Benito Arruñada - Universitat Pompeu Fabra
+ PAGE SUMMARY    |    Eng · Esp · Cat
Spanish youngsters have not been educated to postpone gratification.
See more

Limiting Mortgage Liability Would Be an Error

Arruñada, Benito (2011), “Limiting Mortgage Liability Would Be an Error”, La Vanguardia, March 20, Dinero, 13.
Download file
  There have been plenty of proposals for limiting debtors’ liability in mortgages. Doing so for existing mortgages would be unfair and rash because it would go against the rule of law. But even for future mortgages it would be a costly mistake. In Spain, as in almost all the developed world, mortgage creditors and debtors prefer not to limit liability. When they exercise their contractual freedom in this regard they do not damage third parties, which might otherwise be a reason for prohibiting it. However, real systematic damage would indeed be caused by the proliferation of limited liability (no recourse) mortgage loans.

Creditors and debtors do not limit liability because it would be detrimental for both. It would worsen their incentives and make default more likely. This would increase interest rates and reduce the number and size of loans, as banks would bear the risk associated with the value of the property. Moreover, this risk might be wrongly allocated. Limited liability on residential mortgages is therefore an anomaly that generates all sorts of problems wherever the law obliges contracting parties to use it—in 11 of the 50 States of the USA.

Unlimited liability encourages debtors to make a greater effort to return the loan. If they face difficulties, they will aim to step up their income or consume less, devoting a larger proportion of their income to paying off the mortgage. But, with limited liability, they stop paying when the property is worth about 10% less than the debt. This leads to the strategic default that plagues the above-mentioned 11 American States, affecting 26% of total mortgages and incurring the high costs of foreclosure.

Moreover, with unlimited liability, debtors will utilize personal guarantees that would be difficult to set up using additional contracts. Above all, they will avoid taking out complementary personal loans over a shorter period and with greater interest, which was common practice in the past when there was less inter-bank competition.

Finally, unlimited liability generates a firmer, more lasting commitment, encouraging debtors to invest in the property. This prevents changes in value from damaging their incentives to maintain it. It also encourages “specific” investments of the sort that are of value only to the current occupier of the property.

The effects of liability on banks’ incentives are relatively secondary, especially when compared to the factors that do distort credit decisions, such as political setting of interest rates and compulsory, risk-blind deposit insurance.

In addition, the effect of liability on credit decisions is limited. When appraisals are inflated, this is neither a mistake nor negligence. It is because the lender positively values the debtor’s overall solvency. If the law were to impose limited liability, the bank would continue lending to the solvent debtor the part that is not guaranteed by the property by means of a complementary personal loan.

Finally, unlimited liability would worsen the bank’s incentives to negotiate, speeding up foreclosures. In a recession, the bank would foreclose with a view to selling the property just as soon as it becomes concerned that the debtor may not pay back the loan. Unlimited liability partly explains the opposite, rather striking phenomenon: that creditors do not require debtors to post an additional guarantee when the property starts to be worth less than the debt it guarantees, as they are entitled to do under Spanish law.

In principle, the benefits obtained with these incentives might be worth less than the cost caused by inefficient risk allocation, if debtors were more risk-averse than creditors. But appearances can be deceptive in this regard. To start with, the risk in question is related to changes in the value of the property and it is unclear what advantages mortgage loans have to channel diversification of this type of risk.

It is also unclear who is the optimal risk bearer. It would be easy for the bank to bear the risk of one property, but not that of all the properties in its loan portfolio, given that their values are heavily correlated. It is well known that the real estate market creates huge risks for banks, sparking many banking crises worldwide, because it is highly leveraged and gives rise to vicious circles between credit volume and collateral value.

Moreover, even if the now compulsory liability of lenders issuing mortgage securities in the secondary market were modified, such lenders would not be able to transfer the risk to other investors without separating the credit decision from risk bearing. And this separation would increase opportunism, as was made clear during the recent crisis in the USA, where securitization was behind the growth of subprime loans. As a result, the US Treasury recommended in February 2011 that whoever decides on a loan should retain part of the risk of default.

Lastly, given that most mortgage loans are form contracts, some people wonder whether debtors really know what it is they are signing. But, if limitation of liability were efficient, competition between banks would lead them to offer it. Moreover, this is a salient clause that has been common for centuries. It seems doubtful that debtors would find it difficult to understand, or that they would be less rational on this than on teasing rates or variable-rate loans.

Unlimited liability therefore seems to be optimal on an individual scale, so it would only be reasonable to prohibit it if there were a serious, irreparable failure in the market, because of lack of competition or negative externalities.

Regarding competition, if limited liability mortgages were efficient, even monopolistic banks would offer them, pushing up interest rates and their profit. Instead, failures might well occur the other way round. In those US States in which the law does not allow unlimited liability contracts, what is worrying is that the deposit insurance may encourage certain banks to ignore the limitation of liability, generating a widespread underestimation of risk in the real estate market.

The same can be said about externalities. The crisis has shown the systemic damage that can be caused by limited liability mortgages. When strategic default is encouraged, recessions get worse and greater damage is caused by foreclosures. And society loses the moderating effect arising from the fact that debtors’ rents are imperfectly correlated.

For all the above reasons, it makes sense for unlimited liability to be almost the universal practice for residential mortgages. The situation in the USA is similar to that in Spain in a crucial sense: in the 39 States in which there is real contractual freedom, mortgages are taken on without limited liability.

We would therefore be well advised to retain contractual freedom in this area. The foreclosure process should be reviewed to speed it up and improve auctions but it should not be forgotten that a change in the rule to dilute liability or make foreclosure more difficult would only damage future contracting parties in order to benefit some of the current debtors, not the poorest ones. The latter, if they have a mortgage at all, have no property or rents worth seizing.