The right weights for mortgage risks

Amit Tyagi Abu Dhabi – Ever since hunter-gatherers started to build their own dwellings some 14,000 years ago, homeownership has been a mark of high social and economic status. Today, the United States has built a mammoth financial bureaucracy to promote homeownership, and economists, bankers, politicians, and of course homeowners themselves keenly follow house-price indicators. But, as we have recently witnessed, unless mortgage risks are properly calculated, the dream of homeownership can all too easily become an economic nightmare. Mortgage debt has become households’ biggest liability throughout the developed world. In the decade preceding the financial crisis, US mortgage debt increased almost threefold, while the real economy grew by only one-third. At its 2007 peak, US mortgage debt stood at $10.6 trillion – more than double the combined GDP of China and India. The six major banking crises in the advanced economies since the mid-1970s were all associated with a housing bust. A real-estate price crash has also been a key cause of emerging-market crises, such as the 1997-98 Asian collapse. Research shows that lost output during recessions accompanied by a housing bust is double or triple what it might otherwise have been had real-estate prices held up. Moreover, housing busts tend to prolong recessions by almost three years. Given the enormous social and economic costs when high levels of mortgage debt meet a crash in house prices, one would expect regulators to be fully engaged on the issue. Unfortunately, they are not. Although the Basel Committee on Banking Supervision, the primary global benchmark for bank regulations, proposed reforms after the 2008 crisis that are aimed at strengthening the financial system, mortgage regulations have changed little. The key to mortgage regulation is the concept of “risk weight,” a measure that increases with the probability that a borrower will default. For example, a loan to AAA-rated Microsoft has a 0% risk weight, meaning that it is virtually risk-free. A loan to the Indian government will have a risk weight of 50%, and one to a riskier Argentina will be weighted at 150%. When it comes to mortgages, the standard risk weight is 35% (down from 50% a decade ago). But banks can use internal mathematical models to calculate their own risk weights. This is akin to marking one’s own exam paper; unsurprisingly, banks’ risk scores are usually on the safe side. Consider, for example, a $200,000 bank loan to buy a $235,000 home. If the bank attaches a 10% risk weight to the loan, the risk-weighted equivalent is $20,000. If the amount of bank capital needed to protect that risk-weighted equivalent is 10%, then the bank requires just $2,000 of its own capital to fund the mortgage, $198,000 of which the bank itself borrows. All it takes to wipe out the bank’s own money is a 16% drop in the house’s price (from $235,000 to below $198,000). US house prices – as measured by the S&P/Case-Shiller National Home Price Index – fell by more than 30% from June 2006 to March 2009, while European countries, including Spain, Ireland, and the United Kingdom, experienced similar or greater declines. The problem is that banks continue to offer big mortgages while having little of their own “skin in the game.” The IMF estimates that the average mortgage risk weight in Europe and Asia was 14% and 15%, respectively, with some banks using a risk weight as low as 6%. Recently, some regulators have become suspicious of banks’ do-it-yourself risk weights. In 2012, the UK authorities asked banks to apply a temporary minimum 15% risk weight to their mortgage portfolios, though that guidance expired in July. The Swedish regulator has recommended that mortgage lenders raise their risk-weight floor from 15% to 25%. Although welcome, these steps may still be too little; more important, they lack flexibility. A better option would be to implement time-varying, pro-cyclical floors, forcing banks to hold more capital during housing booms. And, whereas a single, fixed floor might push banks toward riskier loans that offer higher returns for the same amount of capital, a dynamic system would allow lower floors for the lowest-risk mortgages and require higher floors for riskier loans. Moreover, while risk weight is an important macro-prudential lever, it must be used in conjunction with regulators’ other tools. These include mortgage eligibility criteria, such as loan-to-value and loan-to-income ratios, and dynamic provisioning (that is increasing banks’ mandatory loan-loss reserves during housing booms). It is important to calibrate the impact of risk weights on housing-market cycles relative to other measures, not least because excessive and overlapping intervention could itself fuel dangerous cycles. The reliance of modern banking regulation on risk weighting is nowhere more important than in the mortgage market. If regulators get it wrong, the entire edifice could collapse – yet again.The views expressed in this column are the author’s own, and do not represent those of the National Bank of Abu Dhabi. Amit Tyagi is Vice President and Head of Group Credit Risk and Portfolio Management at the National Bank of Abu Dhabi. Copyright: Project Syndicate.

Opinions

The right weights for mortgage risks