Barney Frank’s Bad Idea
Last month Barney Frank unveiled the House plans to fix the financial services industry. One of the provisions (section 1501) will require that any creditor who originates a loan to retain some of the ultimate risk of non-repayment of the loan. The provision is an apparently sensible response to the pathologies in the originate-to-distribute (OTD) model of mortgage lending that we saw at the height of the subprime boom. The basic idea is that originators were insufficiently incentivized to monitor the credit worthiness of applicants, and therefore manufactured a huge volume of ultimately toxic financial assets. The idea is to fix the problem of agency costs by aligning the incentives of loan originators with loan holders. Despite the plausibility of the proposal, I think that it is ultimately a bad idea.
First, it is a bad idea because it addresses a symptom rather than a cause of financial rot. The problem with the mortgage-brokers-as-villains narrative is that it fails to explain why the brokers could do a land office business selling toxic junk to a voracious secondary market. One explanation – the one implicit in section 1501 – is that brokers were taking advantage of purchasers, selling them supposedly sound financial assets that the purchasers were too unsophisticated or blinded by greed to realize were junk. To state this assumption explicitly is to see its limitations. The purchasers of mortgages were not unsophisticated consumers or little old ladies entrusting their savings to fast talking swindlers. These were a bunch of extremely wealthy, extremely sophisticated, extremely large financial institutions. It is rather unlikely that these guys were “fooled” by the mortgage brokers.
A more plausible story, in my opinion, looks at the underlying supply and demand for credit. First, why did the mortgage brokers go into the subprime market? At least in part the answer is that they could afford to do so. With the short term wholesale funding on which they relied to originate loans costing them essentially nothing, it was extremely inexpensive to originate loans. At the same time, the massive subsidization of the subprime market through implicit guarantees to the Fannie and Freddie, the so-called “Greenspan Put” on which Wall Street relied, and various (admittedly much smaller) direct subsidies created a massive demand for the assets churned out by the mortgage brokers. Add to this the impact of monetary and Chinese balance of payments factors on asset prices, and the notion that the subprime crisis was really the result of agency costs in the OTD model looks implausible. Absent macro-economic and regulatory distortions, I suspect that market competition and reputational sanctions are sufficient to keep the OTD brokers honest. Given those distortions, we have seen spectacular examples of those who did have skin in the game responding perversely to the perverse incentives with which they were presented.
If this were all, the risk retention provision would simply be useless. Unfortunately, it is more than simply a regulation aimed at a phantom villain – in this case the OTD model. As written it is likely to have positively perverse consequences. Section 1501 will create regulations requiring any lender originating a loan to retain some of the risk associated with the loan. Such a rule will potentially play havoc with entirely ordinary and unobjectionable credit transactions. Consider a business that sells its goods or services on short-term or medium-term credit, creating a pool of accounts receivable. It is standard practice for the business to pledge such accounts receivable as collateral on a bank loan. However, should the business default on the loan, under current law the bank would foreclose on the collateral – in this case receivables – and sell them off to satisfy the business’s debt. This foreclosure sale, however, would necessarily mean that the business would no longer retain any of the risk associated with the receivables that it generated, violating section 1501 of the proposed act. In other words, in the name of eliminating what is essentially a symptom rather than a cause of the financial panic, the House proposal seems to put a stake through the heart of garden variety receivables financing.
It gets potentially worse.
It is pretty standard for banks to loan money against inventory. Often the inventory is sold on credit. Inventory financers look to the receivables generated by these sales to satisfy their loans, and under Article 9 of the UCC their security interest automatically attaches to the receivables as proceeds. The analysis above, however, suggests that these proceeds-based security interests are open to the same problems under section 1501 as transactions where receivables are used as original collateral. We are now quite a ways away from the exotic world of Wall Street credit derivatives, potentially sweeping up such thoroughly ordinary transactions as taking a security interest in goods on a retailer’s shelves.
The irony, of course, is that should section 1501 have this consequence, its effects could be bargained around using asset securitization. Rather than pledging the receivables themselves as collateral, an originator could securitize them through a SPV in which it retained some residual risk. The securities created by the SPV could then be held by the originator and they (as opposed to the underlying receivables) could be pledged as collateral to a lender. There is something a bit perverse, however, about creating an extra level of credit derivative complexity in order to bargain around the problems created by regulations designed to simplify credit derivative driven complexity.