Research Suggests Low-Income Home Buyers Not to Blame for Financial Crisis

March 19, 2015

Professor Manuel Adelino studies the intersection of real estate finance, household finance, and entrepreneurship. As an assistant professor of finance at Fuqua, he's especially interested in how housing markets affect real economic activity, both in the good times when home prices are going up, and the bad times of tumbling values and increasing default rates.

Adelino shares his latest work, and how he came to study the financial crisis, in a Fuqua Q&A.


Q: Your latest research addresses the notion that the mortgage crisis was caused by loans extended to low-income buyers. Explain the prevailing theory of the housing market collapse and why you wanted to examine it.

There is a common perception that innovation in the way mortgages were underwritten in the period before the financial crisis meant that low income neighborhoods and households gained access to credit in a way that was not possible before. The proponents of this view argue the supply of credit to low-income households fueled increasing house prices, and was the source of the crash that brought down the financial system in 2008. We wanted to examine real data to see if this was the true root of the collapse.

Q: How did you go about this research?

This working paper (with Antoinette Schoar at MIT and Felipe Severino from Dartmouth) uses data on all mortgages originated in the United States between 2002 and 2006, where we see the size of the mortgage that was taken up, as well as the income reported by the buyers. We also use data from the IRS at the ZIP Code level. Additionally, we have access to a sample of mortgages for which we can see how they were performing monthly, i.e. whether they were still current on their payments, or whether they defaulted. We combine all these data sets and correlate the growth in mortgage origination with both income levels and the growth in income during the boom period.

Q: What did you find, and what are the implications?

Our paper makes a very simple point: that the overwhelming majority of mortgages go to middle income and relatively high income households. This was true in the '90s, but it was still true during the housing boom. A lot of policy and even academic research has focused on why there was so much credit for low income households, and largely missed the role played by the rest of the income distribution. In fact, most of the losses to the financial system stemming from mortgage defaults can be attributed not to low income households defaulting but rather to a very large increase in default rates of middle income and high income households. This realization has important consequences for how we think about the housing boom and the financial crisis, and for how we design policy for preventing such crises in the future.

Q: You were working on your PhD at MIT when the housing bubble burst. How did that influence your research interests?

I was among the first cohort of doctoral students in finance and economics whose research was strongly influenced by the crisis. At the time, I was especially interested in how mortgage securitization—the sale of mortgage debt as securities—affected the ability that banks had to renegotiate home mortgages with underwater borrowers. This was a very important problem in 2009 and 2010, as there was a lot of policy coming out of Washington that was directed at promoting mortgage modifications (including, among others, the Making Home Affordable Act of 2009) that would help delinquent borrowers get back on their feet. Some observers argued that the securitization of mortgages during the housing boom was preventing lenders from making changes to the contracts, and that if banks still held these mortgages on their balance sheets things would be very different. My work with researchers at the Boston and Atlanta federal reserves (Kris Gerardi and Paul Willen) shows that securitized mortgages and those that were held on banks' balance sheets are actually renegotiated at very similar (and very low) rates. We argue in that paper that there is no evidence that securitization is an important impediment to renegotiation, and that banks have other reasons for not wanting to change mortgage terms. We pointed out in particular that many borrowers "self-cure", i.e. they are able to become current on their mortgages again without help. Obviously, in these cases, banks do not want to modify the terms on the contracts. Also, banks do not want to appear "soft", as many borrowers would want to strategically default to obtain similar concessions. Interestingly, the reasons banks had for not renegotiating are very much the same as those being used now in Europe by creditor nations to refuse to renegotiate with Greece and other southern European countries.

Q: You have also studied the connection between housing markets and entrepreneurship. What did you discover?

In a recent paper with David Robinson and Song Ma (both in the finance group at Fuqua) we show that new businesses are very important for the job creation process when a region's housing market is booming or contracting. The focus has traditionally been on the role played by small businesses in job creation, but our work shows that the majority of small businesses are old businesses that simply do not grow. Even when cities are booming, and lots of jobs are being created in the services sector (retail, food, accommodation, and construction), the existing small businesses do not react, and generally do not account for a substantial portion of the new jobs. Instead, over two-thirds of all net new employment comes from new firms being created. This research is part of a broader agenda where we want to understand why those existing businesses do not grow, constraints they face. Financial constraints may play a role, but we are convinced that other factors—like the ability to manage larger organizations—matter more for preventing small business growth.

This story may not be republished without permission from Duke University's Fuqua School of Business. Please contact for additional information.

Contact Info

For more information contact our media relations team at