Mark Calabria: Coming Full Circle on Mortgage Finance
April 21, 2016
America’s mortgage finance system was once called “the envy of the world,” at least by Americans. Wisely, few other countries choose to emulate our system. The current limbo of Fannie Mae and Freddie Mac offers a once-in-a-generation opportunity to rethink our mortgage finance system. There is much we can learn from both our own history and that of other countries. We should return largely to an “originate-and-hold” model for mortgage finance. But as with any malady, the road to a cure begins with admitting you have a problem.
What are some of the problems with the current system? First, it is characterized by massive leverage on the part of financial institutions. My “back-of-the-envelope” calculation is that the ultimate holders of mortgage risk in the United States were leveraged almost 60-to-1 at the time of the crisis, based upon their minimum required capital standards. This requires a loss of only about 2 percent before the system becomes insolvent. Such leverage was largely the result of the interaction of the Basel III capital accords and mortgage securitization.
Despite oft-heard claims that the current system has largely worked fine for over 80 years, the truth is that our securitization-based system is relatively recent. Until the early 1980s and the implosion of the savings and loan industry, most mortgages were held as whole loans on the balance sheets of depositories. Insurance companies held most of the remaining loans. The securitized share did not even break double digits until around 1980. Our current system arose from the ashes of the savings and loan crisis. When the system was first tested in a housing downturn, it failed. It would be more accurate to say that the current securitization-based system has never survived a housing downturn, rather than to claim long-standing stability.
The recent crisis revealed that securitization has largely been a capital arbitrage that simply shuffled mortgages around from one balance sheet to another (Ambrose, LaCour-Little, and Sanders 2005). We should not forget that. Mortgages, under any system, have to rest on someone’s balance sheet. Like water running downhill, the risk has flowed to the most leveraged sectors of our financial system. This is a recipe for instability.
Nor has the current system delivered gains in terms of homeownership. The homeownership rate today is lower than it was before the era of securitization. The racial homeownership gap is at a 100-year high. Interestingly enough, over the last century, the racial gap was smallest in 1980, at the beginning of the securitization era. In fact, the growth of mortgage securitization coincided with an increase in the racial gap. So whatever the merits of our current system, it has not delivered on the “American Dream.”
The failure of our mortgage finance system to deliver broad-based gains in homeownership or even housing affordability should not come as a surprise, as the primary obstacles to achieving those goals are not related to mortgage finance. The primary barrier is the high price of housing relative to income. Rather than continuing to substitute easy credit for a lack of income growth, the better path would be to address both income growth (or lack thereof) and housing price growth directly. Our current system has left families drowning in debt, hardly the American Dream and completely unnecessary. As Elizabeth Warren recognized in her book, The Two-Income Trap, additional easing of mortgage standards has only left families more indebted as they bid ever higher against each other for a limited supply of housing. The clear solution is to increase that supply of housing, not continue down the path of low down payments.
The most striking evidence of the need for an increased supply of housing can be found in California. Because of the legislated loan limits for Fannie Mae and Freddie Mac, their footprint in California was relatively constrained before 2008. Congress increased the loan limits for high-cost areas in July 2008. This change greatly expanded the footprint of the government-sponsored enterprises (GSEs) in California. Did California housing become more affordable? No, just the opposite. According to the California Association of Realtors, 38 percent of first-time buyers could afford a median-priced home in San Francisco just before the GSE loan limits were raised in 2008.1 That number has dropped to 24 percent. Increasing the GSE footprint in San Francisco made housing less affordable, not more. That shouldn’t be a surprise, as the housing affordability crisis in California has almost nothing to do with the mortgage market.
Who's Watching the Credit Risk?
Twenty years ago, Lew Ranieri (2000) succinctly laid out the goals of mortgage securitization:
The goal was to create an investment vehicle to finance housing in which the investor did not have to become a home loan savant. He or she did not have to know very much, if anything, about the underlying mortgages.
Mission accomplished. We have a system, as we did leading into the crisis, where investors did not either care or know about the underlying credit risk. While some might see that as a critical feature of the system and one to be preserved, I believe it is the most fundamental flaw of our system. A system where investors do not care about credit risk is one where toxic mortgages are not only allowed but encouraged.
To substitute for investor due diligence, we have placed our faith with regulators and politicians. The Dodd-Frank Wall Street Reform and Consumer Protection Act creates new standards for mortgages under the qualified mortgage and qualified residential mortgage rules. But, as I have detailed elsewhere, these new rules are unlikely to increase credit quality and may even increase delinquencies in the next downturn (Calabria 2014). Perhaps that should not be surprising. Our politicians have long been more interested in expanding ever cheaper credit than in promoting economic and financial stability. Any reformed mortgage finance system must align the incentives of investors, borrowers, lenders, and taxpayers. Insulating investors from credit risk fails to do so.
Broader Access Requires Local Knowledge
Even though the growth in securitization accompanied an expanded racial homeownership gap, many argue that securitization increases credit access. Associating securitization and greater credit access without qualification overlooks how lending decisions are made today. The expansion of automated credit scoring turned mortgage underwriting into an assembly line. This development reduced the cost of underwriting for standard borrowers, those with good credit and able to make a down payment.
The standard underwriting system relies heavily on hard, objective information, which is usually financial. Subjective, or soft, information is difficult, if not impossible to utilize in a securitization-based system. Expanding access responsibly requires gathering that subjective information. It also requires that mortgage lenders have appropriate incentives to review soft information and, when appropriate, extend credit based upon it. This is one reason why researchers have found that branch banking networks are critical for expanding access to low-income or thin credit–file borrowers (Ergungor 2010).
One could argue that those who pool mortgages, such as Fannie and Freddie, could rely on local originators to gather and base approval decisions on subjective knowledge. But as the recent crisis demonstrated, mortgage-pooling firms have no incentive to incorporate soft information. Going back to an originate-and-hold model is likely the best avenue for expanding access to borrowers who are truly “creditworthy.” The originate-and-hold model also helps at-risk borrowers avoid default by reducing barriers to modifying mortgages during downturns, because conflicts between investors, trustees, and servicers would be minimized.
Litigating Mortgage Out of Reach
The most critical elements of a functioning mortgage market are respect for property and contract. Mortgage providers are reluctant to extend credit if they do not believe their rights will be respected and protected. This includes the ability to foreclose on the underlying collateral. While every foreclosure is a tragedy, a system that did not allow foreclosures would be even worse. Unfortunately, we continue in the direction where consumer protection law makes legitimate foreclosure more and more difficult. If we wish to see mortgage credit widely available, litigation risk must be reduced. As researchers at the Federal Reserve Bank of Richmond have demonstrated, consumer “protection” policies that appear on the surface to help borrowers can actually leave those borrowers worse off (Athreya, Tam, and Young 2009).
Wrong Liquidity at the Wrong Time
Securitization is often defended as an important provider of “liquidity.” As the last decade demonstrated, however, securitization’s "liquidity effect” is procyclical, providing “too much” liquidity in good times and not enough in bad times. Looking at the jumbo mortgage market, Federal Reserve System researchers found that banks that were heavily dependent on secondary market activities cut back their lending during the crisis, while banks that engage more heavily in portfolio lending continued to extend credit (Calem, Covas, and Wu 2013).
Debt or Equity Funding
An oft-heard defense of the securitization model is that without it, there would not be sufficient funding for the US mortgage market. This claim ignores that funding for the secondary market must come from somewhere. For the most part, it has come from the institutions it was supposed to have replaced. Financial institutions and the Federal Reserve are the primary holders of agency securities. Instead of linking borrowers with retail investors, our current system shuffles around claims among financial institutions. If banks or insurance companies can hold a dollar’s worth of agency mortgage-backed securities, they can hold a dollar’s worth of whole mortgages. One might retort “but then they’d have to hold more capital,” but this claim only exposes the system for what it truly is: massive leverage. If we think increasing leverage is what the mortgage market needs, we could just as easily increase the leverage of banks and insurance companies. That would end badly, but so does disguising that leverage via Fannie Mae and Freddie Mac.
Equally relevant is the division between debt and equity. Because of distortions such as the tax preference for debt, financial market participants prefer to fund investment via debt rather than equity. Yet a dollar that goes into debt could just as easily fund equity, were we to reduce or remove those distortions. A far larger percentage of our mortgage finance system should be equity rather than debt. This should be true for lenders, investors, and borrowers.
Avoiding Another Savings and Loan Crisis
Why go back to an originate-and-hold model when, as demonstrated by the savings and loan crisis, it can fail too? Because it wasn’t the originate-and-hold characteristics of thrifts that caused the crisis. First, it was the lack of market discipline that resulted from the explicit guarantees of deposit insurance. Unfortunately, most reform plans move to an equally flawed explicit guarantee. While we must reform deposit insurance—including a roll-back of its extension under Dodd-Frank—a less leveraged portfolio-based system would rely more on equity and less on government guaranteed debt.
Second, the savings and loan crisis—and the banking crises of the Great Depression—were partly the result of a geographically fragmented system that lacked diversification. In fact, creating the GSEs was largely a Band-Aid to offset the flaws in our then predominately unit banking system. Fortunately, Congress and the states finally removed branching restrictions in the aftermath of the savings and loan crisis. Our largest banks can directly access the capital markets without Fannie, Freddie, or the Federal Home Loan Banks.
Countries such as Canada have managed to have similar rates of homeownership to the United States, with lower defaults and fewer crises, without relying upon securitization to fund the mortgage market (Bordo, Redish, and Rockoff 2015). Even the National Association of Home Builders recognized that among EU countries, those with higher homeownership rates actually had fewer owners with any mortgage debt.2 We should recognize that encouraging more mortgage debt is not the same as encouraging homeownership. In fact, the result has generally been the opposite.
What to Do with Fannie and Freddie?
Because I am largely advocating a return to the originate-and-hold model, what should we do with the secondary market’s biggest players? There may well be value in a smaller secondary market and in the current GSEs. To retain whatever value there is, the current GSE charters should be converted to national bank charters and the GSEs reorganized as bank holding companies (BHCs). Not only could the GSEs continue to pool and securitize mortgages as BHCs, but they could also originate, collect deposits, and engage in other bank activities.
Given the high concentration among the largest banks, adding two large BHCs would immediately bring increased competition to that market. It would level the playing field between the GSEs and the largest banks, in both directions. This would require that the GSEs operate under the same rules as other BHCs. Securities law exemptions and favorable tax treatment would disappear. But the goodwill and human capital within the GSEs would remain. Shareholders would also benefit to the extent that the companies had value. This would require the GSEs to meet bank capital levels. Selling off the government’s preferred shares would assist in this regard.
It is worth noting that converting Fannie and Freddie to BHCs solves one problem while exacerbating another: the too-big-to-fail status of our largest banks. Ultimately all BHCs, including the newly created Fannie and Freddie BHCs, should hold substantially more capital. That capital should also largely be in the form of common equity and calculated on a non-risk-adjusted basis.
Securitization is a false god that failed us. While not without some value, its virtues have been exaggerated, if not illusionary, while its costs have been hidden or ignored. The same holds for government guarantees of credit risk. Hiding costs rarely reduces them. Painful experience has continued to show the opposite. A more stable and affordable housing market would be best served by returning to an originate-and-hold model of mortgage finance. The past failures of that model—its fragmented nature and its extensive government guarantees—can be more easily addressed without a continued resort to massive leverage. We need not prohibit securitization or even mandate originate and hold; we simply need to remove the various subsidies and distortions that tilt the field toward securitization. While modest tweaks of the current system might be more politically palatable, such small tweaks will only push off the issue until the next downturn in the housing market (and yes, there will be another housing downturn). The only responsible path is to fix the fundamental flaws in our current system.
- “Housing Affordability Index,” California Association of Realtors, accessed June 10, 2016, http://www.car.org/marketdata/data/ftbhai/. ↩
- Michael Neal, “A Cross-Country Comparison of Homeownership Rates,” National Association of Home Builders, June 19, 2015, http://eyeonhousing.org/2015/06/a-cross-country-comparison-of-homeownership-rates/. ↩
Ambrose, Brent W., Michael LaCour-Little, and Anthony B. Sanders. 2005. “Does Regulatory Capital Arbitage, Reputation, or Asymmetric Information Drive Securitization?” Journal of Financial Services Research 28 (1): 113–33.
Athreya, Kartik, Xuan S. Tam, and Eric R. Young. 2009. Are Harsh Penalties for Default Really Better? Richmond, VA: Federal Reserve Bank of Richmond.
Bordo, Michael D., Angela Redish, and Hugh Rockoff. 2015. “Why didn’t Canada have a banking crisis on 2008 (or in 1930, or 1907, or…)?” The Economic History Review 68 (1): 218–43.
Calabria, Mark A. 2014. Mortgage Reform under the Dodd-Frank Act. Working paper. Washington, DC: Cato Institute.
Calem, Paul, Francisco Covas, and Jason Wu. 2013. “The Impact of the 2007 Liquidity Shock on Bank Jumbo Mortgage Lending.” Journal on Money, Credit and Banking 45 (1): 59–91.
Ergungor, Ozgur Emre. 2010. “Bank Branch Presence and Access to Credit in Low- to Moderate-Income Neighborhoods.” Journal of Money, Credit an Banking 42 (7): 1321–49.
Ranieri, Lewis S. 2000. “The Origins of Securitization, Sources of Its Growth, and Its Future Potential.” In A Primer on Securitization, edited by Leon T. Kendall and Michael J. Fishman, 31–44. Cambridge: Massachusetts Institute of Technology.