Monday, December 31, 2007
93 Banks are not Rational Actors -- or they have just been subsidized
by Ken Houghton
And the more I don't think about it, the more my shoulder hurts.
Because despite the NYT's attempt to paint it as a positive, the results were ugly:
Now, that's nice, but the "discount" rate, since the advent of the Bush Administration, has been higher than the Federal Funds rate, currently by about 50 basis points (1/2 of 1%).
In the grand tradition of the Spherical Cow joke, "Assume a bank is a rational investor." But wait; isn't that a tenet of economics?
So we have to assume these investors are rational. Which means that they knew what they were doing when they bid 40bp above the Federal Funds rate (higher yield, that is, they are taking a lower price) to lend securities to the TAF.
Let us assume, for the sake of argument, that there is an "anonymity premium" (though how much anonymity there can be when the NYT says "93 banks"—not 92 or 94 or some other number—is left as an exercise). It's not going to be 40 basis points. So there is, theoretically, money being left on the table—by precisely those institutions that know when money is being left on the table.*
Unless it's not.
And in that case, we can easily construct the equivalent pricing in the market. The formula is straight algebra: Principle * Interest Rate * (Number of Days)/360.
The banks borrowed $20 billion for 28 days at 4.65%. They're spending $72 1/3 million dollars to exchange some securities for cash. If those securities had been traded at those same prices, at the Federal Funds rate of 4.25%, the banks would have spent $66 1/9 million.
The banks, in theory, left $6 2/9 million on the table.
So, if we look at economic reality—in which a bank would not do that—we have to come to the obvious conclusion: the value the Fed was willing to place upon those securities was at least 8.6% higher (40/465) than the price those securities would have fetched were they used as collateral in the open market.
The "technical" term for the TAF is "corporate welfare." Anyone who tells you otherwise is trying to sell something—probably overvalued securities.
*I'm belaboring this point, but it's central. This is not a consumer hoping to spend $1 on a present and finding out that the present costs $1.01 or even $1.05. This is a financial institution that survives based on not paying $1.01 for an asset that is worth $1.00. While Homo economicus has always been a simplifying proxy for consumers, it's a fundament for financial management.
I've been trying not to think about this quote from CR:
The Federal Reserve said on Wednesday that it had collected bids from 93 financial institutions in its first auction of short-term credit, a turnout that suggested banks might be more inclined to borrow from the Fed under its auction-based system.
And the more I don't think about it, the more my shoulder hurts.
Because despite the NYT's attempt to paint it as a positive, the results were ugly:
Winning bids were awarded at a 4.65 percent interest rate, lower than the Fed’s so-called discount rate, which normally regulates borrowing from the central bank. The discount rate was lowered to 4.75 percent last week.
Now, that's nice, but the "discount" rate, since the advent of the Bush Administration, has been higher than the Federal Funds rate, currently by about 50 basis points (1/2 of 1%).
In the grand tradition of the Spherical Cow joke, "Assume a bank is a rational investor." But wait; isn't that a tenet of economics?
So we have to assume these investors are rational. Which means that they knew what they were doing when they bid 40bp above the Federal Funds rate (higher yield, that is, they are taking a lower price) to lend securities to the TAF.
Let us assume, for the sake of argument, that there is an "anonymity premium" (though how much anonymity there can be when the NYT says "93 banks"—not 92 or 94 or some other number—is left as an exercise). It's not going to be 40 basis points. So there is, theoretically, money being left on the table—by precisely those institutions that know when money is being left on the table.*
Unless it's not.
And in that case, we can easily construct the equivalent pricing in the market. The formula is straight algebra: Principle * Interest Rate * (Number of Days)/360.
The banks borrowed $20 billion for 28 days at 4.65%. They're spending $72 1/3 million dollars to exchange some securities for cash. If those securities had been traded at those same prices, at the Federal Funds rate of 4.25%, the banks would have spent $66 1/9 million.
The banks, in theory, left $6 2/9 million on the table.
So, if we look at economic reality—in which a bank would not do that—we have to come to the obvious conclusion: the value the Fed was willing to place upon those securities was at least 8.6% higher (40/465) than the price those securities would have fetched were they used as collateral in the open market.
The "technical" term for the TAF is "corporate welfare." Anyone who tells you otherwise is trying to sell something—probably overvalued securities.
*I'm belaboring this point, but it's central. This is not a consumer hoping to spend $1 on a present and finding out that the present costs $1.01 or even $1.05. This is a financial institution that survives based on not paying $1.01 for an asset that is worth $1.00. While Homo economicus has always been a simplifying proxy for consumers, it's a fundament for financial management.
Labels: corporate welfare, fiduciary responsibility, FRBOperations, monetary policy
Saturday, December 01, 2007
Political Connections save Householders
by Ken Houghton
There is, of course, no reason to believe there would be anything suspect about such a transaction. Contrast, for instance, this statement:
with the reality:
And the follow-up statement to it:
It's also fraud.
Then again, it's the lovable and much-loved-by-the-Beltway Rudolph, so it must not be important.
In the midst of claims that Northeast real estate is going to drop 15-40%, some transactions are off-market:
As for Hazleton,[Pennsylvania,] the Giulianis are buying more than gas there these days. On Nov. 1, records show, the couple purchased Nathan's childhood home from her parents for an unspecified sum.
There is, of course, no reason to believe there would be anything suspect about such a transaction. Contrast, for instance, this statement:
"Nobody was trying to hide anything."
with the reality:
One document dated June 26, 2000, shows how money from five such offices - the Mayor's Office of People with Disabilities, the Community Assistance Unit, the Assigned Counsel Administrative Office, the Loft Board and the mayor's liaison to the United Nations - was used to prepay an American Express account to the tune of $60,000...
Carbonetti said that the document - dated four days before the end of the city fiscal year - simply showed how unused money from agencies was being used to prepay bills. [emphases mine]
And the follow-up statement to it:
"It's fiscally responsible to anticipate predictable expenses and prepay them," he argued.
It's also fraud.
Then again, it's the lovable and much-loved-by-the-Beltway Rudolph, so it must not be important.
Labels: 2008, fiduciary responsibility, High Finance, Housing Bubble, Mannion, Politics, Republican Party, RudyG
Tuesday, July 17, 2007
All Right, I Give Up: Rating Agencies cannot find barn door to close
by Ken Houghton
However, closing the barn door after the horses escape is one thing; not knowing how to tell if the horses have escaped, or where the door is, is another.
Via Naked Capitalism, the FT discusses rating agency mis- or nonfeasance:
Anyone interviewing for a job as an MBS analyst who didn't mention most of the above would not get a second interview. Except, apparently, at Moody's.
There still should be other agents acting first. (The most reasonable argument against regulation is that, by the time regulators have the information, the problem may be being solved.) But rating agencies are at worst the last resort of the small investor.
Oh, well. So much for that theory?
I'm disinclined to blame rating agencies, who are explicitly not fiduciaries, for delays in downgrading, largely because they will not have updated information so quickly as, say, the service provider or the owner of the securities.
However, closing the barn door after the horses escape is one thing; not knowing how to tell if the horses have escaped, or where the door is, is another.
Via Naked Capitalism, the FT discusses rating agency mis- or nonfeasance:
[Josh] Rosner [a consultant at research firm Graham Fisher} points to an April report from Moody's that showed the rating agency did not consider debt-to-income ratios as a primary piece of data in their mortgage models, although this is generally considered as one of the three key predictors of mortgage default.
In the same report Moody's said it would for the first time request loan level data detailing the structure of adjustable-rate mortgages, the servicer, the month of the first interest rate adjustment and other data that would allow them to analyse risks. S&P admitted this week that it does not receive this kind of granular data on performance of individual loans within the mortgage pools backing the bonds that it rates.[emphasis mine]
Anyone interviewing for a job as an MBS analyst who didn't mention most of the above would not get a second interview. Except, apparently, at Moody's.
There still should be other agents acting first. (The most reasonable argument against regulation is that, by the time regulators have the information, the problem may be being solved.) But rating agencies are at worst the last resort of the small investor.
One revelation that analysts have described as "extraordinary" this week is that S&P has no specific estimate of how much turmoil in the housing market would be needed to force downgrades of the AAA and AA ratings that have been left untouched in this round of downgrades and constitute the bulk of the principal value of most mortgage-backed deals. Moody's also said in an interview that it had no such estimate.
Oh, well. So much for that theory?
Labels: fiduciary responsibility, Moral Hazard, mortgage, subprime
Thursday, July 12, 2007
This Could Be Fun: A Third Post I Don't Have to Write
by Ken Houghton
English translation: Your pension fund is SNAFU.
The perversity of it is that in a vibrant market, you shouldn't have to buy a CDO. But that's another post.
This is why, all those posts ago, I compared this "meltdown" to the Structured Note market. I tried at one point to go short a Structured Note (don't ask) and cover it in the Repo market. After almost a full day, no one could find it. As with CDOs, the assumption is that you will Hold to Maturity. You bought it, you own it—even if it breaks.
And, most preciously:
It's arguably not collusion when everyone, working in their own self interest, does the same thing—just as it wasn't collusion when they bought the paper in the first place. And the glut of hedge funds created in recent years makes it fairly clear that a "barriers to entry" argument will fly about as well as, say, Amaranth did late last year.
If I were thinking lawsuit, I'd be using terms such as "fiduciary responsibility" and "prudent" and looking at mutual fund investments in hedge funds: low-hanging fruit may not be the tastiest, but it is the most accessible.
Felix Salmon gives a briefing of how difficult it would be to show collusion in CDOworld. Select excerpts:
Firstly, the big banks are not generally big holders of CDO tranches. The whole reason why CDOs exist in the first place is that they're a mechanism for moving risk off the balance sheets of the sell side, and onto the balance sheets of the buy side.
English translation: Your pension fund is SNAFU.
Part of the reason is that they don't need to collude in order to hold on to their CDOs. CDOs, by their very nature, are a buy-and-hold investment. There's almost no liquidity in them, and the explicit tradeoff in the CDO market is that investors get a higher coupon by giving up liquidity.
The perversity of it is that in a vibrant market, you shouldn't have to buy a CDO. But that's another post.
You can't short something which never trades, and as far as I know no one is writing credit protection on CDOs. There are lots of people writing credit protection on MBSs, including subprime-backed MBSs, but they're a different instrument entirely.
This is why, all those posts ago, I compared this "meltdown" to the Structured Note market. I tried at one point to go short a Structured Note (don't ask) and cover it in the Repo market. After almost a full day, no one could find it. As with CDOs, the assumption is that you will Hold to Maturity. You bought it, you own it—even if it breaks.
And, most preciously:
One option is to address the general topic without quoting my question [presumably the post's title, "Why There Aren't Anti-Collusion Lawsuits in the CDO Mess," in question form) and without mentioning magic words like "antitrust" and "collusion."
Frankly, I'm not clever enough to answer this question without quoting it and without using the magic words.
It's arguably not collusion when everyone, working in their own self interest, does the same thing—just as it wasn't collusion when they bought the paper in the first place. And the glut of hedge funds created in recent years makes it fairly clear that a "barriers to entry" argument will fly about as well as, say, Amaranth did late last year.
If I were thinking lawsuit, I'd be using terms such as "fiduciary responsibility" and "prudent" and looking at mutual fund investments in hedge funds: low-hanging fruit may not be the tastiest, but it is the most accessible.
Labels: Antitrust, Economics, fiduciary responsibility, hedge funds, subprime
