Can you explain how Geitner's latest is supposed to work?

March 27, 2009 9:24 AM

Geitner's latest: If the problem is the spread between fantasy values that allow a bank to say it's solvent and market values that say it's insolvent, how does a plan that protects private investors from most but not all of the downside get them to overpay for these assets, narrowing the spread?
posted by markcmyers to Law & Government (14 answers total)

There's a number of different definitions of "solvent" here that make a big difference. It's easy if you think about it in extreme terms:
- say I have some book of loans that are definitely not performing, but they're not _all_ in default. I still make some money from them.
- at the same time, my ability to sell these loans is pretty weak: no one has any money, and the minimal yield they offer isn't that exciting because there's more money to be made elsewhere, and its too expensive/impossible to borrow money against them to juice the returns.

Now, provided my already crap book of loans doesn't get worse and I am making some spread over my funding costs, even if I may look on paper insolvent under the definition that what I own is worth less than what I owe, clearly, if I can operate for a while, eventually I will invert that situation. I'll just sit on the cash from those loans and eventually become solvent, by any reasonable definition, while transiting through other states of insolvency (in this case, namely insolvent with regard to statutory net capital requirements and insolvent in that my liabilities outstrip my tangible net assets. But on an operating cash basis I may remain solvent)

The tricksy bit is that I can't actually access enough funding in this environment to continue to operate. If I could, at a reasonable price, I *would* survive, but everyone is too afraid of my solvency, or whats on my books, or the complete opacity of the FDIC seizure/bailout process, no one wants to lend to me. If they new I had a normal, not that good book of business that was reasonable, I can access wholesale funding and eventually earn my way back to solvency by all the dimensions.

So, what I'm getting at, is that for some banks, the problem is _not_ the spread between the fantasy values and the market values, its the fact that fundamentally, the financial system is lubricated by trust, in the form of credit, and without the ability to trust in what's going to happen, wholesale funding to banks and bank-like entities disappears. No one can really price the default risk anymore, so the banks can access funding and that actually will make them die.

Another way to think about this is that...'overpay' is a little tough to gauge here. Say that we look at some book of loans and we are very, very confident that it will yield at least X%. It could be way higher, but we are conservative dudes and this is a buyer's market. Say X is small in this case. 5%? 4%? Way better than treasury yields right now, but maybe not worth tying our money for these returns. We, as regular schmoes, can't really access cheaper funding than the rate on these, but the fed can. So the fed's basically saying, "hey, what if we essentially let you borrow at our effective rates? Doesn't that make this deal look a little more promising?" So whilst everyone else is deleveraging the fed is leveraging to mitigate the pain.

Now of course, some of the banks are fucked anyway, but do you see where I'm going with this? How you could by one or another definition look insolvent but still be a viable going concern, particularly if smooth functioning of the credit system was restored?

Apologies if this doesn't make any sense I literally haven't slept in like forty zillion hours. And by "literally" I mean "figuratively".
posted by jeb at 9:38 AM on March 27, 2009

Because you aren't overpaying for the assets if they are undervalued. The bid-asks on the assets are tremendous because of fear not because one side is delusional and the other reasonable.

The non-recourse leverage is designed to juice returns for investors to put the preceived risk/reward back in line. Then the theory is that by essentially creating an auction to bid on the assets you will find out what the value is that will give investors a reasonable return on their equity, while allowing the banks to find a real liquid market value to mark their assets to. In theory since everyone is using the same amount of non-recourse leverage you should compete the returns down to a pretty low level. The lower rate of return the investors will require the higher the assets will get sold at, the better shape the banks will be in, and they will start lending again.

If you believe (and I would genuinely like someone to show me their math if this is their belief) that the current bid for the assets is correct then the Geitner plan is doomed to fail.

The alternate explanation is that the government needs to find a way to clean up the banks balance sheets by buying bad assets but doesn't beleive the current market is correct. In a perfect world the Treasury would hire a bunch of guys to try and figure out what the assets are worth and buy them from the bank. If a bank ended up insolvent after that exercise sell it or nationalize it.

The reality is that if the Treasury unilaterally came out and said "we'll pay X for Y" it would be a politcal kerfuffle of amazing proportion. Either they'd have to keep the valuation methodolgies totally opaque or they'd end up in interminable arguements about what sort of assumptions for default rates and recoveries to put into the cashflow models.

Basically think of the Geitner plan as paying a fee to someone to figure out what the true value of the assets are in a way that keeps the government out of the price setting role.
posted by JPD at 9:38 AM on March 27, 2009

Jeb isn't that more the argument for "suspend mark to market and statutory capital standards and let banks earn their way out of the hole", whereas the Geithner plan is more "people are freaked out because they assume the banks are insolvent based on current unreasonable market prices for certain scary assets - so what we need to do is fix the market for scary assets, then people will realize that banks are not in as dire a shape as thought, so the credit markets will open back up"
posted by JPD at 9:45 AM on March 27, 2009

The real question here though is when precisely an asset becomes worthless, since after all the line between a "rational" and a "fantasy" market is ultimately subjective: it's possible the collection of lint you have in your basement will someday be worth millions. Everyone seems to agree that some of the assets in question are near-worthless if not entirely worthless, although naturally no one knows for sure: again, the nature of any market is such that even seeming worthwhile investments will suddenly drop in value, or vice versa. The Geithner plan suggests that these assets will yield eventually for patient investors, and that the plan will help to both better locate the "market price" for what these assets are indeed worth, and also to help re-build confidence and unfreeze credit lines. I wish I shared that confidence.
posted by ornate insect at 9:59 AM on March 27, 2009

There has been some detailed analysis of the cost vs. returns performed by some bloggers. The short of it is that the outside investors stand to make a bunch of money, while the government will make less or be stuck with the losses.

The “Geithner Put”, part 1 (many links from here, including Modeling an FDIC Robbery).

The more cynical view is that the banks themselves will game the system, paying an excess for the assets, in order to pass the losses onto the FDIC, and thus potentially sounder banks who pay insurance or the taxpayer, and that the plan is very complex to help hide that fact. Personally, I share this view. To paraphase some examples I have seen around, which I probably will mess up:

Say that a bank bought $100M worth of toxic stuff. They have marked it down to $80M already, but the current market value is $30M. If they sold it at that value, they would be insolvent, taken over by the government, and then they would lose their bonuses and any value in their stock options. So instead, they lend some of the dreck to the Fed under one of the many programs available, and get some cash - lets say $5M. Now, they give this $5M to a friendly hedge fund which can participate in the auctions.

The hedge fund then bids on the assets at face value: $100M. Since the government is providing non-recourse funding for up to 97%, the hedge fund only has to kick in $3M, leaving them with $2M. The bank gets $100M in cash, and pays back the $5M to the Fed, leaving it with $95M in cash. Yay! Now, not only are they not insolvent, but they look like they were geniuses all along because they sold it at face value. The hedge fund has $2M in pure profit from the bank, and $3M in crap, which they don't care if they lose, since it wasn't their money in the first place. If it makes money, great, but if it doesn't all they lose is $3M that wasn't theirs in the first place. So who loses? Well, the government has pitched in $97M to help get this off the banks books. The banks were overvaluing it at $80M, and the market thinks it is worth $30M, so the real value is in there somewhere - lets say $50M, which means they eventually lose $47M.

In the end, the bank had paid $5M to avoid a loss of $50M. The hedge fund made $2M of that. The taxpayers eat $47M.

Good work if you can get it.
posted by procrastination at 10:46 AM on March 27, 2009 [1 favorite]

Ok. Why do you think the assets are worthless? I'll give you a really rough example of why I think the assets are worth more then the current market prices.

Lets take Alt-A RMBS 2006 Vintage. (I'm not a structured credit guy by any means so lots of caveats apply)

If you look at the current default rates, do the math on how default rates and unemployment interact, assume unemployment rates in excess of 12% - you get a base case default rate. Then its a question of recoveries. If you look at the loans themselves in the assets you can see what the LTV of the home was so you know how much of the downpayment you have to eat through. So for example lets look at something thats like an '06 vintage first lien Alt-A RMBS. Figure you have an average LTV of 90%, written at the peak of the market. Home prices decline 40% (in-line with the very worst sub-markets in CA, NV, AZ, FL). So my collateral goes from 1 to .6 - but I don't share the loss until my ten bucks is eaten through, the homeowner does. So worst case on a defaulted loan my recovery is 64%. Ok so what are default rates. As of October ( I know lame, but I was being quick with google) 20% of alt-a loans from 2006 were "seriously delinquent (this actually overstates the real number quite a bit because it disregards prepays from refinanced, and a large % of serious delinquents ended up getting back current) out of the pool. So lets say 17.5% are delinquent? Ok so .82.5*100+.17.5*.64 = 93.7%. And the current default rates are closer 4% and after this many years most of the people who were doomed to default from day 1 have defaulted. Most of the future defaults will come from increased unemployment - which like I said is a knowable thing kind of.

I have to admit though that I don't know enough to say comparing this to the current '06 vintage RMBS AAA Index on markit.com that currently trades at 32 is a fair comparison. But based on my shallow understanding, its pretty reasonable- although the credit quality in the index is weaker, I'm looking at the entire pool for my default rates and the index is supposed to be overcollaterlized to get the AAA rating. Also the index itself has little liquidity so it may not be the best indicator we have, but short of calling an MBS desk that's probably the best we can do.

So 93.7 vs 32. does that seem like a functioning market to you.

You do the same math for things like leveraged loans as well. There is a reason why experienced distressed debt guys think this is going to be the greatest period in the history of the asset class. They are just scared to catch falling knives because they don't have permanent capital.

I'm not saying there aren't worthless assets out there that were colossal f-ups (HELOCS which people thought would act like home loans but ended up acting like credit cards are an asset class that is essentially worthless. Junior leveraged loans in some LBO's of questionable businesses, etc) but the market is clearly broken.
posted by JPD at 10:59 AM on March 27, 2009

Why do you think the assets are worthless?

Not sure if this question is directed my way, but just to clarify: I did not say they were worthless, only that value and worth are entirely relative to the market. Eventually, there may be a market for just about anything one can name. Eventually. Presumably these assets and complex mortgage-backed securities and asset classes were cherry picked before they wound up at the Fed's fire-sale, and presumably if there was a lot of value here then they would have been squeezed dry by investors and portfolio managers prior to the credit crisis. This is the bottom of the barrel, is it not? They were taken off bank books as a stop-gap measure to free up credit, if I understand this correctly, because they were dragging down the market. Of course, it is indeed possible that some of them will regain and exceed current values, but does it not seem unlikely that all will do so? I mean if they all exceed their current values, then maybe on a macro level we've just re-inflated a bubble that will pop once again, and maybe the game is indeed being rigged such that the Fed is acting merely as a temporary warehouse for assets that, in the long run, have no business being re-introduced to the market. I'm not claiming omniscience here; far from it. Just wondering what this stuff amounts to.
posted by ornate insect at 11:11 AM on March 27, 2009

value and worth are entirely relative to the market That makes absolutely no sense. Value is value. If something is worth ten but someone tells me its worth five, it isn't worth five.

Presumably these assets and complex mortgage-backed securities and asset classes were cherry picked before they wound up at the Fed's fire-sale, and presumably if there was a lot of value here then they would have been squeezed dry by investors and portfolio managers prior to the credit crisis. This is the bottom of the barrel, is it not?

These assets were not taken off bank-books. That what this entire exercise is about. I have no idea what "Squeezed dry by investors and portfolio managers" means. You can't squeeze these assets. There is no excess value to be extracted from them.

Ignoring the time value of money and its relationship to interest rates the most a bond can ever be worth is par. It isn't an equity. You get paid what is says on the label. No more, maybe less. No one denies these assets are worth less then par (original cost). Its a question of how much less. The Geitner plan is a bet that these assets are worth more then the market thinks. I happen to think that's the case. I think I have good reason to think that as I showed in my example above. If you disagree with my assertion please explain why. I am you might be surprised to hear, very interested to hear from people who disagree and their rationale for it.

I mean if they all exceed their current values, then maybe on a macro level we've just re-inflated a bubble that will pop once again, and maybe the game is indeed being rigged such that the Fed is acting merely as a temporary warehouse for assets that, in the long run, have no business being re-introduced to the market.

I have no idea what you mean by this? I am saying the deeds of 03-07 were a disaster, just not to the degree that the market currently thinks it is. This plan isn't reigniting any bubble. Look at savings rates (up) , look at home equity withdrawl (down) those trends aren't going to change. Americans are delevering, the consumer is realizing his way of life was unsustainable.

If you had asked someone five years ago if they though it was possible that a AAA RMBS could have value of only 94 you would have been laughed out of the room. The idea that an entire pool could go underwater was unthinkable.
posted by JPD at 11:40 AM on March 27, 2009

Response by poster: Procrastination, I apologize if I oversimplify your comment, but you seem to be saying that the bank and the hedge fund will collude in looting the government, under a possibly implicit agreement. If so, it's brilliant (but not at all nice) on the part of Geithner. Government money fills the hole in the bank's balance sheet under cover of a market mechanism. The only problem: How long will it be before the FDIC has to be bailed out? Soon, I suspect, we'll be hearing that the FDIC is too big to fail, then it'll be Congress to the rescue once again, a gun held to its head.
posted by markcmyers at 1:05 PM on March 27, 2009

I agree that the loophole exists and pray its get dealt with unfortunately I'm not so sure it will- but Geithner is responsible for the FDIC so he has no interest in screwing it.

No matter how the FDIC fails it'll get bailed out. The government has done everything it can to make this clear.
posted by JPD at 1:54 PM on March 27, 2009

If the problem is the spread between fantasy values that allow a bank to say it's solvent and market values that say it's insolvent, how does a plan that protects private investors from most but not all of the downside get them to overpay for these assets, narrowing the spread?


By reducing the downside of a risky bet, you increase the expected payoff and therefore the investors are willing to pay a higher price.

For example: Assume that you are buying a bunch of risky assets and each has a 50:50 chance of earning either $100 or losing $20. The expected return on these assets from simple probability is

(0.5 * 100) + (0.5 * (-20)) = 40.

So you would expect to make $40 each on a pile of these assets. Investors could bid up the price they pay for these assets to somewhere just below $40 each and still expect to make a profit.

Now assume the government guarantees that you can lose no more than $10 each on your assets. The expected return is now

(0.5 * 100) + (0.5 * (-10)) = 45

So you would expect to make $45 each on the pile of assets. Investors could now bid the price up to near $45 and still expect to make a profit.

By eliminating the some of the downside, there is a higher probability of profit so investors are willing to bid higher.
posted by JackFlash at 1:57 PM on March 27, 2009

you seem to be saying that the bank and the hedge fund will collude in looting the government

Yes, that is exactly what I am saying can happen. It is probably possible to game the system to stick the FDIC with major losses. Apparently the FDIC will have some say in how the process works, and will act to protect their interests to some degree, but the taxpayers haven't been too well protected so far, so I don't know they will start now.
posted by procrastination at 2:03 PM on March 27, 2009

Response by poster: Wait a sec, I think I see where the real leverage is here. First my hedge fund buys ten million shares of Citigroup at $2.50. Then I go in with the government on a hundred million in toxic assets, risking 3 million of my own money. I assume other hedge funds are thinking the same way I am and are also buying this stuff. As the toxic assets are removed from Citi's balance sheet, the stock price inevitably goes up. I sell my ten million shares at $3, covering 100% of my potential loss on the toxic assets and pocketing 2 million in additional profit. If the toxic assets appreciate, I split the profit with the government. If they don't, it's not my problem. And Geithner is happy, because he and I have started to re-inflate the bubble. Through market manipulation, we've once again created value where there isn't any!
posted by markcmyers at 2:37 PM on March 27, 2009

Sure, that sounds good too. If you look at the Citigroup stock price, it was up 300% after the program announcement, so it seems lots of people are thinking along the same lines. I have seen a couple of other potential tricks suggested elswhere. There is money to be made screwing the taxpayer, at least until the off-balance sheet stuff comes home to roost.
posted by procrastination at 2:53 PM on March 27, 2009