slump was just beginning to accelerate and foreclosures were beginning to rise. This meant CDOs, particularly those with subprime mortgages contained within, began to see the default of first tranches. They became very “risky”.
(CDOs are broken into three “tranches”, with prime mortgages in the largest tranche, mid-prime in a much smaller tranche, and subprime in a tiny tranche. This allowed for a certain level of default in the subprime tranche before the mid-prime tranche was affected, and so on. The creators of CDOs worked out the balances based on average mortgage default rates, but when the housing market slumped foreclosures soared. The mid-prime and even prime tranches thus came under threat of default, as the securities were set up in a cascading fashion.)
To this day we still don’t know how many prime tranches, or even mid-prime tranches, are close to default. We don’t even know how many CDOs actually exist, or even who can legally lay claim to “owning” a mortgage therein. Mortgage aggregators packaged up hundreds of mortgages which were then sold to banks, who then created CDOs, who then sold the CDOs to hedge funds and other investors all over the world. No one can put a price on these things.
It was on that basis that no one was prepared to buy any CDO. If there was no bid, then on a strict “mark to market” assessment those CDOs were worth zero. However, not every tranche of every CDO is going to default, and realistically the bulk may well make it through to maturity unscathed. Therefore they still had some “value”, but how could one determine what that might be?
Thus the phrase “mark to myth” was coined in Wall Street vernacular to replace “mark to model”. From November on, institutions began heavily marking down the value of those assets as Statement 157 required. And the next quarter they marked them down again, and the next quarter again. Wall Street had assumed institutions had been kidding themselves about valuations and simply crossing their fingers that value would return. Hence the “myth”. Extensive write-downs were the proof. But each time house prices fell further in a quarter, down the value of CDOs would go once more. Wall Street kept hoping for “kitchen sink” downgrades that would signal the bottom of valuation, but they were never forthcoming.
The result? Bear Stearns went first, Lehman went two weeks ago, Fannie, Freddie and AIG have been bailed out, Merrill Lynch is no more, Goldman Sachs and Morgan Stanley are just hanging on, IndyMac is gone, Washington Mutual is gone.
Let’s go on. IKB Bank is gone, Northern Rock is nationalised, Bradford & Bingley is nationalised, Fortis is partially nationalised.
And the big one – US Congress is currently negotiating an all-encompassing, last ditch, high cost rescue plan. This is because anyone still left standing is also now at grave risk of going down.
When institutions were forced to write down the value of their toxic debt, it diminished their capital bases, leading to forced capital raisings, slashed dividends, and most importantly, credit rating downgrades. As soon as their credit was downgraded, institutions saw their problems compounded. They needed even more capital to survive. They saw mass selling of their debt. They became illiquid, then they became insolvent, then they collapsed.
All because of one little accounting rule.
Okay. Now you’re going to immediately say “It’s not FASB’s fault! It’s all those greedy Wall Street fat cats and Main Street mortgage brokers and credit-worthless home-buyers fault!”. And I agree entirely. However…
We know that not all CDOs will fully default into oblivion. Thus we know not all CDOs are worth zero. If we ever needed any proof it lies in the fact Hank Paulson and Ben Bernanke have proposed a plan to buy all of the securities – some US$700bn worth – at “hold to maturity” prices. This implies that the US government is prepared to pay the price for these securities that the institutions believed them to be worth before they were forced to begin writing down their value – to “mark” to a non-existent “market”.
There has been some confusion stemming from this “hold to maturity” (HTM) price concept, mostly because President Bush has suggested the securities will be bought at “fire-sale” prices instead. A fire-sale price would be what the institutions are carrying the securities on their books at now, or less. However, the two may not be so different.
An HTM price of any debt security is calculated by discounting both the redemption value at maturity and the coupon stream to that point back to today’s dollars. The discount rate is US Treasury plus an additional credit risk spread. Under this valuation, a AAA-rated security should actually be worth more now than it will be on redemption. CDOs were sold as AAA, with credit rating agency approval, but we now know that to be a farce (or even fraudulent).
However, if the security in question has no rating, meaning it is “junk”, that credit risk spread will be very large and today’s value of the security much lower. A simple rule of thumb is to say if a security has a 50% chance of defaulting then it should only be worth about 50% of its face value at this time. If many CDOs are still at risk of defaulting as house prices continue to fall, then maybe that default chance is 80%, suggesting only 20% of face value. Once again, no one knows.
Paulson and Bernanke have suggested buying all toxic debt at HTM value. What is this value? We don’t know. But they intend to determine this value by holding a “reverse auction”. This means institutions will offer cheaper and cheaper prices to the government to be rid of the debt that is threatening to send them bankrupt. If enough institutions are sufficiently desperate, then a “fire-sale” price will result. Once the lowest price has been set, all other institutions will be forced to either accept this price or not, and if not, write down the value of their securities to this new “mark to market” price anyway.
(Please bear in mind that this is my interpretation only, nothing is particularly clear yet within The Plan, and it is changing every day anyway).
If an institution sells all its toxic securities then it will crystallise its losses. Any hope of holding the securities to maturity and writing their value back up is gone. This is the punishment. The government will now own the upside. Had The Plan been implemented last week financial markets would have stabilised, and credit would have begun to flow once more. Congress is holding off to gain further concessions for the taxpayer, which is well-meaning but misguided. Wall Street would have taken a very, very long time to return to any sort of level of profitability anyway. Yet Main Street would have been saved.
But wait!
Last night talk in the US was that the FASB, in conjunction with the Securities & Exchange Commission, were considering re-qualifying accounting rules, perhaps introducing another Statement which would relax the stricter rules put in place last November. They are considering allowing an interpretation of “fair value” for a security which is more a case of the owner’s opinion, or best estimate, than the “mark to market” implication of Statement 157.
What is “fair value”? Well one assumes it lies somewhere in between “hold to maturity” value and “fire-sale” value. Either way, institutions would be free to assume perhaps a more realistic value, but certainly a more “friendly” value. In other words, a higher value. They could “write up” the value of their toxic securities!
This would immediately increase their capital, stave off a credit downgrade, and save them from potential bankruptcy. Their credit spreads would come down, and they may even begin to start lending to each other.
It is unclear at this point exactly what the SEC and FASB intend, but realistically they may even prove to be the world financial market’s saviours. They may even make The Plan redundant.
Or we may have the extraordinary situation of Congress finally agreeing on a Plan, the government offering to buy the toxic debt, and institutions saying “No thanks, I’ll just write their value back up”.
If you are a Lehman Bros employee you would be looking on with dismay. “Are you telling me we may never have had to bankrupt?!”
I reiterate that this is all just conjecture and vague announcement at this point. But if the SEC and FASB do enact some new, relaxed ruling, then maybe all our problems are solved. That is why financial stocks led the rally on Wall Street last night. In many cases those stocks rallied back above the point at which they fell on Monday. This would be a tremendous boon.
But we also know it would be more than just a boon. It would most likely put us right back where we started from – unrealistic valuations of complex and opaque securities. Incentive to mark too high. Potential for another great fall. Another fantasy world, another house of cards.
We can only hope that somewhere within all this mess elected officials and appointed regulators can sort themselves, and the market, out.
