Posted on

US government debt

Greg Mankiw [Harvard] recently quoted a snippet without comment from this opinion piece by Kenneth Rogoff [Harvard]:

Within a few years, western governments will have to sharply raise taxes, inflate, partially default, or some combination of all three.

Reading this sentence frustrated me, because the “will have to” implies that these are the only choices when they are not.  Cutting government spending is the obvious option that Professor Rogoff left off the list, but perhaps the best option, implicitly rejected by the use of the word “sharply“, is that governments stabilise their annual deficits in nominal terms and then let the real growth of the economy reduce the relative size of the total debt over time.  Finally, there is an implied opposition to any inflation, when a small and stable rate of price inflation is entirely desirable even when a country has no debt at all.

Heck, we can even have annual deficits increase every year, so long as the nominal rate of growth plus the accrual of interest due is less than the nominal growth rate (real + inflation) of the economy as a whole and you’ll still see the debt-to-GDP ratio falling over time.

Via Minzie Chinn [U. of Wisconsin], I see that the IMF has a new paper looking at the growth rates of potential output, and the likely path of government debt in the aftermath of the credit crisis.  Using the the historical correlation between the primary surplus, debt, and output gap, they ran some stochastic simulations of how the debt-to-GDP ratio for America is likely to develop over the next 10 years.  Here’s the upshot (from page 37 of the paper):

IMF_US_debt_profile

Here is their text:

Combining the estimated historical primary surplus reaction function with stochastic forecasts of real GDP growth and real interest rates—and allowing for empirically realistic shocks to the primary surplus—imply a much more favorable median projection but slightly larger risks around the baseline. If the federal government on average adjusts the primary surplus as it has done in the past—implying a stronger improvement in the primary balance than under the baseline projections—the probability that debt would exceed 67 percent of GDP by year 2019 would be around 40 percent (Figure 4). Notably, with 80 percent probability, debt would be lower than the level it would reach under staff’s baseline by 2019. [Emphasis added]

So I am not really worried about debt levels for America.  To be frank, neither is the the market, either, despite what you might have heard.  How do I know this?  Because the market, while clearly not perfectly rational, is rational enough to be forward-looking and if they thought that US government debt was a serious problem, they wouldn’t really want to buy any more of that debt today.  But the US has been selling a lot of new bonds (i.e. borrowing a lot of money) lately and the prices of government bonds haven’t really fallen, so the interest rates on them haven’t really gone up.  Here is Brad DeLong [Berkeley]:

[A] sharp increase in Treasury borrowings is supposed to carry a sharp increase in interest rates along with it to crowd out other forms of interest sensitive spending, [but it] hasn’t happened. Hasn’t happened at all:

Treasury marketable debt borrowing by quarterTreasury yield curve

It is astonishing. Between last summer and the end of this year the U.S. Treasury will expand its marketable debt liabilities by $2.5 trillion–an amount equal to more than 20% of all equities in America, an amount equal to 8% of all traded dollar-denominated securities. And yet the market has swallowed it all without a burp…

I don’t want to bag on Professor Rogoff. The majority of his piece is great: it’s a discussion of fundamental imbalances that need to be dealt with. You should read it. It’s just that I’m a bit more sanguine about US government debt than he appears to be.

Posted on

CDS hilarity

I’m paraphrasing James Hamilton here.

A credit default swap is a contract that pays out if a specified event occurs on the underlying security. Normally, and in this case, the security is some debt and the event is a default on that debt.

There was a pile of $29 million in debt. Specifically, they were (based on) subprime loans in California and a bunch of them were already delinquent.

A brokerage firm from Texas started offering (i.e. selling) credit default swaps on the $29 million. Since so many of the underlying loans were delinquent, it seemed a sure thing that a default would occur and the big boys in New York were happy to buy the CDS contracts.  In fact, they were so sure that the debt would default that they were willing to pay up to 80 or 90 cents for a $1 payout in the event of a default.

Two important things then played a role:  First, credit default swaps are traded “over the counter”, so if you buy one from me you don’t know how many other people have also bought from me or how many they each bought.  Second, there are (currently) no regulations on credit default swaps and in particular, there is no limit to the scale of the CDS market against a particular asset.

In this case, the big banks paid about $100 million for CDS contracts that would pay out $130 million if the debt defaulted.

The brokerage firm took the $100 million, paid off the debt entirely (so it didn’t default) and walked away with $70 million.

Posted on

Once more on bankers’ pay

Megan McArdle makes a perfectly sensible point when she writes:

More than one smart analyst thinks that the yearly bonus regime encouraged both traders and their managers to take excess risk. I’m not sure, as an empircal matter, that I buy this argument. Most of those bankers who were allegedly gambling for free with (implicit) taxpayer money in fact lost half or more of their own fortunes in the ensuing crash. From this I infer that they did not, in fact, realize that they were gambling.

I still think that some regulation on bonuses is warranted. Indeed, I think it warranted precisely because the bankers didn’t fully appreciate the risks they were taking. By holding bonuses in escrow for, say, five years, we serve to increase the risk aversion of those bankers.  Megan implies partial agreement with the conclusion, if not the logic, a little later on:

But enforcing, say, a multi-year bonus scheme wouldn’t be terribly destructive, and it might help.

Continuing immediately on, she writes:

On the other hand, if the government starts meddling with the level of compensation, this will be disturbing both because it will not do good things for the American financial services industry, and because, well, who the hell is the government to start telling private firms that are not receiving any taxpayer money how much to pay their employees?

In general I’d agree, but we should also consider the recent work by Thomas Philippon and Ariell Reshef suggesting that remuneration in the finance sector relative to the rest of the economy for a given level of education has been especially high lately.  Here is an ungated version of their paper.  Here is the abstract:

We use detailed information about wages, education and occupations to shed light on the evolution of the U.S. financial sector over the past century. We uncover a set of new, interrelated stylized facts: financial jobs were relatively skill intensive, complex, and highly paid until the 1930s and after the 1980s, but not in the interim period. We investigate the determinants of this evolution and find that financial deregulation and corporate activities linked to IPOs and credit risk increase the demand for skills in financial jobs. Computers and information technology play a more limited role. Our analysis also shows that wages in finance were excessively high around 1930 and from the mid 1990s until 2006. For the recent period we estimate that rents accounted for 30% to 50% of the wage differential between the financial sector and the rest of the private sector. [emphasis added]

… which is prima facie evidence in support of some sort of regulation on remuneration in the finance sector.

Posted on

The Chrysler bankruptcy

This is not a post about how Chrysler might work going forward, nor a post about how dastardly the hold-outs are.  This is a post about the distribution of haircuts and the move from White House-led negotiation to bankruptcy court.

There are broadly four groups of creditors:  The (sole remaining) shareholder, the union/pension-fund, the bond holders and the US government.

Clearly the shareholder should be wiped out.  The question is how much of a haircut everybody else should take.

I believe that by law, the US government would take the smallest haircut (get the largest fraction of their money back) as they’re super-senior, then the bond holders in decreasing order of seniority and the union/pension fund should get the biggest kick in the teeth.  The hold-outs were secured creditors, which means that if the company is liquidated they get a pretty senior claim on the proceeds.

[Update: Duh.  The government isn’t a super-senior bond holder, it’s a preferred share holder, which means that it’s claims, in principle, ought to be subordinate to the bond holders]

As I understand it, the deal on the table had the order differently.  The unions were getting back something like 60c on the dollar, the government 45c on the dollar and the bondholders 25c on the dollar (those numbers are made-up, but indicative).

That conflict between what would normally happen and the deal on offer was what gave rise to this sort of comment from Greg Mankiw:

The Rule of Law — Not!

Via the WSJ, here is the view from a “secured (sic) creditor” of Chrysler:

“Like many others I made the mistake of buying what I believed was ‘value,'” Mr. Gwin says, adding that investors who bought at the time believed the loans were worth more than their market price. “We did not contemplate having our first liens invalidated by a sitting president,” he adds.

As the President intervenes in more and more industries, a key question is how he does it and what he is trying to achieve. Is he trying to reorganize insolvent firms while, as much as possible, preserving the rights of stakeholders as established under existing contracts? Or is he trying to achieve a “fair” outcome as he judges it, regardless of preexisting rules and agreements? I fear it may be the latter, in which case politics may start to trump the rule of law.

Mankiw has an uncanny ability to irritate me at times and although he has a bloody good point, even a vitally important point, this post did irritate me because I suspect that most bankruptcy arrangements aren’t fair, for a few reasons:

First, bond-holders, like equity holders, are ultimately speculators.  We differentiate the seniority of their claims legally, but the fact is that a guy holding a Chrysler bond is just as much of a punter as the dude holding one of the shares.  They (presumably) had the same access to information about Chrysler’s future and they (hopefully) both knew that their investment came with risk.  The idea of one subset of one factor of production being largely protected from the risk of the company’s failure is silly.

Second, employees are not speculators in the same way that the providers of capital are.  The cost of taking your money out of a company’s bonds or shares and moving it to another company is negligible.  The cost of taking your labour out and moving it to another company is significant.  At the very least, you are often geographically tied down while your money is not.  Therefore the socially optimal decision would help insure the employees against the risk of the company failing but leave the capital to insure itself.  Since US unemployment benefits (the public insurance framework) is so measly, it seems reasonable to grant employees partial access to the assets of the company.

Third, in every company to some extent (although varying depending on the industry), the employees are the company.  At an extreme, ask what a law firm would be worth if you fired all the lawyers.  Therefore, even if labour were perfectly mobile, there is a game-theoretic basis for giving the employees a stake in the game:  Principal-Agent problems exist all the way down to the floor sweepers.  This is an argument for German-style capitalism where the workers are also minority shareholders.  You might argue against workers’ representatives on the board of directors, but I do think they ought to have a share holding.

Fourth, even if all of the above balanced out to zero, there might (might!) be be beneficial social welfare to ensuring that the company is an ongoing concern rather than liquidated.  When they pushed Chrysler into bankruptcy, the hold-outs were doing so because they would get more money under liquidation than the deal on the table.  If there is a benefit to social welfare in keeping the company open, there ought to be a way to force the bond-holders to take a hefty haircut rather than liquidating the assets, even – and this is where Professor Mankiw might really get upset – if it wasn’t Pareto improving (the needs of the many …).

Nevertheless – and this is why Mankiw managed to get under my skin on this occassion – I am glad that Chrysler has gone into bankruptcy.

I am glad because even though I largely agree with the White House’s proposal, and even if my four points are all true, it is not the job of the executive to be making these decisions.    There are entire institutions set up for it.  The bankruptcy courts and the judges who preside over them specialise in this stuff.  By all means the White House might make a submission for consideration (as the executive of the country, not just as a stakeholder), but it should be up to the judge to decide.

I suspect, or at least like to imagine, that Barack Obama knows all this already (he is a constitutional lawyer, after all) and that he pushed the negotiation down the path it has taken because politically he needed to be seen to be trying to “save” Chrysler from bankruptcy and economically ne needed to avoid the market turmoil that would have ensued from a sudden move to bankruptcy rather than the tortuously gradual one we have seen.

Posted on

Is America recapitalising all the non-American banks?

The recent naming of the AIG counterparties [press release, NY Times coverage] reminded me of something and this post by Brad Setser has inspired me to write on it.

Back in January, I wrote a post that contained some mistakes.  I argued that part of the reason that the M1 money multiplier in America fell below unity was because foreign banks with branches in America and American banks with branches in other countries were taking deposits from other countries and placing them in (excess) reserve at the Federal Reserve.

My first mistake was in believing that that was the only reason why the multiplier fell below one.  Of course, even if the United States were in a state of autarky it could still fall below one as all it requires is that banks withdraw from investments outside the standard definitions of money and place the proceeds in their reserve account at the Fed.

And that was certainly happening, because by paying interest on excess reserves, the Fed placed a floor under the risk-adjusted return that banks would insist on receiving for any investment.  Any position with a risk-free-equivalent yield that was less than what the Fed was paying was very rapidly unwound.

Nevertheless, I believe that my idea still applies in part.  By paying interest on excess reserves, the Fed (surely?) also placed a floor under the risk-adjusted returns for anybody with access to a US depository institution, including foreign branches of US banks and foreign banks with branches in America.  The only difference is that those groups would also have had exchange-rate risk to incorporate.  But since the US dollar enjoys reserve currency status, it may have seemed a safe bet to assume that the USD would not fall while the money was in America at the Fed because of the global flight to quality.

The obvious question is to then ask how much money held in (excess) reserve at the Fed originated from outside of America.  Over 2008:Q4, the relevant movements were: [1]

Remember that, roughly speaking, the definitions are:

  • monetary base = currency + required reserves + excess reserves
  • m1 = currency + demand deposits

So we can infer that next to the $707 billion increase in excess reserves, demand deposits only increased by $148 billion and required reserves by $7 billion.

In a second mistake in my January post, I thought that it was the difference in growth between m1 and the monetary base that needed explaining.  That was silly.  Strictly speaking it is the entirety of the excess reserve growth that we want to explain.  How much was from US banks unwinding domestic positions and how much was from foreigners?

Which is where we get to Brad’s post.  In looking at the latest Flow of Funds data from the Federal Reserve, he noted with some puzzlement that over 2008:Q4 for the entire US banking system (see page 69 of the full pdf):

  • liabilities to domestic banks (floats and discrepancies in interbank transactions) went from $-50.9 billion to $-293.4 billion.
  • liabilities to foreign banks went from $-48.1 billion to $289.5 billion

I’m not sure about the first of those, but on the second that represents a net loan of $337.6 billion from foreign banks to US banks over that last quarter.

Could that be foreign banks indirectly making use of the Fed’s interest payments on excess reserves?

No matter what the extent of foreign banks putting money in reserve with the Fed, that process – together with the US government-backed settlements of AIGs foolish CDS contracts – amounts to America (partially) recapitalising not just its own, but the banking systems of the rest of the world too.

[1] M1 averaged 1435.1 in September and 1624.7 in December.  Monetary base averaged 936.138 in September and 1692.511 in December.  Currency averaged 776.7 in September and 819.0 in December. Excess reserves averaged 60.051 in September and 767.412 in December.  Remember that the monthly figures released by the Federal Reserve are dated at the 1st of the month but are actually an average for the whole of the month.

Posted on

How to value toxic assets (part 6)

Via Tyler Cowen, I am reminded (again) that I should really be reading Steve Waldman more often.  Like, all the time.  After reading John Hempton’s piece that I highlighted last time, Waldman writes, as an afterthought:

There’s another way to generate price transparency and liquidity for all the alphabet soup assets buried on bank balance sheets that would require no government lending or taxpayer risk-taking at all. Take all the ABS and CDOs and whatchamahaveyous, divvy all tranches into $100 par value claims, put all extant information about the securities on a website, give ’em a ticker symbol, and put ’em on an exchange. I know it’s out of fashion in a world ruined by hedge funds and 401-Ks and the unbearable orthodoxy of index investing. But I have a great deal of respect for that much maligned and nearly extinct species, the individual investor actively managing her own account. Individual investors screw up, but they are never too big to fail. When things go wrong, they take their lumps and move along. And despite everything the professionals tell you, a lot of smart and interested amateurs could build portfolios that match or beat the managers upon whose conflicted hands they have been persuaded to rely. Nothing generates a market price like a sea of independent minds making thousands of small trades, back and forth and back and forth.

I don’t really expect anybody to believe me, but I’ve been thinking something similar.

CDOs, CDOs-squared and all the rest are derrivatives that are traded over the counter; that is, they are traded entirely privately.  If bank B sells some to hedge fund Y, nobody else finds out any details of the trade or even that the trade took place.  The closest we come is that when bank B announces their quarterly accounts, we might realise that they off-loaded some assets.

On the more popularly known stock and bond markets, buyers publicly post their “bid” prices and sellers post their “ask” prices. When the prices meet, a trade occurs.[*1] Most details of the trade are then made public – the price(s), the volume, the particular details of the asset (ordinary shares in XXX, 2-year senior notes from XXX with an expiry of xx/xx/xxxx, etc) – everything except the identity of the buyer and seller. Those details then provide some information to everybody watching on how the buyer and seller value the asset. Other market players can then combine that with their own private valuations and update their own bid or ask prices accordingly. In short, the market aggregates information. [*2]

When assets are traded over the counter (OTC), each participant can only operate on their private valuation. There is no way for the market to aggregate information in that situation. Individual banks might still partially aggregate information by making a lot of trades with a lot of other institutions, since each time they trade they discover a bound on the valuation of the other party (an upper bound when you’re buying and the other party is selling, a lower bound when you’re selling and they’re buying).

To me, this is a huge failure of regulation. A market where information is not publicly and freely available is an inefficient market, and worse, one that expressly creates an incentive for market participants to confuse, conflate, bamboozle and then exploit the ignorant. Information is a true public good.

On that basis, here is my idea:

Introduce new regulation that every financial institution that wants to get support from the government must anonymously publish all details of every trade that they’re party to. The asset type, the quantity, the price, any time options on the deal, everything except the identity of the parties involved. Furthermore, the regulation would be retroactive for X months (say, two years, so that we get data that predates the crisis).  On top of that, the regulation would require that every future trade from everyone (whether they were receiving government assistance or not) would be subject to the same requirementes.  Then everything acts pretty much like the stock and bond markets.

The latest edition of The Economist has an article effectively questioning whether this is such a good idea.

[T]ransparency and liquidity are close relatives. One enemy of liquidity is “asymmetric information”. To illustrate this, look at a variation of the “Market for Lemons” identified by George Akerlof, a Nobel-prize-winning economist, in 1970. Suppose that a wine connoisseur and Joe Sixpack are haggling over the price of the 1998 Château Pétrus, which Joe recently inherited from his rich uncle. If Joe and the connoisseur only know that it is a red wine, they may strike a deal. They are equally uninformed. If vintage, region and grape are disclosed, Joe, fearing he will be taken for a ride, may refuse to sell. In financial markets, similarly, there are sophisticated and unsophisticated investors, and unless they have symmetrical information, liquidity can dry up. Unfortunately transparency may reduce liquidity. Symmetry, not the amount of information, matters.

I’m completely okay with this. Symmetric access to information and symmetric understanding of that information is the ideal. From the first paragraph and then the last paragraph :

… Not long ago the cheerleaders of opacity were the loudest. Without privacy, they argued, financial entrepreneurs would be unable to capture the full value of their trading strategies and other ingenious intellectual property. Forcing them to disclose information would impair their incentive to uncover and correct market inefficiencies, to the detriment of all …

Still, for all its difficulties, transparency is usually better than the alternative. The opaque innovations of the recent past, rather than eliminating market inefficiencies, unintentionally created systemic risks. The important point is that financial markets are not created equal: they may require different levels of disclosure. Liquidity in the stockmarket, for example, thrives on differences of opinion about the value of a firm; information fuels the debate. The money markets rely more on trust than transparency because transactions are so quick that there is little time to assess information. The problem with hedge funds is that a lack of information hinders outsiders’ ability to measure their contribution to systemic risk. A possible solution would be to impose delayed disclosure, which would allow the funds to profit from their strategies, provide data for experts to sift through, and allay fears about the legality of their activities. Transparency, like sunlight, needs to be looked at carefully.

This strikes me as being around the wrong way.  Money markets don’t rely on trust because their transactions are so fast; their transactions are so fast because they’re built on trust.  The scale of the crisis can be blamed, in no small measure, because of the breakdown in that trust.

I also do not buy the idea of opacity begetting market efficiency.  It makes no sense.  The only way that information disclosure can remove the incentive to “uncover and correct” inefficiencies in the market is if by making the information public you reduce the inefficiency.  I’m not suggesting that we force market participants to reveal what they discover before they get the chance to act on it.  I’m only suggesting that the details of their action should be public.

[*1] Okay, it’s not exactly like that, but it’s close enough.

[*2] Note that information aggregation does not necessarily imply that the Efficient Market Hypothesis (EMH), but the EMH requires information aggregation to work.

Other posts in this series:  1, 2, 3, 4, 5, [6].

Posted on

Whyte is wrong to think that Brown is wrong

Writing in Friday’s FT, Jamie Whyte argues that Gordon Brown is wrong to think that regulating bankers’ bonuses to stop the culture of short-term thinking will avoid future financial crises.  He writes:

[I]magine you are the manager of a lottery company. Your job is similar to a banker’s. You sell tickets (make loans) that have a certain probability of winning a prize (of defaulting). To ensure long-run profits, you must set a price for the tickets (charge a rate of interest) that is sufficient to pay out the lottery winnings (cover the cost of defaulting borrowers).

But suppose you were a greedy lottery company manager, concerned more with your own bonus than with your shareholders’ interests. Here is a trick you might play. Offer jackpots, ticket odds and ticket prices that in effect give your customers money. For example, offer $1 tickets with a one-in-5m chance of winning a $10m prize. A one-in-5m chance of winning $10m is worth $2 . So each ticket represents a gift of $1 to its purchaser.

With such an attractive “customer value proposition” you would leave your competitors for dead. And if you limited ticket sales to, say, 1m a year, the chances are no one would win the prize. In most years you will earn $1m in ticket sales and pay nothing in prizes. When someone finally wins the $10m prize, and your company collapses, that will be a problem for shareholders and creditors; you will probably have pocketed a few nice bonuses already.

To prevent such wickedness, Mr Brown may insist that lottery managers be paid bonuses on the basis of long-term profits: five years’, let us say. No problem: simply set the prize at $100m and the chance of winning at one in 50m. Then you will be unlucky if anyone wins in a five-year period, and you can be confident of walking away with a fat bonus.

This is why, even if Mr Brown were right that short-term bonus plans caused the financial crisis, his proposed remedy would not help. Whatever time frame he mandates, it will always be too short. For, like lottery managers, bank managers can manipulate the “risk profile” of the bank so that large losses, although inevitable in the long run, are unlikely during the mandated period.

I like Mr. Whyte’s analogy, but as far as I can see, there are three problems in his logic.  For the sake of some numbers to talk about, I’ll consider the idea of a five-year delay in high-end bankers having access to their bonuses.

First, he’s missing the fact that for his lottery company to offer a prize of $100 million, it’s going to need some backers with much deeper pockets than if his prize is only $10 million.  Whyte quite correctly points out that risk has been mispriced, but provided that it’s got some price, scaling up without a larger customer pool (the equivalent of increasing the leverage of your bank) must come with extra costs.  Even if the wholesale market is willing to stand behind you, one option is to increase the duration until the size needed to outflank it would require bank mergers that would run foul of competition law.

Second, a key feature long-term bonuses is that they accumulate.   If bonuses are awarded annually but placed into escrow for five years, then even if the bad event doesn’t happen until year 10, there will be five years of bonuses available for claw-back.  All the bankers are currently giving up one year of bonuses.  By putting bonuses to one side, we magnify the value at risk faced by the bankers themselves.

Third, we need to recognise that nobdy lives forever, and while one year might not be so long when measured against a career, five years is a serious block of time.  The reputational effects of any failure would be increased and, I hope, institutional memory would be improved.

As I say, I agree that a mispricing of risk lies at the heart of the credit crisis.  I simply disagree with Mr. Whyte on why it occured.  I’m not sure why he thinks it occured, but I think that part of the cause is the short-term nature of bank incentives.

Posted on

How to value toxic assets (part 5)

John Hempton has an excellent post on valuing the assets on banks’ balance sheets and whether banks are solvent.  He starts with a simple summary of where we are:

We have a lot of pools of bank assets (pools of loans) which have the following properties:
  • The assets sit on the bank’s balance sheet with a value of 90 – meaning they have either being marked down to 90 (say mark to mythical market or model) or they have 10 in provisions for losses against them.
  • The same assets when they run off might actually make 75 – meaning if you run them to maturity or default the bank will – discounted at a low rate – recover 75 cents in the dollar on value.

The banks are thus under-reserved on an “held to maturity” basis. Heavily under-reserved.

He then gives another explanation (on top of the putting-Humpty-Dumpty-back-together-again idea I mentioned previously) of why the market price is so far below the value that comes out of standard asset pricing:

Before you go any further you might wonder why it is possible that loans that will recover 75 trade at 50? Well its sort of obvious – in that I said that they recover 75 if the recoveries are discounted at a low rate. If I am going to buy such a loan I probably want 15% per annum return on equity.

The loan initially yielded say 5%. If I buy it at 50 I get a running yield of 10% – but say 15% of the loans are not actually paying that yield – so my running yield is 8.5%. I will get 75-80c on them in the end – and so there is another 25cents to be made – but that will be booked with an average duration of 5 years – so another 5% per year. At 50 cents in the dollar the yield to maturity on those bad assets is about 15% even though the assets are “bought cheap”. That is not enough for a hedge fund to be really interested – though if they could borrow to buy those assets they might be fun. The only problem is that the funding to buy the assets is either unavailable or if available with nasty covenants and a high price. Essentially the 75/50 difference is an artefact of the crisis and the unavailability of funding.

The difference between the yield to maturity value of a loan and its market value is extremely wide. The difference arises because you can’t eaily borrow to fund the loans – and my yield to maturity value is measured using traditional (low) costs of funds and market values loans based on their actual cost of funds (very high because of the crisis).

The rest of Hempton’s piece speaks about various definitions of solvency, whether (US) banks meet each of those definitions and points out the vagaries of the plan recently put forward by Geithner.  It’s all well worth reading.

One of the other important bits:

Few banks would meet capital adequacy standards. Given the penalty for even appearing as if there was a chance that you would not meet capital adequacy standards is death (see WaMu and Wachovia) and this is a self-assessed exam, banks can be expected not to tell the truth.

(It was Warren Buffett who first – at least to my hearing – described financial accounts as a self-assessed exam for which the penalty for failure is death. I think he was talking about insurance companies – but the idea is the same. Truth is not expected.)

Other posts in this series:  1, 2, 3, 4, [5], 6.

Posted on

How to value toxic assets (part 4)

Okay.  First, a correction:  There is (of course) a market for CDOs and other such derivatives at the moment.  You can sell them if you want.  It’s just that the prices that buyers are willing to pay is below what the holders of CDOs are willing to accept.

So, here are a few thoughts on estimating the underlying, or “fair,” value of a CDO:

Method 1. Standard asset pricing considers an asset’s value to be the sum of the present discounted value of all future income that it generates.  We discount future income because:

  • Inflation will mean that the money will be worth less in the future, so in terms of purchasing power, we should discount it when thinking of it in today’s terms.
  • Even if there were no inflation, if we got the money today we could invest it elsewhere, so we need to discount future income to allow for the (lost) opportunity cost if current investment options generate a higher return than what the asset is giving us.
  • Even if there were no inflation and no opportunity cost, there is a risk that we won’t receive the future money.  This is the big one when it comes to valuing CDOs and the like.
  • Even if there’s no inflation, no opportunity cost and no risk of not being paid, a positive pure rate of time preference means that we’d still prefer to get our money today.

The discounting due to the risk of non-payment is difficult to quantify because of the opacity of CDOs.  The holders of CDOs don’t know exactly which mortgages are at the base of their particular derivative structure and even if they did, they don’t know the household income of each of those borrowers.  Originally, they simply trusted the ratings agencies, believing that something labeled “AAA” would miss payment with probability p%, something “AA” with probability q% and so on.  Now that the ratings handed out have been shown to be so wildly inappropriate, investors in CDOs are being forced to come up with new numbers.  This is where Knightian Uncertainty is coming into effect:  Since even the risk is uncertain, we are in the Rumsfeldian realm of unknown unknowns.

Of course we do know some things about the risk of non-payment.  It obviously rises as the amount of equity a homeowner has falls and rises especially quickly when they are underwater (a.k.a. have negative equity (a.k.a. they owe more than the property is worth)).  It also obviously rises if there have been a lot of people laid off from their jobs recently (remember that the owner of a CDO can’t see exactly who lies at the base of the structure, so they need to think about the probability that whoever it is just lost their job).

The first of those is the point behind this idea from Chris Carroll out of NYU:  perhaps the US Fed should simply offer insurance against falls in US house prices.

The second of those will be partially addressed in the future by this policy change announced recently by the Federal Housing Finance Agency:

[E]ffective with mortgage applications taken on or after Jan. 1, 2010, Freddie Mac and Fannie Mae are required to obtain loan-level identifiers for the loan originator, loan origination company, field appraiser and supervisory appraiser … With enactment of the S.A.F.E. Mortgage Licensing Act, identifiers will now be available for each individual loan originator.

“This represents a major industry change. Requiring identifiers allows the Enterprises to identify loan originators and appraisers at the loan-level, and to monitor performance and trends of their loans,” said Lockhart [, director of the FHFA].

It’s only for things bought by Fannie and Freddie and it’s only for future loans, but hopefully this will help eventually.

Method 2. The value of different assets will often necessarily covary.  As a absurdly simple example, the values of the AAA-rated and A-rated tranches of a CDO offering must provide upper and lower bounds on the value of the corresponding AA-rated tranche.  Statistical estimation techniques might therefore be used to infer an asset’s value.  This is the work of quantitative analysts, or “quants.”

Of course, this sort of analysis will suffer as the quality of the inputs falls, so if some CDOs have been valued by looking at other CDOs and none of them are currently trading (or the prices of those trades are different to the true values), then the value of this analysis correspondingly falls.

Method 3. Borrowing from Michael Pomerleano’s comment in rely to Christopher Carroll’s piece, one extreme method of valuing CDOs is to ask at what price a distressed debt (a.k.a. vulture) fund would be willing to buy them at with the intention of merging all the CDOs and other MBSs for a given mortgage pool so that they could then renegotiate the debt with the underlying borrowers (the people who took out the mortgages in the first place).  This is, in essense, a job of putting Humpty Dumpty back together again.  Gathering all the CDOs and other MBSs for a given pool of mortgage assets will take time.  Identifying precisely those mortgage assets will also take time.  There will be sizable legal costs.  Some holders of the lower-rated CDOs may also refuse to sell if they realise what’s happening, hoping to draw out some rent extraction from the fund.  The price that the vulture fund would offer on even the “highly” rated CDOs would therefore be very low in order to ensure that they made a profit.

It would appear that banks and other holders of CDOs and the like are using some combination of methods one and two to value their assets, while the bid-prices being offered by buyers are being set by the logic of something like method three.  Presumably then, if we knew the banks’ private valuations, we might regard the difference between them and the market prices as the value of the uncertainty.

Other posts in this series:  1, 2, 3, [4], 5, 6.

Posted on

How to value toxic assets (part 3)

Continuing on from my previous thoughts (1, 2, 3, 4) …

In the world of accounting, the relevant phrase here is “fair value.”  In the United States (which presently uses a different set of accounting requirements to the rest of the world, although that is changing), assets are classified as being in one of three levels (I’m largely reproducing the Wikipedia article here):

Level one assets are those traded in liquid markets with quoted prices.  Fair value (in a mark-to-market sense) is taken to be the current price.

Level two and level three assets are not traded in liquid markets with quoted prices, so their fair values need to be estimated via a statistical model.

Level two assets are those whose fair value is able to be estimated by looking at publicly-available market information.  As a contrived example, maybe there is currently no market for a particular AA-rated tranche of CDOs, but there are recent prices for the corresponding AAA-rated and A-rated tranches, so the AA-rated stuff should be valued somewhere in between those two.

Level three assets are those whose fair value can only be estimated by appealing to information that is not publicly observable.

These are listed in the U.S. Financial Accounting Standards Board (FASB) Statement 157.  In October of last year, the FASB issued some clarification/guidance on valuing derivatives like CDOs when the market for them had dried up.

Brad DeLong, in early December last year, was given a list of reasons from Steve Ross why we might not want to always mark-to-market (i.e. assume that the fair value is the currently available market price):

  • If you believe in organizational capital–in goodwill–in the value of the enterprise’s skills, knowledge, and relationships as a source of future cash flows–than marking it to market as if that organizational capital had no value is the wrong thing to do.
    • Especially as times in which asset values are disturbed and impaired are likely to be times when the value of that organizational capital is highest.
  • If you believe in mean reversion in risk-adjusted asset values, mark-to-market accounting is the wrong thing to do.
  • If you believe that transaction prices differ from risk-adjusted asset values–perhaps because transaction prices are of particular assets that are or are feared to be adversely selected and hence are not representative of the asset class–than mark-to-market accounting is the wrong thing to do.
  • If you believe that changes in risk-adjusted asset values are unpredictable, but also believe:
    • in time-varying required expected returns do to changing risk premia;
    • that an entity’s own cost of capital does not necessarily move one-for-one with the market’s time-varying risk premia;
    • then mark-to-market accounting is the wrong thing to do.

Other posts in this series:  1, 2, [3], 4, 5, 6.