The Virtues of Price Caps

In the last post I discussed the potential benefits of price caps in the small loan market, one of which was to bring the price down to what consumer price shopping would produce if it were present in that market. Now I would like to turn to the potential benefit of price caps in even the most (albeit still quite imperfectly) price-competitive credit market, the mortgage market.

While superficially appearing to be about price, the primary potential benefit of credit price regulation is that it can rein in risk. Even in the small loan market, the primary problem is not paying high, noncompetitive prices, but the risk of not being able to pay off the principal and then being trapped in debt servitude to a loan shark. This trap imposes social costs and high psychological costs on the borrower. The primary problem in the mortgage crisis has also been risk, the risk of default and foreclosure. Risk is intimately tied to price in both situations, but setting a “fair” or “efficient” price seems to me to be to be secondary. (Then again, I am culturally tone-deaf, so maybe fairness in pricing is really what has motivated usury restrictions over the centuries; some historical accounts, however, place the risk of debt servitude as the primary motivator).

How do we rein in risk? Several possibilities spring to mind. First, we could, in theory anyhow, educate borrowers and require disclosure of risk such that borrowers would demand only low-risk loans. All we know about the psychology of risk, however, demonstrates that to be a hopeless project. Second, we could require a party with limited ability to bear risk to retain risk, so as to give that party an incentive to limit risk. Third, we could constrain loan product structures so as to eliminate those features that increase risk. A fourth approach is to regulate the underwriting process directly. Finally, we could regulate prices so as to give lenders incentives to limit risk. 

Dodd-Frank and its proposed implementing regulations engage in a mish-mash of the first four approaches through consumer financial education (discussed in one of my prior posts), risk-retention requirements, prohibitions on and disincentives to structuring loans in particular ways, and requirements and incentives to underwrite loans using particular criteria.

Dodd-Frank's risk retention, loan structure, and underwriting provisions can be found in, e.g., the ability-to-repay (ATR) standard, the qualified mortgage (QM) definition, risk-retention requirements, and the qualified residential mortgage (QRM) definition. In brief: all mortgages will be required to meet the ability-to-repay standard; lenders will receive some protection from liability to consumers for alleged violations of that standard for loans meeting the QM definition; risk-retention requirements will force lenders to retain some risk rather than unloading it all onto the secondary market; and lenders will be released from risk-retention requirements for loans meeting the QRM definition. For the most part, the QM definition discourages lenders from selling loans with particular structures and the QRM definition encourages lenders to engage in particular underwriting practices, although there are some underwriting rules built into the QM definition and all QRM loans must also meet the QM definition.

The basic risk-retention requirement is that originators must retain 5% of credit risk for non-QRM loans. This provides individual lenders with an incentive to keep risk on non-QRM loans within that lender’s risk tolerance. In addition, a QM loan generally cannot charge upfront points and fees in excess of 3% of the amount borrowed, such that much of the price of the mortgage can only be collected through interest, over time. Although not what is typically meant by "risk retention," this price structure restriction in effect means that lenders retain an interest in seeing that borrowers make payments over time. However, as I have discussed elsewhere, lenders can tolerate more risk than can homeowners, renters and communities, so risk retention alone is insufficient to produce the correct amount of risk for society. 

Dodd-Frank’s mortgage loan structure provisions, which in some sense parallel the Small Loan Acts’ amortization and repayment schedule requirements I discussed in my last post, are more promising. The mortgage loan features that are prohibited or disfavored include negative amortization, interest-only payments, balloon payments, adjustable rates that adjust quickly or by more than a certain amount, and most prepayment penalties. These product features play on psychological biases that can lead consumers to take on significant risk with little or no realization that they have done so. Past analyses of these loan features indicate that, ceteris paribus, they increase risk of default and foreclosure dramatically, which provides some reason to think that Dodd-Frank’s loan structure provisions will be successful in reducing risk. The costs of these provisions are real, in that some borrowers will end up with less than ideal loans, but most of these consumers will probably be able to obtain mortgages with safer structures. The biggest danger is that these provisions could restrict future loan structure innovation that would provide value to consumers and society, such as the income-linked loans suggested by Robert Shiller. Regulators must be alert to this possibility and amend these regulations as needed.

Many of Dodd-Frank's underwriting provisions are more problematic. They swing between extremes, from a broad, ill-defined ATR standard applicable to all mortgages to precisely specified underwriting rules about credit history, debt loads, and loan-to-value (LTV) ratios. 

The ATR standard requires “a reasonable and good faith determination based on verified and documented information that, at the time the loan is consummated, the consumer has a reasonable ability to repay the loan….” Although it is possible that aggressive courts will flesh this out in a manner that chills low-risk lending, the more likely problem is that the standard is so amorphous as to be toothless. We do have some experience with this; since 1994, HOEPA has prohibited engaging in a pattern or practice of making certain high cost mortgages “without regard to the consumers’ repayment ability, including the consumers’ current and expected income, current obligations, and employment,” but that HOEPA provision appears to have done little or nothing to reduce risk. The new ATR standard is stronger – it applies to all mortgages, is paired with documentation requirements that will provide courts with a paper trail, and can be used defensively in individual foreclosure actions rather than needing to be proven on a pattern or practice basis. But, other than in the most extreme cases or where smoking gun evidence demonstrates a lack of lender good faith, the ATR standard will continue to require courts to decide what a “reasonable ability to repay” is, a question outside their expertise as a technical and normative matter. 

The underwriting rules, while they could end up so mild as to provide no more risk reduction than the market would on its own, pose a more serious risk of being so rigid as to cut off access to credit that would have posed little risk. Credit history rules that make sense in one period make little sense in the next. For example, while a foreclosure might have once been such a black mark as to reasonably preclude borrowing for many years, borrowers experiencing foreclosure today may be quite creditworthy soon (and sooner than their underwater former neighbors). The debt load regulations include complex formulations for calculating debt-to-income ratios and residual income, yet even the most complex will not fully capture the true affordability of the loan.  LTV ratio requirements pose the greatest concern. As recent work by Roberto Quercia, Lei Ding, and Carolina Reid demonstrates, 80% or even 90% LTV ratio requirements that aim at creating an equity cushion (as opposed to, for example, a 97% LTV ratio requirement, which I would say is aimed more at ensuring the borrower has the resources and ability to save money for housing needs at all and to encourage a borrower mindset of investment rather than speculation) could substantially restrict access to credit without substantially reducing foreclosure risk. The effect on access to credit for low-income borrowers and minority borrowers, who are less likely to have accumulated wealth or have access to intergenerational wealth transfers, could be particularly devastating. Moreover, the equity stripping of predatory lending feasted on low LTV ratio loans; a high LTV ratio loan, while increasing the odds that a borrower will lack sufficient equity to sell or refinance in face of an income or expense shock, also gives the lender an incentive to lend to borrowers who are unlikely to experience such shocks.

True, the underwriting formulations lenders develop on their own are over- and under-inclusive, but they evolve quickly with experimentation. Federal regulations micromanaging underwriting will not. 

Consider instead the lowly usury restriction – so crude, so Biblical-era – and yet…

To greatly simplify: Let’s set the price cap at P. If a lender makes only one loan at price P, it must limit expected losses on that loan to something less than P. The component of expected losses due to foreclosure is the product of the probability of foreclosure and the losses the foreclosure would produce for the lender. Let’s call that probability of foreclosure R (risk). Let’s call the maximum risk of foreclosure that a price P can sustain on a single loan R of P. 

If a lender originating a pool of loans can only charge price P, it can only make a profit if the pool of loans collectively presents a low enough risk of loss that P covers that loss.  Not every loan needs to make a profit, just the pool, so if some loans can be made at a price above the risk of loss posed, then others can be made at price P even if their expected losses exceed price P.  But limiting price to P will provide lenders with a very strong incentive to originate loans with an average foreclosure risk no larger than R of P, and to the degree that the market for lower risk loans is price-competitive, the average risk lenders will be able to profitably bear will be lower. 

The danger of price caps is that we will set P at the wrong point, either allowing for too much risk or permitting too little risk.  But this is no different than the danger of allowing too much risk or permitting too little risk posed by regulations that micromanage loan underwriting. 

The primary virtue of using price caps to limit risk is that lenders can then decide how to keep mortgage risk within this R of P, selecting whatever combinations of loan structure and underwriting criteria they determine, based on ever-changing data, will do so.  So while a price cap may facially appear to be a blunt instrument, it actually allows for more nuanced, creative, responsive, and accurate underwriting than does Dodd Frank’s proposed set of credit history, debt load, and LTV ratio rules. 

A second virtue of price caps is that they would be easier to enforce, and thus more likely to be enforced, than detailed underwriting rules. This is not to say that a price cap ought to take the simple 6% or 8% form of yore – price caps need to move with the price of money, account for the relative greater size of fixed costs for smaller loans, and be paired with rules that cut off various tricks that could otherwise be used to circumvent the caps. But the resulting regime still seems easier to define and apply, and more difficult for lenders to evade, than rules that attempt to micromanage underwriting.