On last week’s This American Life, Chris Arnold of NPR did a good segment on loan servicers and why they do or do not modify loans for delinquent borrowers (starting around the 10-minute mark). There isn’t a lot that avid readers won’t know already; the central message is that it would be better for everyone involved – including lenders and investors – if more loans were modified. It also doesn’t address the legal issues created by collateralized debt obligations where the tranches have different priorities. But if you’re confused about the basics, it’s worth listening to.
Still, there were a couple things that were new or interesting to me:
- Scott Simon, a managing director at PIMCO (the world’s biggest bond fund manager), said that he thinks loan servicers should be modifying more mortgages; that seems like a pretty clear vote from the investor side.
- The segment brings up the issue of computer systems, which is something I hadn’t thought of but should have. Apparently, most if not all of the big, bank-owned servicers don’t have computer systems (software) that can estimate the net present value of a foreclosure as opposed to a modification, taking into account zip code-specific repair costs, broker’s fees on the sale, closing costs, foreclosure-specific legal costs, and expected sale proceeds. Big-company information technology is something I know well, and I can say with a high degree of confidence that if they started designing these things in 2007, they won’t be done until sometime next year at the earliest, and there’s a good chance they won’t work, and even if they do they will have difficulty handling the load. On the other hand, one good product manager and ten good developers in Silicon Valley could probably build something better in about 12-18 months. I sure hope the fate of the economy doesn’t depend on custom homegrown software.
By James Kwak
The Baseline Scenario 
I listened to that program, and it really annoyed me how they said something like (and I’m paraphrasing very liberally), “There are also other reasons banks can’t modify mortgages, including regulations…”
And yet, they never really tell you what those regulations are – though I suspect that I know (if they modify the mortgage they might run afoul of regulations that force them to hold a certain amount of capital or assets or whatever). It very well might be that they’re a more significant than the IT problem that they describe, but they didn’t go into them so they didn’t really allow the listener the opportunity to make that judgment themself.
As best as I recall, they never mentioned the usual “moral hazard” argument that by facilitating modifications they would encourage others to fall behind and ask for modifications. Since I’m sick of hearing that argument, I guess I’m pretty glad they skipped it, but I suppose there are legitimate questions to be raised in that vein.
As to developing the software, I think you are overstating the difficulty, at least of the software engineering aspects. Starting with the observation that almost anything is better than the nothing they have, something as simple as an Excel spreadsheet would be a good start. The real challenge is collecting the data on the various costs. Once again, though, some crude but reasonable approximations would be better than sitting in the dark like they’re doing now. Besides, the show tells us there is at least one firm, having specialized in troubled mortgages, that’s already put this together. Anyone ask if they’re willing to sell? Barring that, the big banks presumably have done enough foreclosures already to be able to search their own databases for the various fees. That leaves the tough one, the resale value. I suspect there are any number of real estate consulting firms who can provide that data. Bottom line, I don’t think it would take 2 years, or even 12 months if someone were actually motivated to make this happen.
Maybe I’m naive, but don’t mortgage lenders learn how to do calculations in their training? Certainly, they could figure out the monetary numbers on the difference between what a foreclosure would cost them, compared to what a loan modification would bring them…by using regular math, like with pencil, paper, and calculator?
Even calculating a simplified example would be better than refusing to modify troubled loans.
The NPR report that says lack of information/the ability to assess delinquent loans limits the ability of the majority of big banks/mortgage loan servicers to properly price loans flies in the face of a fundamental assumption of our “free market capitalism”. The “Efficient Market” theory says all information that can be know, is known and reflected in the price.
It may be true that big banks may need years to develop effective programs to assess defaulting loan values. But, I bet in our dynamic enterprenuerial system, a few smart (unemployed) investment bankers/computer geeks could calculate reasonably good values on an Excel spreadsheet in a few days. They could make tons of money on the arbitrage value of reasonably good information versus no information. Alternatively they could license the software to the Banks who could make tons of money with the improved, even if not perfect information.
Again this “lack of valuation” issue does not make sense if our theory of markets and capitalism is valid.
The implicit assumption seems to be that “the lack of computer systems” is merely a problem of writing the software that will apply a series of models and formulas to accurately predict the foreclosure / modification present values in the loan database. I’m not sure that’s correct.
I think the problem is that the present downturn is proving we don’t have these trusted formulas models – and so even if we had the computer system, they might not do much good.
If we could predict those present values, couldn’t we also make significant progress towards pricing the mortgage-backed securities? If there were good general agreement on what prices would emerge from foreclosure sales or widespread modifications – the market for the toxic legacies would probably begin to clear as the sellers came to grips with reality when a value consensus emerges in their field.
Indy makes a crucial point – and I think. Some of the others are getting a little swept up in the technology argument.
The bottom line is that most banks can’t – or don’t want to – accurately price the assets on their books, MBS included.
We know – and have empirical evidence backing it up – that modifications are preferable to foreclosures for most situations. Everybody wins, from the homeowner to the bank to the investor. We don’t need a complicated workbook or software application to tell us that.
I also think this hits on something I have picked up on – in today’s financial industry we have fancy software and uber-powerful servers which allegedly give us the information we need to make sound judgements. In reality, however, these models and computers are as susceptible to flaws as the people who wrote them. We don’t always treat the information we get from these devices with the caution we should. My point: why write a program to tell us what we know – modifications are almost always preferable AND each situation is unique and needs to be dealt with individually. We can start by giving the right people the right incentives to make this happen.
I think the problem is that the present downturn is proving we don’t have these trusted formulas models – and so even if we had the computer system, they might not do much good.
What about PIMCO? They seem to have hit upon a good formula…I find it difficult to believe others couldn’t do they same.
If we could predict those present values, couldn’t we also make significant progress towards pricing the mortgage-backed securities? If there were good general agreement on what prices would emerge from foreclosure sales or widespread modifications – the market for the toxic legacies would probably begin to clear as the sellers came to grips with reality when a value consensus emerges in their field.
From what I understand, the only reason this hasn’t happened already is because early on the government put out the message that it would be willing to pay above the market rate for these toxic assets, and so banks had no incentive to try to find their true value and sell them – all they had to do is sit around and wait for the government to buy them at a price much higher than the market would have. So I don’t think this is a legitimate reason why the banks can’t readjust the mortgages.
One thing your model would need to take into account is the probability that the borrower would still default on a modified mortgage, which is high.
I think the idea that modifying a loan will yield more than foreclosure sale is a carryover from the days when loans actually made economic sense to begin with, as opposed to being made according to how fast they could be packaged and sold.
If the loan was made based on traditional underwriting criteria, and the borrower just happened to have a spot of bad luck (lost a job, etc.), sure, maybe modifiying the loan would save the arrangement. But many of these loans were made without regard to whether the borrowers could afford them (stated income loans, loans sold with the assumption that they would be refinanced in the near future, etc.) and many of the borrowers have little to no initial investment to lose.
With these loans, you’d really just be trying to determine what the value of the loan is when it is priced on par with the borrower’s housing alternatives. I don’t think it is a foregone conclusion that this will yield more than a foreclosure sale.
Bingo!
I think that’s the problem. Many of those who are defaulting now couldn’t afford the house/mortgage in the first place no matter what the terms of the loan were, and the deteriorating labor/economic environment amplifies the problem and makes them not afford the modified mortgage either.
So modifying the mortgages now would be a waste of time and money, which the banks right now can not afford, nor are they in the mood of doing (short of political pressure).
Действительно удивили и порадовали :) Никогда не поверил бы, что даже такое бывает :)
A month at most perhaps to identify and contract for the data required, a week to agree on the math, another week to create and test the programming, another week to make a pretty interface and then a week to test the whole thing offline. Even this sounds generous, doesn’t it? And this amounts to two months. Programmers, what say ye? Straw men?
Bond Girl has it right below. It doesn’t matter. We’re going to be seeing such horrendous unemployment that the lenders’ time is much better spent trying to squeeze every last penny out before the pennies are gone.
I took Kwak’s post to be on target. Sure, not everyone will qualify for a modification, but many should, many more than the banks seem inclined to give at this point. On Planet Money this week, there was someone who would qualify for Obama’s mortgage plan, and was getting nowhere with their bank, but magically, after NPR reporters kept calling, he got approved the day after the podcast aired.
I agree that even primitive spreadsheets with a couple of knowledgeable realtors in a region should be able to do better than absolutely nothing. It won’t take a fancy-shmancy project, and as I recall there are lot’s of folks in the RE to be hired now.
So, the good is being lost because folks want the perfect solution in the middle of a meltdown. Total FAIL imo.
Always hazardous to extrapolate from your own particulars to the general public (see Edmund Andrews), but here goes.
While the first wave of defaults (esp. in CA, FL, NV) were largely to those who never should have been given the loan, now we are well into the second wave (generalized nationwide) and it should no longer be presumed that mortgages were unaffordable to begin with.
We are starting to see the fruits of a terrible economy, strangled credit, high unemployment and fear that the bottom has not been hit (rightly so).
Believe me, when my husband and I bought our house in ’04 we COULD afford it. After a lay-off in December we cannot. New jobs in this economy pay less and are harder to get than they were two years ago. That’s across the board.
So please stop assuming current defaulters are the same as last year’s defaulters.
I’m not making this assumption at all. There will be multiple waves of defaulters who used exotic products and there likely is a new wave of defaulters due to economic circumstances. My point was that we cannot treat these borrowers in the same way. Servicers should work with borrowers that have traditional loans that are working their way through some misfortune. But people who were speculating on real estate or their future earnings (to put it nicely), or just flat-out committing fraud… I think it is unlikely that much will come out of working with them, and the government forcing banks to try to treat these borrowers as if they were in your position will just draw out the process of dealing with reality.
Candidly, without regard to the imperfect command and control technologies of the servicers, the fundamental reason that mass mortgage mods are going to be very tough to acheive is that they are high touch propositions on the borrowers’ side of the modification.
An annectdotal analogy–last summer my wife’s step dad died. He left a modest, non-taxable estate, almost entirely to his wife of 30 some odd years. There has been utterly no controversy or dispute involved in the resolution of his final affairs, and the only complication has been his handwritten will.
There is, however, enough substance to the estate that it must be probated, a final tax return has to be prepared, etc. There has been a great deal of handholding for my on elderly mother-in-law, confusion about various loose ends, and so on.
The whole process will end up taking about a year. It will be handled professionally, compassionately and efficiently. Still, it takes some time and it costs some money.
People who think that automated systems, regulatory mandates and political directives can ‘fix’ the mortgage delinquency situation are missing the human element. It’s a whole lot easier to automate the process of shovelling money out to people in boom times than it is to work out a fair resolution of to an unpleasant situation in bad times, even assuming the best of intentions, completely honorable conduct and a reasonable, agreed goal or objective.
Sorry.
Well, Marie, I’m sorry to add insult to your injury, but if after 3-4 months of being laid off you can’t afford your house anymore, that means you did not afford your house to begin with.
“Afford” is not the same as being able to pay the monthly sum, it means being able to weather adverse situations/environments and keep paying.
With the economy creating fewer jobs while monthly job loss is around 665,000 according to this article in the ny times http://www.nytimes.com/2009/05/25/business/economy/25foreclose.html?_r=1&hp, there will be a lot of responsible borrowers with good credit and a 6 months to 8 months rainy day savings no longer able to afford their affordable mortgage. That is the reality
The incentives throughout the current system doom it to prolonging a recovery. Originators are hoping asset prices will soon rebound wherein they can minimize their losses. Servicers are trying to ring as much fee income as they can, so they have an incentive to drag their feet and ultimately foreclose.
This is deregulation out of control
Garbage-in, garbage-out was what they taught us for the AppleIIe’s in 1988, and it still applies today.
Unfortunately, loan modifications, on a large scale, equate to far worse terms than the original terms of the original loan. It’s no wonder family’s continue to loose their home’s even after receiving what is supposed to be considered help. Hmmm… come to think of it… it sounds like the definition of Predatory Lending in it’s truest form to me.
No Patty, you are not naive, but not sure about the banks? Even though it is an easy calculation you would be shocked at the way the banks handle different scenarios
More modifications isn’t going to solve the problem because approximately 50% of the loans that get modified will go back in default within 6 months. People can’t afford to pay due to high consumer debt.
Throwing more money at it by pushing loan mods is not the answer.