Saturday, October 20, 2007

Public and Private Bank Regulators

Revised - see excerpts added at end.
Revised 10/23 - see joint press release regarding capital adequacy rules for conduits at end.

I am not an expert on banks and banking, although I have been a banker and I have read a few books on the subject. Because of that, consider this an opinion piece based on some observations that I would like to share regarding the current credit crises. If you plan to use any of this information for anything important please remember to verify.

My conclusion is that too much of our financial system is currently regulated by rating agencies. This is accomplished by (get ready for this one) - using the market distortion of bank regulation to avoid bank regulation. Now this is not a rant against rating agencies. In fact I like rating agencies and they play a very important role in our economy. But I believe they have been making too many decisions that impact us all in what has effectively become a dual system of public and private regulation. Here is my reasoning:

On one hand, the banking industry is structured to take excessive risk because of all the safeguards built in by regulators. We don’t want banks to fail. The results are bad – people lose money, fewer loans are available and that slows the economy, etc. So, to protect us from that, banks get certain privileges. First, they have a lender of last resort – the FED. Banks, assuming they are not insolvent, can pledge collateral to the FED and borrow from it if they run into a liquidity problem. If they have no assets the FED is willing to take and not enough liquidity to meet withdrawal demands, then they are most likely insolvent. But even then, there is the “Too Big To Fail” doctrine that, we all know from history, is true. The impact of a large bank failure on the economy would be big and bad, so we don’t believe the government would let that happen. Finally, the FDIC insures deposit accounts (within limits) so even if a bank does fail, the depositors get their money back. Look at all the bail out built into the system. This limits the downside risk of lending because in the worst case, who gets hurt really? So, banks have a built-in incentive to go for the gold (make risky high interest loans).

In order to ensure that banks don’t just swing for the fences, there is the flip side called bank regulation. Banks are prohibited from owning certain types of risky assets like stocks and junk bonds. They are also required to hold a certain level of equity capital that is based, in part, on the risk level of the assets they own (the securities, loans, etc.) to provide a cushion against insolvency. If the value of their assets declines, there must be enough equity to absorb it. They also must keep a certain percentage of the deposits they owe to depositors in reserves to help ensure liquidity. These regulations are enforced by regulators and are in place to counter balance the protections we offer banks and prevent them from focusing on volume instead of quality of loans. These checks and balances result in a profitable banking system that is less likely to fail, and that results in better credit ratings for banks. So, all of the regulation helps banks maintain good credit ratings.

Once you take away these regulations, or more precisely, figure out how to avoid them, you take away the check on the imbalance toward too much risk. I believe this is one of the major issues with off-balance sheet vehicles such as SIVs and ABCP Conduits. As noted above, banks are required to have risk-based capital that is calculated based on the risk of the assets they hold. This is the implementation of one of the checks against all of the protections banks are afforded. When it comes to ABCP Conduits, however (including the now infamous SIVs), banks calculate the amount of risk-based capital required by multiplying the line of credit used to enhance the conduit by 10% (the regulations allow them to do this). So, they escape 90% of the risk-based capital requirement. Shouldn’t this lower their credit rating? (Now, I believe these numbers are generally correct, but as I said I am not an expert on this and I do not have access to a law library at this time. I did find some support for these numbers here: http://www.chapman.com/media/news/media.485.pdf.)

Now we see why banks love this structure. They take the interest rate differential between the assets funded by the conduit and the liabilities funding the conduit and earn a fee from it while only using up 10% of the risk based capital they would be required to maintain if they did this on their balance sheet. This jacks up return on equity by, in effect, getting around the capital requirements.

Let’s look at an example, a make believe bank named PLUS 90 Bank. It has lines backing conduits of approximately $77 billion. Lets be very generous and assume that the assets would be risk-weighted at 20% if they were on the balance sheet, so the bank would be required to keep $1.232 billion dollars in capital to support these assets ($77 x .2 x .08). But, because they are not on the balance sheet, the bank needs only $123.2 million in capital ($77 x .2 x .08 x .1). This $1.1 billion dollar difference is why the banks love this structure. They can avoid the capital requirement and enhance the return on equity. Lets do a simplified example:

Suppose a bank acquired assets with $77 billion and funded that purchase with time deposits. They would earn an interest spread on the investment. Since we are using a risk-weight of 20%, these would be pretty good assets and the interest spread would be on the low side. Assume it is 1.5%. What is the return on equity before expenses?

($77,000,000,000 x .015) / 1,232,000,000 = 0.9375, or 93.75%. Leverage!

What happens if the bank takes this whole thing and moves it off-balance sheet?

($77,000,000,000 x .015) / 123,200,000 = 9.375, or 937.5%! Even if the bank makes a 1% fee from the off-balance sheet entity, that’s still ($77,000,000,000 x .01) / 123,200,000 = 6.26, or 626.00%.

Looked at another way, with a 10% capital requirement, the bank can provide 10 times as much credit off-balance sheet as it could on-balance sheet (in fact, they can probably leverage it even more if they can avoid the 10% bank line; is this why there are troubles afoot with the SIV structure?). See what I mean about the regulations? This structure makes a mockery of the basic idea of limiting bank exposure through the regulatory framework. So what, then, does limit bank exposure? Well, the limit is really the market for commercial paper. This market is driven, in large part, by the perceived quality of commercial paper and that is where the rating agencies come in.

Rating agencies examine the structure of conduits and the quality of the assets they hold. They also look at the credit lines from the banks that provide liquidity to the conduit should there be a problem, so in part the rating assigned a conduit depends on the rating of the bank. See, the rating of the bank that is in part determined through the regulatory structure is then used to backstop a conduit, which is used to avoid those very same regulations. Eureka! (For more learning on how rating agencies rate conduits you can go here http://americansecuritization.com/uploadedFiles/Moodys_ABCP.pdf where you will find Moody's Update on Bank-Sponsored ABCP Programs: A Review of Credit and Liquidity Issues from September 2007.) (See below for some excerpts.)

As long as rating agencies opine on a particular conduit, the money flows. If there were a problem and the bank had to actually fund the credit lines that could impact the rating of the bank and in turn the rating of the conduit. Sounds like circular reasoning to me, sort of like making a loan to the CEO of a company and using her stock as collateral. We know how those work out when there’s trouble afoot. In any event, this structure leverages up the amount that can be loaned given any level of total bank capital as long as the rating agencies opine, hence they are regulating the banking system. This would not be a concern if there were no bank lines behind these structures because then bank solvency would not be an issue. Unfortunately, when they are backed by bank lines bank insolvency can become an issue, and how big of an issue is regulated not by the bank regulators but by the rating agencies.

Here are some details from the Moody’s report referenced above. Note the asset classes that are financed through conduit structures. Sounds like traditional bank lending assets to me. (I do not know if these include SIVs.)

“In the United States, Moody’s rates 113 bank-sponsored programs, with total ABCP outstanding of US$522 billion as of June 30, 2007. Of these, 62 are multiseller programs, with US$461 billion outstanding, 22 are securities arbitrage, with US$67 billion outstanding, and 10 are hybrids with US$32 billion outstanding. Looking at the multisellers, we note that these are highly diversified programs. On average, these programs fund 62 different transactions among 10 different asset types. The largest asset types by outstanding amounts are credit cards at 15%, trade receivables at 13%, commercial loans at 11%, auto loans at 10% and securities at 9%. Mortgages also make up 9% of the total, mostly in the form of warehousing lines that fund newly originated mortgages for short periods of time. Highly rated CDOs comprise about 3% of the assets. Note that US multiseller programs typically have 8% to 10% program credit enhancement, nearly covering any single one of these asset classes.” And,

“Securities arbitrage programs are similarly diversified, averaging 126 different individual securities among 7 different asset types. The concentrations by dollar par amount reflect more the term securities market, with CDOs comprising 35%, commercial mortgage loans 17%, residential mortgage loans 13%, home equity loans 12% and student loans 5%. Of these securities, 89% have Aaa ratings from Moody’s, and another 6% have Aa ratings. An additional 5% are rated Aaa by rating agencies other than Moody’s. This leaves less than 1% rated below Aa.”

Regarding the link between ABCP Conduits and their sponsoring banks:

“As we have noted, all three types of bank-sponsored programs – multisellers, securities arbitrage and hybrids – rely on liquidity and credit enhancement provided by Prime-1 rated entities. For most of these programs, most of this support is provided by the sponsoring bank. Therefore the credit rating of these programs is also linked to the rating of the sponsoring bank. Should a liquidity or credit support provider be downgraded below Prime-1, Moody’s would review the ABCP conduit rating and might downgrade the conduit’s rating.” And,

“Moody’s includes a review of the sponsor and the support providers when assigning a rating to an ABCP program. As a result, we believe that those parties are able and willing to perform according to the terms of the agreements supporting ABCP programs in general, and bank-sponsored conduits in particular.”

I welcome all comments and additional learning on this topic. Please feel free to leave your thoughts and point out any faults in my fact base or logic.

For posts I have done on the SIV issue you can scroll down or click on the link to your right. There are two additional posts from the past week.

Update: Here is the announcement of the rule for bank capital in connection with conduits, which can be found here: http://www.federalreserve.gov/boarddocs/press/bcreg/2004/20040720/default.htm

"For Immediate Release
July 20, 2004
Agencies Issue Final Rule on Capital Requirements for Asset-Backed Commercial Paper Programs

The federal banking and thrift regulatory agencies today issued a final rule (393 KB PDF) amending their risk-based capital standards. The rule permits sponsoring banks, bank holding companies, and thrifts (banking organizations) to continue to exclude from their risk-weighted asset base for purposes of calculating the risk-based capital ratios asset-backed commercial paper (ABCP) program assets that are consolidated onto sponsoring banking organizations' balance sheets as a result of Financial Accounting Standards Board Interpretation No. 46, Consolidation of Variable Interest Entities, as revised (FIN 46-R). This provision of the final rule will make permanent an existing interim final rule.

The final rule also requires banking organizations to hold risk-based capital against eligible ABCP liquidity facilities with an original maturity of one year or less that provide liquidity support to ABCP by imposing a 10 percent credit conversion factor on such facilities. Eligible ABCP liquidity facilities with an original maturity exceeding one year remain subject to the current 50 percent credit conversion factor. Ineligible liquidity facilities are treated as direct credit substitutes or recourse obligations and are subject to a 100 percent credit conversion factor. The resulting credit equivalent amount is then risk weighted according to the underlying assets, after consideration of any collateral, guarantees, or external ratings, if applicable. All liquidity facilities that provide liquidity support to ABCP will be treated as eligible liquidity facilities for a one-year transition period.

The rule, which will be published shortly in the Federal Register, will become effective on September 30, 2004. "

For the rule and a good description of ABCP Conduits here is the link to the FED's publication: http://www.federalreserve.gov/boarddocs/press/bcreg/2004/20040720/attachment.pdf

Check out this piece from the FED from the link referenced above:

"The resulting credit equivalent amount would then be risk-weighted according to the underlying assets or the obligor, after considering any collateral or guarantees, or external credit ratings, if applicable. For example, if an eligible short-term liquidity facility providing liquidity support to ABCP covered an asset-backed security (ABS) externally rated AAA, then the notional amount of the liquidity facility would be converted at 10 percent to an on-balance sheet credit equivalent amount and assigned to the 20 percent risk weight category appropriate for AAA-rated ABS."

So, the rating agencies determination of the risk in the conduit is used to determine the amount of capital the bank must keep which in part determines the rating on the Bank which in part determines the rating on the conduit. I' dizzy!

Sphere: Related Content

1 comment:

Palermo's Blog said...

There was a blog post at The WSJ today titled Poole Sees Improvement but Says Contagion Threat Remains. You can find the full post here: http://blogs.wsj.com/economics/2007/10/22/poole-sees-improvement-but-says-contagion-threat-remains/#comment-5676 . The blog was posted courtesy of Sudeep Reddy.

This was my comment, and it relates to this post regarding banks and bank regulation. I wouldn't mind being wrong about all of this, but I wish someone would clarify the whole thing already.

I have a few of problems with this! The first problem is that the only reason to set up a super-siv is to avoid selling the assets at market prices. In a market economy, one bank’s loss would be another’s gain. But because the banks have leveraged their capital by going off-balance sheet with these structures, the banking system cannot take a meltdown in values so here we are, looking right into the face of the too-big-to-fail problem. This is a regulatory failure in my book. How much leverage, including all off-balance sheet entities, is there in the banking system? Mr. Bernanke? Mr. Greenspan? If this is not the issue then why do we need a super-siv and why is treasury involved (and don’t tell me it was to buy lunch because that’s ridiculous). Transparency PLEASE.

My next problem is how this super-siv gets structured and reported. Why is there all of this secrecy? Has anyone found an actual full description, including details, of how this thing will work? If not, then why not? Are’t these publicly traded and regulated institutions? Take this quote from above: “They may be convinced that this paper is worth 60 cents on the dollar and they don’t want to be forced to sell it at 30 cents on the dollar. I really don’t know very much about that, and I suspect that a lot of the people even who run these funds, whose businesses run these funds, don’t know exactly where they stand on that and are out there trying to figure out what to do.” He suspects that “the people” who run these funds ……. Are’t these the major public financial institutions that the FED regulates? Don’t they run “these funds” for a fee? Is this stuff secret even from the FED? Is it because they need to hide the fact that there will be a transfer of an asset worth $60 booked as an asset worth $80 with the $20 difference supporting bank capital? The assets may be worth $30 but they think that the assets are worth $60? What happened to market value? Sounds like a bit of hocus-pocus to me. Transparency, PLEASE PLEASE.

What about this quote: the super-siv is “’an effort to promote some better price discovery of what those assets are really worth,’ Mr. Poole said. ‘I don’t know enough about it and I don’t know whether it’s actually going to work or not. … But it’s the kind of device that you would expect the markets to create — some devices to start moving toward normal.” What? If you want to discover what something is worth, sell it. That’s how a market economy works. And this stuff about what the markets would create to start moving toward normal – code for hide the losses until we hope these assets increase in value or we can have time to reserve against them like we should have in the first place. What the market would do is transfer the assets for what they are worth today and there would be winners and losers, like always (except when the losers are too big to fail).

If this is all wrong, then someone needs to explain why. I understand the larger issues of a complete loss of CP investor confidence and the impact on the economy, but that’s not a secret. So why is there all of this confusion?

I hate to be so cynical, but until someone explains this thing to the satisfaction of the financial community, it all sounds like a bunch of hide-the-ball accounting to me. Wonder what Andy Fastow thinks about all this.