Key elements of recent policy initiatives, such as the revised HARP program and the Fed's "Twist," have focused on either pushing down fixed mortgage rates or taking advantage of their current low levels. In addition to the Fed's actions, many observers have placed a devotional faith in the Fed's ability, as Lawrence Summers wrote in a recent op-ed piece in the Washington Post, to "...support demand and the housing market by again expanding purchases of mortgage-backed securities." However, an underappreciated threat to the mortgage and housing markets is the reliance of many American borrowers on low Libor rates. This means that the Fed and other policymakers must be aware of the vulnerability of the U.S. housing markets to rate spikes resulting from upheavals in the European financial system.
This dependence results from the fact that many borrowers still have adjustable-rate loans linked to either six- or 12-month Libor. While data on the exact composition of the outstanding mortgage market is limited, it's reasonable to believe that somewhere between one-fifth and one-quarter of outstanding loans have adjustable rates, most of which are indexed off Libor. Moreover, many outstanding ARMs are subprime loans with potentially onerous features. While most prime hybrid ARMs were issued with gross margins of 225 basis points over 12-month Libor, subprime loans typically have reset margins of 500-700 basis points. Depending on how "subprime" is defined, I'd estimate that roughly $400 billion in subprime ABS remains outstanding, which equates to more than 2 million loans.
Low money-market rates have provided a subtle tonic to the housing markets and the economy. Prime ARM borrowers are currently enjoying some of the lowest mortgage rates in memory; a loan with a margin of 225 over 12-month Libor currently has an interest rate of somewhere around 3.25%. While subprime borrowers have much higher rates, the very low levels of Libor have kept their note rates in the area of 5.5%-7.5%, which is close to or below their starting rates. (Since the bulk of these loans were issued between 2004 and 2007, most ARMS are now fully adjustable, i.e., the fixed periods for these loans have expired.)
The danger is that Libor rates begin to climb dramatically as the result of European financial dislocations. This would be reflected in wider spreads between Libor and U.S. government rates such as CMT. A noticeable widening of these spreads has already occurred; since the beginning of August, for example, the spread between 12-month Libor and one-year CMT has widened by roughly 30 basis points. For context, note that this spread (which averaged 29 basis points between January 2001 and June 2007) peaked at 308 basis points in October 2008. A sustained return to that level implies 12-month Libor in the area of 3.25%. In turn, this would push prime ARM rates up to around 5.5% (subject to periodic caps) and immediately pressure household cash flows and spending. More ominously, many subprime loans would reset quickly to double-digit rates. This would have catastrophic consequences for many subprime borrowers already struggling to make mortgage payments and unable to refinance into cheaper loans.
I believe that regulators and policymakers have paid too little attention to this looming threat to the mortgage and housing markets. Lending policy initiatives should include incentives to induce borrowers with ARM loans to refinance into fixed-rate products, as well as ways to cushion the potential rate shock to subprime borrowers. In addition to considering unconventional measures designed to keep Libor rates relatively low in the event of a full-blown banking crisis, the Fed must recognize the mortgage market's continued dependence on low short-term interest rates.
Bill Berliner is Executive Vice President of Manhattan Capital Markets. He is the co-author, with Frank Fabozzi and Anand Bhattacharya, of the recently-released second edition of Mortgage-Backed Securities: Products, Structuring, and Analytical Techniques. His email address is firstname.lastname@example.org.