Mark To Market, A Beginner's Guide
Feb. 29th, 2008 12:20 pm![[personal profile]](https://www.dreamwidth.org/img/silk/identity/user.png){kind=small}
peterbirksThere's an old saying in the financial world that, as usual, contains more than a kernel of truth. That saying is that "the definition of a long-term investment is a short-term investment that has gone wrong".
Back in the 20th century in the US (and for the first five years of the 21st Century in
mainland Europe), you could book investments at something called "historic value". This led to all sort of distortions, with some companies booking investments that were little more than worthless (non-performing loans being a favourite in the banking sector) at their initial value. This led to "book values" being, more often than not, a pile of toss.
So, in 1993 the US accountancy profession came up with a plan that would solve this, or so they thought. Under this plan, you couldn't value your assets at what you bought them for, but at what you could get for them if you sold them in the market tomorrow.
"But!" cried financial institutions (particularly banks with non-performing loans and companies in Europe with rather dubious cross-booked shareholdings at artificially high values) "therein lies potential disaster!"
Unfortunately, banks are past masters at crying wolf, so the accountancy profession told them to STFU.
And, lo and behold, the banks were proved right, although not for the reasons that they thought.
One thing that the banks spotted back in 1993 was that mark-to-market valuations (also known in a rising market as "what some other fool is willing to pay for something") could create an upwards spiral. Banks look to achieve a constant leverage multiple of their balance sheet (remember that phrase, it's important). If they fail to do this, institutional investors start moaning about inefficiencies of capital. In simple terms, in a rising market, more capital is created, which needs to find somewhere to go, which encourages banks to make loans that they would not make in "normal" circumstances.
You can see where this leads, and the banks could see where it was going to lead. The loans, if the interest is being repaid, create more capital in companies that are expanding, thus leading banks to lend more because the asset of their loan gets better. And, with mark-to-market accounting, the positive feedback is exacerbated.
What even the banks (and the academics) failed to notice in 2003 was that the creation of mark-to-market valuations also helped create a whole new world of valuations - via securitisations.
In simple terms, a residential backed mortgage security is a synthetic (although in the rarefied world of finance, it isn't actually seen as a synthetic) product which makes an illiquid asset (lots of people's houses) a liquid asset (a tradable security).
If you then take a lot of these RBMSs and slice and dice them, putting them back together in a form that categorises the risk more "plainly" you get collateralized mortgage obligations (a 'genuine' synthetic). If you do the same with other types of debt, you get collateralized debt obligations.
Why does this matter? Well, it matters in two ways.
One is that it creates more capital that looks liquid (a tradable security) that isn't really liquid (because its fundamental building block is a house). In the world of mark-to-market "fair value" accounting, this is important (far more so than in historic cost accounting) because, instead of valuing the house, so to speak, you are valuing what someone thinks the debt on the house is worth.
The second is that these securitisations shifted the mortgages off the banks' balance sheets. This meant that capital was freed up, so that they could lend more. Indeed, they had to lend more (see above on constant leverage multiple of balance sheet). Huge wodges of capital were going in search of borrowers, any borrowers.
Once the rising asset markets go into reverse, you get a negative loop, and that's what we are getting now. This morning, AIG wrote off $11bn for Q4 -- that's, er, more than 10% of what Bernanke thought last Summer would be needed to cause a systemic crisis (he doesn't talk about that too much these days). Swiss Re said that another CHF240m had disappeared since January 1 -- a sum that these days barely causes a tremor.
With all of this money "vanishing", it's all very well for AIG and Swiss Re to say that it's not real money because in the end things will work out fine and that they don't have to sell the stuff at the moment anyway (the old "historic cost" accounting argument). Things might well work out fine. But, for the moment, it's not just dubious loans that won't be made anymore -- it's loans that would in normal times be seen as "reasonable". Fair value accounting is crucifying capital, which means (and, once again, we come back to
the mantra of "constant leverage multiple of balance sheet") that only the safest of the safe will be able to borrow cash.
This is basically seriously bad news for anyone looking to refinance a loan, be it a company or an individual. You could be a perfectly good customer in "normal" times, but these are not normal times.
For not a few companies, I can't see this meaning anything else but rights issues, lots of rights issues.
_______
Back in the 20th century in the US (and for the first five years of the 21st Century in
mainland Europe), you could book investments at something called "historic value". This led to all sort of distortions, with some companies booking investments that were little more than worthless (non-performing loans being a favourite in the banking sector) at their initial value. This led to "book values" being, more often than not, a pile of toss.
So, in 1993 the US accountancy profession came up with a plan that would solve this, or so they thought. Under this plan, you couldn't value your assets at what you bought them for, but at what you could get for them if you sold them in the market tomorrow.
"But!" cried financial institutions (particularly banks with non-performing loans and companies in Europe with rather dubious cross-booked shareholdings at artificially high values) "therein lies potential disaster!"
Unfortunately, banks are past masters at crying wolf, so the accountancy profession told them to STFU.
And, lo and behold, the banks were proved right, although not for the reasons that they thought.
One thing that the banks spotted back in 1993 was that mark-to-market valuations (also known in a rising market as "what some other fool is willing to pay for something") could create an upwards spiral. Banks look to achieve a constant leverage multiple of their balance sheet (remember that phrase, it's important). If they fail to do this, institutional investors start moaning about inefficiencies of capital. In simple terms, in a rising market, more capital is created, which needs to find somewhere to go, which encourages banks to make loans that they would not make in "normal" circumstances.
You can see where this leads, and the banks could see where it was going to lead. The loans, if the interest is being repaid, create more capital in companies that are expanding, thus leading banks to lend more because the asset of their loan gets better. And, with mark-to-market accounting, the positive feedback is exacerbated.
What even the banks (and the academics) failed to notice in 2003 was that the creation of mark-to-market valuations also helped create a whole new world of valuations - via securitisations.
In simple terms, a residential backed mortgage security is a synthetic (although in the rarefied world of finance, it isn't actually seen as a synthetic) product which makes an illiquid asset (lots of people's houses) a liquid asset (a tradable security).
If you then take a lot of these RBMSs and slice and dice them, putting them back together in a form that categorises the risk more "plainly" you get collateralized mortgage obligations (a 'genuine' synthetic). If you do the same with other types of debt, you get collateralized debt obligations.
Why does this matter? Well, it matters in two ways.
One is that it creates more capital that looks liquid (a tradable security) that isn't really liquid (because its fundamental building block is a house). In the world of mark-to-market "fair value" accounting, this is important (far more so than in historic cost accounting) because, instead of valuing the house, so to speak, you are valuing what someone thinks the debt on the house is worth.
The second is that these securitisations shifted the mortgages off the banks' balance sheets. This meant that capital was freed up, so that they could lend more. Indeed, they had to lend more (see above on constant leverage multiple of balance sheet). Huge wodges of capital were going in search of borrowers, any borrowers.
Once the rising asset markets go into reverse, you get a negative loop, and that's what we are getting now. This morning, AIG wrote off $11bn for Q4 -- that's, er, more than 10% of what Bernanke thought last Summer would be needed to cause a systemic crisis (he doesn't talk about that too much these days). Swiss Re said that another CHF240m had disappeared since January 1 -- a sum that these days barely causes a tremor.
With all of this money "vanishing", it's all very well for AIG and Swiss Re to say that it's not real money because in the end things will work out fine and that they don't have to sell the stuff at the moment anyway (the old "historic cost" accounting argument). Things might well work out fine. But, for the moment, it's not just dubious loans that won't be made anymore -- it's loans that would in normal times be seen as "reasonable". Fair value accounting is crucifying capital, which means (and, once again, we come back to
the mantra of "constant leverage multiple of balance sheet") that only the safest of the safe will be able to borrow cash.
This is basically seriously bad news for anyone looking to refinance a loan, be it a company or an individual. You could be a perfectly good customer in "normal" times, but these are not normal times.
For not a few companies, I can't see this meaning anything else but rights issues, lots of rights issues.
_______
![[identity profile]](https://www.dreamwidth.org/img/silk/identity/openid.png){kind=small}
Good stuff
Date: 2008-02-29 01:48 pm (UTC)Of course you could always engineer a revaluation of assets by bed and breakfasting an asset through any nearby vehicle but people never did it. That does lead to balance sheets which are not reflective of the asset base of a company. For instance, my firm (like most practices) has a value which is almost totally Goodwill. But in normal circs your Goodwill doesn't appear. I had to buy out a partner and so the money I paid her to sod off is now in the balance sheet which makes an even more confusiing balance sheet.
How is the market in these CDOs and similar widgets? I can't help but think that these write-offs are overdoing things and that there'd be Soros-sized amounts to be made by someone who could actually un-slice and dice these things, rip out the safe lending and treat the proper riskier lending as what it is. Right now it seems to be the equivalent of a bit of plastic stuck in the cardboard recycling.
Re: Good stuff
Date: 2008-02-29 02:10 pm (UTC)In a sense, you may be right. If you could buy up the whole lot, unslice and dice them, put them back together, and have a good look, you could discover how toxic things really are. In essence, the "real" loss is how much has been lent that won't get paid back. The "fair value" loss is how much people are willing to pay for something when they don't know what the "real" loss is. At the moment, the sum that they are preopared to pay approximates to zero.
Buffett could make a fortune if he was prepared to gamble, but he isn't. He just spotted that the municipal bond parts of the bond insurers were good value and tried to get hold of them without touching the mortgage stuff. Ammbac told him to sod off and (in Buffett's own words) the others didn't even bother to reply.
Where we are at the moment is maybe half-way along the "toxic" curve. I'd guess that we'll reach the nadir in mid-2008 (August?), by which time all the shit is on the balance sheets. Then, oh-so-slowly, the write-offs start coming back in positive terms onto the balance sheet as liquidity returns.
Insurance is used to this. You put stuff into reserves for future claims and then, when things aren't as bad as you feared, you release them back into profit and loss. Other financial groups aren't so used to it. It's a bit misleading to say that Soros amounts are there to be made because, as I said, the holders of these products, which they have had to write down in value, do not intend to sell those products. AIG isn't going bust. Under historic accounting, it wouldn't even be in a crisis, because it would have been able to keep the products at historic cost (the accountants/auditors would not have kicked up such a fuss). Buffett can't make money from it, because AIG isn't selling. However, Martin Sullivan might well fall on his sword on Sunday.
What the monoliners plan to do is split themselves into municipal bond lenders and mortgage bond lenders (a semi-unslice-and-undice). You will then get new companies being formed that have an equivalent today in Pearl (or the old Resolution) and the running of closed life assurance books. These will put together all the toxic crap at a relatively low price, run them efficiently, and with a bit of luck make a profit by the time everything is shut down.
PJ
Fine Young Cannibals
Date: 2008-02-29 05:28 pm (UTC)I enjoyed your analysis of CMOs/CDOs and the flip-side of leveraged lending. Well, actually, I was somewhat in awe of it. Christ, if you're prepared to devote that much effort to explaining the current market to proles like us, what the hell does your day job's output look like?
I'd be interested to hear your views on the concept (spouted by some Yank via the Telegraph today -- I forget who) that "replacing" bank loans with SWF capital won't work because "it's a different model. Banks work off leverage (ie equity). SWFs work off debt."
Well, that's as may be, but presumably Wall Street is hoping to turn long-term debt into equity. And presumably the SWFs involved believe that they can do that. Otherwise, what would be the point? Or am I missing something?
Damn, I really should have gone to a major public school. Then I could have moved into finance, and the City would be my Lobster.
Re: Fine Young Cannibals
Date: 2008-03-02 08:39 am (UTC)You can restructure virtually any kind of equity into debt and debt into equity, provided the counterparties exist. Much of the problem in the accountancy land of banks and rating agenncies is often deciding whether something should be treated as debt or as equity (the banks want the lenders/investors to think of it as debt, but the rating agencies to think of it as equity). If you often can't really tell whether something is equity or debt, how can it be a different business model?
PJ
Re: Fine Young Cannibals
Date: 2008-03-03 08:01 pm (UTC)I'm probably wrong, but it seems to me that the concept of derivatives has degenerated into some weird kind of anti-mathematical panic, driven by lust, greed, and ignorance. I'm getting sick and tired of hearing that these instruments require "a degree in astrophysics" to understand.
Well, no they don't. That, in fact, would be the last sort of degree that one might find useful. Understanding derivatives involves nothing more than transparency and a whopping big computer program to track the slices; probably, but not necessarily, using Markov and Monte Carlo. (Even then, unfortunately, the Black Swan comes in; but this isn't a Black Swan event. This is pure lust, greed, ignorance, and perverse incentives. I mean, $160 million to get out of the way. Hell.)
Your point on restructuring is well taken. There is a continuum here. I'm just guessing, but I think that the different business models are based on the discontinuities involved, which in the current case seem to involve an almost total lack of liquidity. Leverage (I assume, equity) seems to work well in a liquid market. Debt seems to work well in an illiquid market, assuming that there are assets to back up the debt.
So, as far as I can see, the time series across the debt/leverage continuum depends upon a complex relationship between liquidity and assets. Liquidity is currently tending towards zero. Assets (owing to lack of transparency) are anybody's guess.
There's probably a PhD in this for some fool out there.
However, on your main thrust (and against the Yank's main and obviously protectionist thrust), there is obviously no reason in an (even limited) free market for capital to deny SWFs the ability to buy in to western banks. Indeed, if there were, this would be a perfect example of moral risk. (Even though I wish to be exempted from the qualifier on religious grounds.) You fuck up, you lose.
It is, if I may repeat, a way of turning (what would otherwise have been) long-term debt into equity. That's the continuum for you. The fact that it's practically instantaneous for SWFs at the moment is just a comment on how seriously the banks have fucked up.