Bankruptcy

Rates are Low--Where Are All the Refis?

By Adam Levitin

Jean Braucher's post on the Hubbard-Mayer housing market reform proposal points to a really interesting question:  why is it that despite historically low interest rates, there has been relatively little in the way of refinancings?   This is a critical question because the housing market has traditionally been a prime channel through which the Federal Reserve can use interest rates to affect the economy.  I've seen one estimate that the missing refinancings would put $90 billion into the pockets of mortgaged households (around 50 million of 'em) every year, without affecting the federal budget.  That's real a direct-to-consumer annual stimulus of $1800/mortgaged household. So, what's preventing more refinancing activity?  

The Hubbard-Mayer proposal, unfortunately, never really gets into the question of why there aren't more refis.  Instead, it gallops straight to a proposed fix, which, as Jean points out, is tragically naive when it comes to the institutional realities of servicing.  (Chris Mayer and I have had a version of this argument before regarding cramdown.)  

To be sure, some of this discrepancy between low rates and low refi levels can be explained by the large number of borrowers with damaged credit and/or underwater properties.  Part of it (not very much by my reckoning) can also be explained by the mortgage industry shifting personnel from underwriting to loss mitigation.  And banks might be gun-shy with their underwriting because GSE/FHA putbacks are starting to bite.  But I think we might need to look to the question of transaction costs to find our answer.

There's been a dramatic growth in mortgage refi costs since the financial crisis started.  The spread between primary and secondary market pricing is now 100bps higher than that pre-crisis.  That means that the spread between mortgage rates and mortgage bond coupons is now around 1% higher than it had been pre-crisis.  In other words, there's an additional 100bps of transaction costs that have been added.  And that only accounts for loans that are actually closed; to the extent that transaction costs prevent refinancings, the costs are not reflected in the primary-secondary spread.  

What explains the increased transaction costs?  Some of this is from the primary market, which has become more concentrated as many mortgage banks and some major banks have simply disappeared.  Some is from the secondary market:  the GSEs are charging a quarter point on all loans for adverse market conditions and then up to three points on other loans for various credit characteristics (loan level pricing adjustments).   The GSEs' additional charges might make good sense on new originations , but I can't follow the logic of charging points to discourage refis.  The refi at a lower rate only reduces credit risk, so why price adversely to it?  

I think a critical task for the administration is to determine how much of the absence of refis is due to caution from chastened financing institutions, and how much is due to increased market concentration, including the secondary market duopoly in the conventional, conforming market.  If the issue were merely caution on new originations, I'd be inclined toward the former explanation, but when the GSEs are obstructing the refinancing of credit risk that they already hold, I'm less certain.

For all of its problems, I'll give Hubbard and Mayer some kudos.  They're at least trying to be creative.  Would that we would see such a level of initiative from this administration.  

Full post as published by Credit Slips on September 22, 2010 (boomark / email).