Recently I contacted Eric Belsky, Director of the Joint Center for Housing Studies (JCHS) at Harvard University, to discuss the reliability of quantitative methods in the wake of the mortgage crisis of 2008. We had no sooner sat down in his sunny office in Harvard Square when the issue of modeling came up. I had a list of other issues I wanted to discuss, but this subject kept coming up, and so I went with it.
The term “modeling” refers to a quantitative set of techniques that helps financial institutions assess risk. In the case of mortgages, this means the risk involved when buying a home, both to the financial lender and the borrower. Ideally, the risk for both parties would be determined as less than the expected payoff.
“Every financial body,” Belsky explained, echoing an article he published on the subject, “tries to model performance. I don’t think this will ever leave the system and I don’t think it ever should.”
But quantitative risk assessment lost face in the mortgage crisis of 2008. If it had failed to predict such a calamity, how can we ever trust this method ever again?
The answer, for Belsky, is in modifying these models rather than discarding them completely. Most importantly, the assumptions that inform them have undergone a radical change. Before 2008, a major downturn in the housing market was generally thought to be impossible — or at least highly unlikely. Now, no one in housing finance will see things again in the same light. The 2008 mortgage crisis, according to Belsky, “created an historical record of a big, ugly meltdown… modeling around assumptions that housing would not decline significantly was misguided.”
The effects of such a change in modeling could be dramatic. First, for consumers, assuming higher risk from the outset would ensure that they would provide them with a buffer in the case of a severe downturn. Conversely, they would place themselves in a better position to take advantage of the opposite scenario. Although planning with more risk in mind would mean higher mortgage rates, this additional cost would pay off in the long term.
Another positive effect could be to restore faith and trust in the financial system as a whole — which is crucial for the housing market’s recovery. After 2008, an enormous amount of skepticism has emerged about the accuracy of prices. This, in turn, has led to a tightening of underwriting standards.
But what about the potentially negative effects of this change? Belsky rightly worries about the impact of new risk assessment models on low-income urban communities. As the financial industry increases its standards, low and low to middle-income groups are becoming modern-day untouchables. “This is the social question,” Belsky added, “the urban question. What we worry about now is the effects of overcorrection.”
If low-income communities bore the brunt of the effects of the mortgage crisis, then, it seems that they may also be the victims of its solution.