Thursday, July 03, 2008

Christopher Mayer, Tomasz Piskorski and Alexei Tchistyi Show that Prepayment Penalties Benefit Borrowers with Poor Credit Histories

Their paper find there is a separating equilibrium under which borrowers with blemished credit want prepayment penalties, while those with perfect credit do not. More particularly, they show:

Subprime borrowers with FICO scores below 620 obtain rates as much as 0.7%
lower than similar borrowers with fully prepayable mortgages and default at a lower rate (13% default rate) than comparable borrowers with no prepayment penalties (18% default rate). Our …findings suggest that regulations banning re…financing penalties might have the unintended consequence of raising interest rates, increasing mortgage default, and limiting available credit for the riskiest borrowers.

The paper does only look at fixed rate mortgages, and so it is not clear whether the argument applies to 2-28s. But the finding is well worth considering in light of the current policy debate.


Anonymous said...

Where did you come across this paper? Was it publicized somehow by NBER?

The lenders had no problem selling ppp's as the bubble expanded. Churning -- refinancing the borrowers as soon as permitted -- made the originators tiny fortunes. If you're a subprime borrower and the lender says to you "I'll get you $50k cash out" you will tend to overlook the little $10k penalty while salivating over the big money.

I've wondered where that $10k goes. Does it stop at the originator if the loan has already been sold, or does it get passed on to the holder of the note?

Their discussion of the effects of cash-out refinancings leaves much to be desired.

"we believe that prepayments based on mobility and cash out motives should not unduly bias our findings."

Yes they concede: "data on mortgages typically do not list the type of prepayments, such as a property sale, a straight refinancing (for a lower rate), or a cash-back refinancing."

What? Why is this? There is a check box on the fannie/freddie standard mortgage application that requires you to state specifically whether or not it is cash out. If this data is not aggregated like all the others, then there is some *very* suspect data collection and/or dissemination going on.

30 Year Fixed Rate loans? Cherry Picking. If you're going to do more than throw gobbledeegook down on paper with some equations, expecially the kind that influences policy and consumer behaviour, then you'd better account for the other types of loans as well.

Personally I don't find "limiting available credit for the riskiest borrowers" much of a negative "unintended consequence."

Chicken in every pot,yes. Subprime borrower in every home, with taxpayers holding their risk bag? No.

Austin Kelly said...

The penalty usually stays with the servicer,and I think in most cases big originators keep the servicing, which is why Wachovia was able to waive the penalty on all their option ARM borrowers.

The authors don't know about cash out refis because they are looking at data on the old loan. The reason the loan is terminated is not captured on a loan record. The reason for taking out the new loan, cash out, rate-term, purchase, etc., is captured on the record for the new loan.

Didn't Anthony Pennington-Cross and a couple of co-authors do a paper looking at prepay penalties effect on rates for 2/28s and 3/27s, and find not much of an effect? Fixed rate subprime was a pretty small part of the total market - the action is all in the 2/28s and 3/27s, as the previous poster noted.

The paper that I would LIKE to see written is one that examines how optimal or suboptimal is the exercise of the refi option for people with prepay penalties. You'd need to know that in order to assess whether the break on the rate (if any) received in return for a prepay penalty was adequate compensation, and it would be fascinating in its own right. I suspect that refi behavior for subprime borrowres with prepay penalties is at least as "woodheaded" (in some direction) as is refi behavior for prime borrowers with conventional fixeds.

Anonymous said...

"Wachovia was able to waive the penalty on all their option ARM borrowers"

I had wondered about this as well. I did loans up until a year ago, though none with Wachovia. Over the years it was fairly easy to negotiate a waiver of the ppp, but towards the end it became more and more difficult to the point where they would just say no. Then I found out why. The big lenders had ramped up their "retention" departments and the retail branch was contacting my borrowers directly (Either they were using the trigger lists created when borrowers had credit run or retail had access to the new application info). They *would* waive the ppp, whereas the wholesale branch would not.

In practice, on every loan that I originated, I would prepare a matrix for the borrower. My brokerage fee was fixed with them ahead of time and matrix would show them rate and fees with and without ppp, with and without discount points, etc. I would discuss the ppp with borrower and point out that it was a benefit only if they planned to stay put and not refinance for the pp period. (Standard industry spiel was that people tend to refi or move every 5 years on average).

I don't have any practical experience with sub prime borrowers or loans except one which was a full doc 30 year fixed loan for someone with a very low fico score. They were approved with the automated fnma underwriting system.

I *would* alert borrowers to dropping rates even when they had ppp's, and they would contact me as well to see if it made sense to refinance. I would do a simple analysis for them to show them when their break-even would take place, but more often than not the analysis would show them that it wasn't worth their while since very few of them expected to be in the same home for more than a few year (that constant trading-up we saw).

I didn't sell *any* option arms at all. For most, it was a matter of conscience. With savvy investors, I would usually lose the deal to some other shark who would pre-approve the borrower for far more money than I was comfortable with.

Back on topic -- my experience tells me that ppp's did not benefit borrowers with poor credit histories, especially since they were refinancing frequently and cashing out. If they did a 30 year fixed loan, either they or the originator, or both, had character and they were not prone to churning.

Anonymous said...