Loan modifications are becoming increasingly common because of the declining housing environment. Although it still remains uncertain how these modifications will play out, ABS players, specifically investors, need all possible tools to cope with the potential repercussions of changes in loan terms on the securities they are invested in. In this roundtable, we found out how market players are using data analysis to understand the impact of loan modifications on ABS portfolios. The discussion was moderated by ASR editor Karen Sibayan.
ASR: What are the most common modifications we have been seeing?
MACDOWELL: The bulk of the modifications that we are seeing up to this point, except for a handful of exceptions, are rate reduction modifications. You have got one or two servicers that are pursuing debt forgiveness in their modification programs. It boils down to each servicer has a different set of incentives that dictate the way they service, and that underlying motivation reveals itself in the form of who is doing loan mods, what type of mods they are doing, and not just loan mods, but loss mitigation as a whole.
HURST: There has been some rate modification and P&I recapitalization going on as well. That's probably the other type of modification that's gotten the most use. I haven't seen very much of the debt forgiveness being used.
MACAULEY: I think I would second that as well. There are one or two servicers that have made more progress on principal reduction, particularly in the subprime space. But as far as data is concerned, the deep history has been in rate modifications.
HUNTER: Anecdotally, the only company that's actually been very aggressive in forgiving debt has been [Ocwen Financial Corp.]. At least that's what all the press releases they put out say. But everybody has said recapitalization has been probably the biggest piece other than rate mods.
ASR: The increase in loan mods presents challenges to data modeling. How are you facing these challenges?
KRONICK: It creates another variable in what you have to build into your analysis and in doing so as a trader and someone who would be investing money in it, you have to build in a certain yield that you need to get paid out for the risk of some kind of modification coming through and negatively affecting your duration or the principal you are going to get back on the bond. So you are definitely looking at a slower process on a bond-by-bond basis to determine the modifications as well as an increased amount of variables and scenarios that could potentially affect the outcome of your bond.
GREEN: The biggest question I get from customers who use our analytical platform is: How do I identify a modification? I think a lot of people haven't even identified the modifications to start modeling them. There is no uniform way they are reported. You have to calculate them yourself first. Is it an interest rate mod? When was it modified? By how much? Was there capitalized interest and an interest rate mod? Is it just a cap mod? By how much? By what date? Was it modified twice? You have to answer all these questions first, before you can even start modeling it. And then, the big challenge comes as you model it. I would estimate there are more than 400,000 securitized loans have been modified one way or another. Before you can start modeling, you need to identify those 400,000 plus loans and effectively monitor their performance.
CALABRESE: I have had a number of client meetings discussing modifications, and from a modeling perspective, how do you really take these into consideration when you are trying to figure out a return of principal, and a number of my guys, both sophisticated hedge funds that have been invested in the mortgage market for quite some time and new players that feel that they have a new way of looking at the market, none of them have really been able to figure out a coherent or concrete method to use to anticipate the effects of loan modifications. What that's done to them is it has basically prevented them from going down into those sectors in which the modification question really becomes a significant component of the principal return on an evaluation of a security and, therefore, they have remained at the top part of the capital structure primarily in super-senior option ARMs, for example, or senior Alt-A hybrids, things of that nature that have significant credit enhancement versus delinquencies. This is where they feel that, even if there is a modification program put into effect that could be properly modeled, at least they are participating in the market at the beginning and be looking to expand their modeling systems to take into account the modification process as it becomes more widespread.
ASR: The market is now looking at what the most effective way to modify loans is. Can you talk about this and how the market is tailoring modifications according to specific products, such as Jumbos, ARMs, second-lien loans, and fixed rates?
ROSENTHAL: Most people would agree that the principal reduction has been the most effective when or if implemented. It hasn't been implemented enough because we are in a falling housing price environment. The biggest problem with a rate modification or capitalizing the advances in a falling housing market is that the borrower is going to be back for better terms as they become more and more underwater. They have already experienced a prior delinquency and will experience it again. That's why the re-default rates are through the roof on the modifications that are rate-only right, because of the falling housing market. The servicers that have the appropriate net present value calculation are adding value from a cumulative loss perspective to the securitization. The struggle is the expectations are so high for cumulative losses that that marginal benefit to the cum losses is not offsetting the pain of the extension risk associated with the triple-A securities when more modifications are going through or the curtailment of interest on bonds in the case of an Ocwen service deal. When they capitalize the interest, they are actually stealing the excess spread and creating interest shortfalls, and it has created a big to-do in the past on interest shortfalls. Net net, from a loss perspective, these modifications should work. The re-default rates, you can calculate this if you had just a simple net present value of go to full foreclosure versus give the guy a 300-basis-point break. If they re-default at 75%, it's still a positive net present value most likely to the securitization, and the servicers will go ahead and mod. With that math, there could be mass modifications continuing as long as they can. Operationally, they can't handle all the modifications. So these high re-defaults still give you a positive net present value, but not enough to make a difference for us investors to offset the pain of the extension risk associated with modifications and the reduction of coupon that could also cause interest shortfalls depending on where you are in the capital structure.
MACAULEY: On the principal versus interest rate reduction, I will concur to some extent that the loans that are principal reduction that are modified tend to have a higher delinquency rate. What I have seen is you might have 80% of the loans already delinquent out of that pool that's getting modified versus interest rate reduction which is a lot lower. However, they tend to track very closely with each other after the modification. So now with principal reduction, once again, the data is very sparse, particularly for other collateral types like option ARMs and Alt-As. Most of this is coming from the subprime space where we have seen enough modifications. The concept of loan modification should be an iterative process, depending on what works, and that's what the servicers are saying. If they believe that on a loan-by-loan basis, they can figure out exactly what works, then that's basically what drives their modifications in bulk. To get to the specifics about subprime versus option ARMs and Alt-A, there are not as many modifications in the option ARM space, but given the fact that these rates are already very low - they are roughly about 5 to 5 1/4 today and they could go down to as low as 4% in about a year - the concept of rate reduction with option ARMs is probably not as effective as you would get with subprime, which are between 8 to 8 1/4. So reducing the interest rate on a subprime loan would make a very significant impact on the mortgage payments. With other products like option ARMs, I see principal reduction as potentially being part of the strategy for successful modification.
KRONICK: On loan modifications, to put it pretty simplistically, the most effective loan modification is a modification of borrowers who actually want to pay down their mortgage, which when you are talking about a 75% re-default rate and net net, it would be a positive. If you talk about a streamlined process for modifications, I would liken it to a streamlined process for originating loans, whereby you got into an area where you streamlined something and you had everything in a little box, and eventually you came out with a great deal of poor loans and the risk factors and the var rates for those loans were very bad at the end of the day. There have been some servicers out there who own the servicing and have a vested interest in the servicing. They are doing it on a loan-by-loan basis and are actually determining whether borrowers actually want to pay off their mortgage. There is nothing worse than a modified loan that eventually defaults again because you are just delaying the inevitable. If a borrower defaults again, the likelihood is that you are going to have a higher severity on that loan, which will be detrimental not only for the overall housing environment, but also for the structures that these loans are in. There should be more of an effort to modify loans of borrowers, even if they weren't currently delinquent, if they have current income that they can prove and they have a willingness to pay off their mortgage.
HURST: That's a really important point because if you look at the loan modifications, you are trying to find not only a way to create a product in which borrowers have a willingness but something that provides an incentive actually to repay the loan. A lot of the foreclosures that have occurred have been because once you have lost the equity in the home that you own, it's better for you to just walk away, or at least that's what some people are saying. So anything that could restore some equity in the loan in a housing depreciation market that has slowed and possibly stabilized - including a combination of maybe a rate reduction and a principal reduction - is where we need to get to. I have seen some research that said that housing depreciation is actually starting to stabilize. Foreclosures are actually driving those numbers, and they have come down to a certain point to which they haven't gone any lower than the 40%, 50% and 60% that you see on the severity side of the subprime part of the market. We have a low interest rate environment going into this year, we really don't have an option ARM problem until 2010 or 2011 when they start to recast and have to deal with repeatedly taking the low payment. It's actually a low interest rate environment. That's going to be good for that product. They are already at a very high level of not having equity. But what I think is going to happen there is, in 2010 and 2011, if we go back into a higher interest rate environment, that's probably going to be a problem on the option ARM side. What I'm looking for is there a way to combine principal reduction and weighted average coupon reduction and then try to get the individual involved to restore the equity in the house and have an incentive to make the payment.
GREEN: If you are reducing the principal, you are paying interest on a lower principal amount, so you are automatically lowering the interest payment. It's important to note that the overwhelming majority of mods have been on subprime ARMs, and they have been interest rate modifications usually on two-year ARMs and these loans have 8-, 9-, 10-, 11-percent coupons. So the idea of changing your coupon to 5% or 4% relieves a lot of stress on the borrower. In the case of an option ARM, which generally float over the 12-month moving treasury average or CMT index, we know the index rate is around 30-50 basis points, and margins are very low (i.e. 200-300 basis points). Many option ARMs are really paying only 3% to 4% interest. You already have a very low interest rate - so the issue with option ARMs becomes only negative equity, while with subprime borrowers, the issue has been negative equity and a very stressful interest payment. To the extent that borrowers really want to stay in their home, you want to work with them, but a lot of investors feel that borrowers have just been walking away from their home because they can rent more cheaply. If a borrower is employed and has the ability to pay his mortgage, but is 20 points under water in home equity, a 20% reduction in principal may induce him to remain in the home and thus continue to pay his mortgage.
HUNTER: The streamlined modification programs have, in some cases, actually exaggerated or made the problem worse. What's happened is that lenders have gone in and they have offered just refinancing. Especially the [Office of the Comptroller of the Currency (OCC)] or the [Federal Deposit Insurance Corp.(FDIC)]-regulated lenders because their regulator was telling them you have to do something to help these borrowers. So they were doing the streamlined approach, and just out of curiosity, I started following [Washington Mutual]. It's a regulated institution and there was a lot of push early on to modify loans. I found that in my tracking of the index, for instance, WaMu seems to have a higher recidivism rate than anybody else. They are modifying more loans than anybody else and also having a higher re-default rate. But that's probably the influence of being regulated - the regulator just saying do something. So they probably took a streamlined approach. If the borrower has a willingness, then you have an opportunity to work with him. A lot of lenders have just been quicker to react than to do something that is going to be a good long-term fix.
ASR: That's a good segue to the next question. Are servicers modifying the right loans and is there too much government pressure that the right loans are not being modified?
MACDOWELL: I don't think it's necessarily political pressure. I think you have to look at each individual servicer and what their interest is with respect to the deal, as well as with respect to their servicing platform. If you are servicing for third parties, you are really concerned about attracting new assets to the platform - and this requires demonstrating to third parties an ability to proactively address loss mitigation. At the same time, keeping the servicing platform functional as a viable business requires maintaining a cost structure that works within the confines of the existing servicing fee arrangements. In the current environment, financing servicer advances has increased significantly - and the volume of required advances has gone through the roof. In addition to performing duties for the trust, servicers must also determine whether loan modifications, or other loss mitigation techniques, make their platform more viable from a cost perspective. How can they most effectively recapture outstanding advances and pay down the costly financing? Alternatively, there are servicers aggressively using property auctions that result in large-scale liquidations - again, in many cases this has the effect of reducing the longer-term cost of servicing and possibly enhancing the economics of the servicing operation. In contrast, there are those servicers whose affiliates retain an interest in the particular securitization trust - resulting in a materially different incentive structure than the third-party servicers. They may want to backload losses. They may want to look at triggers from a certain perspective to preserve their interests - and loan modification may allow them to manage to a certain trigger threshold for the benefit of their interests. It's not so much political influence as it is organizations looking out for their own interests, striving to remain viable in an effort to retain employees, garner more business and be positioned to survive in the current environment. Observing each individual servicer's interest will provide investors with an indication as to how servicers are likely to approach loan modifications.
ROSENTHAL: Unless the FDIC takes control. Then it's political.
KRONICK: The question also is you feel like you have to get in front. There is a feeling that you have to get in front of what the IndyMac Bank situation was whereby you will eventually get into a situation where there is a mandate that becomes the best solution. So you have to lead your way toward that best solution. It feels like the FDIC is trying to hold the baton and lead the way in loan modifications in general. I'm sure that even though there isn't probably overt political pressure, there probably is some pressure felt that you should get in front of the wave that might be coming.
MACAULEY: From a modeling perspective, one of the challenges of figuring out the benefits of loan mods is the moral hazard issue, where if you target only delinquent loans, you have to factor in the possibility of how many borrowers that are current may choose to go delinquent in order to qualify for a modification. In modeling losses, the current-to-delinquency transition rate is one of the major factors in projecting losses. From that perspective, a loan modification strategy that targets current and delinquent borrowers, which may be based on employment or on DTI, will to some extent alleviate that moral hazard concern and allow investors and strategists to properly project the losses and cash flows of loan modifications.
ROSENTHAL: Are you seeing a lot of modifications in performing loans in your analysis, current loans, are they getting modified down?
MACAULEY: We are seeing that for some servicers. It's servicer specific.
ROSENTHAL: In the pooling servicing agreement, there is a question whether you can modify if a borrower is not in default. If it's a termination of imminent default, then you can modify the loan. But if you could modify - if you gave out more modifications to performing borrowers - you could mitigate the moral hazard issue.
MACDOWELL: I agree with the streamlined approach, or any approach for that matter that adds clarity to the market. Trying to invest and trade the securities based upon a set of cash flow assumptions is a challenging task when loan modifications are introduced. Fundamentally, loan modifications can have a significant impact on the cash flows to the trust. From my perspective, it's one thing to ascribe a re-default assumption to loans that have already been modified; the real challenge from a modeling perspective is identifying the borrower eligibility requirements used by each individual servicer. Right now, I can't tell if there is any consistency in the market with respect to the types of borrowers who are being modified. Some servicers are focused exclusively on delinquent borrowers; others are more proactive and try to anticipate default by modifying "current" loans close to ARM reset dates. So, when I evaluate a bond, I'm basically saying, let's take all the adjustable-rate loans, look at their current interest rate, assume 100% are going to be modified down to 4% and use that as a starting point for our WAC deterioration forecast. Then, we evaluate the servicer, collateral, borrower attributes, vintage and other factors to adjust from there. The baseline is essentially saying everybody is going to be modified, and until there is some level of consistency in the market, in order to protect against the risk of a mass modification program, those are the assumptions that prudent investors are going to use.
KRONICK: That also brings up a question of the effectiveness of a mass modification program in reality versus in theory. It's easy to think that this was a problem in the loan market that was five-plus years in the making, and to think that it's something that can be resolved in a year or two with the mass modification program is a little bit simplistic. To get massive modifications on a pool level, not only do you have to offer out the modification to the right person, but you have to have that person actually respond in a positive way to the servicer that they want to be modified. One of the things that is going to be a big problem with modifications, whether they be mass modifications or not, is getting a borrower who has consistently been avoiding mail from their servicer, anything regarding their mortgage, to open up an envelope or answer a call to someone who is trying to offer them a modification.
ROSENTHAL: I will go back to the moral hazard issue. The nonperforming guys are the ones that got principal reduction that are re-defaulting at 80%, right? They were not performing and then they got a principal reduction and now they are not performing again. Then the house goes down again until at some point you will liquidate or maybe it will stick. But they won't open up the mail because once they go down, they get flooded with solicitations from all sorts of industries. So if they are current, they might respond and maybe it's the refi. In the agency market, we are trying to incent the agency borrower to refinance to a lower payment with lower rates. The equivalent of the non-agency side would be to stimulate the current borrowers to either modify or refinance, so it's going to have to be modification because you don't want to get an appraisal. It would be a mod to go down. Maybe the solution is to do more on the current side versus the nonperforming side.
HUNTER: Servicers say that only 30% of the delinquent customers even contact them. Those are the ones that even go into foreclosure. Forget the modifications. If you can get in touch with only 30% of the people and then you have a smaller portion of them that you are able to offer a mod to, the numbers keep getting smaller and smaller. If you had a customer who is current and you offered him something, he is probably more apt to open the envelope. Some of the guys who are delinquent start getting offers that for $3,000 they will get a modification on their loan. It's like if I had $3,000, I could make my mortgage payments. But these people, they are being solicited by people who just want to make money off of them. And even the servicers are saying it's a difficulty for us. We are able to do it for free, and then you have all these companies out there that are pitching to these people I will help you get a modification but you have to pay me, sort of like a bankruptcy attorney.
HURST: It brings up the issue of do they have enough staff or are they organized enough, given the profitability you are talking about that either doesn't exist or has evaporated. This was a cost that was never really anticipated in this market. For them to go out and hire all the people and put up a call center and make the phone calls to get the modifications done in an organized fashion, I'm not sure that's really going to happen. At least from the people that are currently servicing, I don't think the money is available to make that kind of effort without some assistance.
CALABRESE: From an investor perspective, it's a unique time that we are in right now because pretty much all the securities that we are talking about are being run to worst-case scenarios, and any kind of anticipated benefit from loan mods is not being priced into the equation. That's kind of like a free option that you are getting by participating in the market at this time. But to cross over into the servicer and hence the loan modification sector, there has been a number of recent pretty large actions by some pretty large investors on the servicer side. Further back, Wilbur Ross taking down the American Home Servicing platform. Then you had the Fortress/Quicken Loans partnership. And then, most recently, the IndyMac acquisition. You can definitely see that there are some major players in the space. I don't know if they are taking a bet on it, but they are positioning themselves to take advantage of it. On a smaller-scale basis, I have a very close friend of mine who had a loan lead company that sold packaged leads to mortgage lenders, and what he is doing in the great spirit of the U.S. - the industrial spirit and the creative spirit - is converting his loan introduction company and reversing that process into a loan modification company, and in doing so, accessing those same brokers that wrote those loans to the people. Particularly in South Florida, where I just was checking out the current state of the market on a street-level basis, a lot of these lenders, particularly in the subprime space, the broker involved was dealing with their family members and relatives and what-not. That could be the first step in getting people to pick up the phone and actually trying to reverse course on a lot of these mortgages - through people who were introduced through relatives or relationships or what-not. What's happened recently in these large acquisitions of these platforms is anticipation that the U.S. mortgage borrower incented to pay off his mortgage and there are very few people who are just willing to throw back the keys.
ASR: Investors are envisioning a worst-case scenario and are pricing it in. How much pain are investors feeling because of loan mods?
KRONICK: I don't think they feel much pain right now from loan mods because they are not as widespread. Non-servicing advances are probably the most pain you are seeing in terms of investors. If you were to ask most investors, they would probably say that a massive loan modification would not necessarily be beneficial to what they are analyzing and the way that they are analyzing bonds and the yield profiles that they are looking at for bonds. Right now most investors are running the worst-case scenarios, but potentially the worst of the worst-case scenarios could be a successful loan modification program due to just lowering coupon and extension of duration.
ROSENTHAL: In a default scenario, you recover some principal. In a mass modification scenario, you recover no principal until way out, considering you are modeling them into a new 40-year loan. Who knows when you will get principal back?
KRONICK: Effectively, a successful modification will be a loan whereby someone will never probably, in their lifetime, want to refinance unless they move. So the duration of that loan is going to be extremely long. It's like a 3% IO rate increasing over years, and that is a zero principal paying loan or very small principal paying loan. This is versus a defaulted loan where you get anywhere from 30% to 50% - you get some principal back. For people who are investing in P&I bonds, which is the majority of the market right now, that is the better way.
ROSENTHAL: The biggest example of a bond that would be hurt the most through the modifications is what's happened to front pays. Front pay subprime is extended out way beyond anyone ever thought. It traded in the 90s. It's now in the 80s, 70s, 60s, depending on the bond. So whether we are pricing it in now more than we were six months ago, there has been pain, but not from actually seeing it happen.
MACDOWELL: Everybody, due to uncertainty with respect to loan modifications as well as the implications surrounding proposed policy efforts, will continue pricing in the worst-case scenario. If the market is able to achieve some level of consistency in terms of what loans are being modified, how the process for modifying a loan takes place, how loans are underwritten, and focus on modifications that provide the borrower with sustainable monthly payments - that would aid in the stabilization of the securities market. Additionally, to come full circle in the feedback loop, receiving access to that information in a format that investors and researchers can analyze would be extremely useful. I recently read a study by Alan White that reported only 54% of the modified loans in the study's sample reduced the borrower's monthly payment, and 23% resulted in payment increases. Now, payment increases are generally going to fall out of my algorithms that attempt to identify modifications, because when I go through 1010data, part of the criteria for identifying a modification is that the interest rate decreases by some amount, which is obviously going to reduce the monthly payment. That's somewhat offset by any arrearages that are capitalized. But if it does not become clear that borrowers are receiving sustainable modifications processed in a consistent manner, the market is going to continue to price securities to the worst-case scenarios.
GREEN: Some investors are just buying the interest. They don't expect any principal. To the extent that you extend their interest payments, it's a win. There are other investors that are investing in principal and interest, and to the extent that you extend somebody else's interest payments but reduce their principal, they are going to be less apt to buy that security. They are going to demand a lower price. Investors may not bid 40 cents, but might bid 30 cents because of a deferred write down. Investors need more clarity in terms of what to expect. They don't really have much historical loan modification data. You have to roll your sleeves up and look at them yourself, and see what different servicers are doing, and ask questions such as: Should I be reducing my WAC by two points? Historically is this what servicers are doing? Should I reduce it by 4%? When I price my bond on Intex, should I use a 2% coupon on all modified loans? Which loans should I project to be modified? So investors have all these considerations that make it very challenging and, to some extent, reduce liquidity of the overall market.
HUNTER: The B-piece investors have made out in here because of the modifications, especially the ones that are capitalized.
MACDOWELL: Except for Libor.
HUNTER: Yes. Except for Libor. But they haven't gotten their write-off. We noticed last month in the ABX Index that there were some massive modifications done. Like GMAC last month put through tons of modifications. As a result, their prepayment rate slowed to nothing. They had negative amortization other than liquidations. We went back and found out it was because they did so much in terms of modifications. It slowed down the prepayments and cash flow to the top layer of the waterfall, and that's a problem. But the good news is everybody has said, 'Guess what, if that's going to happen, I'm going to price this to the worst-case.'
MACAULEY: Before we leave this, I want to point out that there are other streamlined modification programs that have been discussed and implemented by the government, like the HOPE For Homeowners program, which adds another twist to this discussion. In that case, the modified loan leaves the pool as it's refinanced into an [Federal Housing Administration] program.
ROSENTHAL: A home run.
KRONICK: To the taxpayer, no.
MACAULEY: It hasn't caught on as much for other reasons, but I just want to point out that it's another twist that needs to be considered.
HUNTER: Actually, I agree with that. The HOPE For Homeowners program would be great. Except that's almost one of the biggest problems. The investors, especially banks that own the asset as opposed to the securities, or some security holders, they would love to see that because, 'I bought the bond at 40. I will give you a 30-point discount.' But some of the banks, they are unwilling to take the hit upfront because they are afraid someone is going to make something on them. I think the HOPE For Homeowners program hasn't been as successful because no one wants to take the hit on their books. They may have reserves against it. They may already have reserved and it may already have been priced in, but they are thinking that these are money-good loans eventually and they don't want to have that money-good loan go away.
ROSENTHAL: Or they don't want to write off their second lien to give the hope to the homeowner.
MACAULEY: My point is that from a modeling perspective, looking at the different choices of loan modifications out there, part of that has to be what if the Hope For Homeowners program ends up being more favored. It isn't so far, but there are discussions out there of how to make it more effective as well.
HUNTER: I agree. I think that would be a good use of the program, except that lenders don't want to take the hit because they don't want to run it through their balance sheet.
ASR: Is there data that could actually help you guys predict if a loan modification is going to be successful or not?
MACAULEY: Well, debt-to-income ratios, which are currently being used, could be more dynamic.
KRONICK: A borrower's current income would probably be really nice to have as well as credit card debt. The majority of the loans that we are talking about here were low-doc or no-doc loans. So, to the extent you can get much more documentation, getting housing appraisals is fairly easy. Getting the actual debt load on the borrower outside of his home mortgage payment is the more difficult thing to get and is probably the most valuable thing that you can have.
ASR: Is there not a single place where you could get updated credit information and debt-to-income?
KRONICK: Legally, probably not.
GREEN: Currently we have customers who get updated credit information about borrowers and us an advance statistical matching technique to match the borrower to the loan. The capability exists today to for investors to statistically obtain this information without violating any laws by focusing the match on 100% non-identifiable information. This is a statistical match and thus investors cannot be 100% sure that they have made the correct match between borrower and loan, but customers have said that access to this type of information helps them more accurately predict a borrower's propensity to default.
HURST: There is also a lot of data out there, at least one or two services that will provide you with the ability to search the property files in order to match the first, the seconds, the thirds, and the fourths. But that also - at least from my perspective - would be very expensive. I do know that people are accessing it and coming up with a combined loan-to-value on a mortgage property, but I don't want to have to pay $10,000 for each mortgage that I do that research on.
GREEN: I think there is a critical mass of capital under management: If investors are able to buy a bond for 20 cents that they know is worth 40 cents because they have updated credit information, then the investment in the information pays for itself many times over. So it depends on a lot of factors, but a lot of our customers have been subscribing to these type of services, and so far they say they have gotten pretty good results. The feedback has been extremely positive.
ASR: There are different ways that a loan mod affects investors in the different parts of the capital structure. Can you elaborate on that point?
ROSENTHAL: We can do subprime because that's where most of the modifications are. We could talk about option ARMs that everyone is just pricing in there; just 40 is not cheap enough.
CALABRESE: It's 45 now.
ROSENTHAL: If you just look at the triple-A stack in subprime, typically they were tranched front pay, second, third, fourth and last cash flow. For the first and the second, it's clearly negative if there are more modifications because it will extend those cash flows and you thought they were going to come in at a shorter duration and they come in longer. The last cash flow, which is the one that is the most tricky in the stack because if the loss mitigation actually does reduce the losses to that marginal benefit where your last cash flow bond might actually be money good, let's assume that - and we all are - it's going to get whacked by losses. Well, we actually like when it gets hit by losses because its principal becomes pro-rata if you buy the right bond. So you don't want the reduction in loss of that bond that you expected to get principal on, because it was going to get hit and you bought it at 30 cents. Modifications could actually hurt that bond because the losses could come in lower. You could extend. Now you are not a 10-year bond, you are a 15- or a 20-year bond. I'm having a hard time finding a bond in the capital structure where they are pricing today where modifications areactually a really good thing. I think that maybe modifications could help the M1, M2, M3 stack in the double-A because you can buy those for three to six cents on the dollar, and maybe the loss mitigation will actually save your bond and you might get principal, but since you bought it at three to six cents on a dollar, that's a home run. Maybe there is a positive there. In credit IOs, it's bad because the principal reduction is coming and it's just going to blow through the credit IO bond faster than you thought with the principal reduction on mods. It will accelerate the losses absent the extension. So you have to find maybe the M4 at two or one, but this isn't that exciting for us. Bonds trading at one, two, three cents on the dollar, it's not that exciting. We focus on some of these intricacies and I'm interested in other feedback.
KRONICK: I would agree with you on the subprime. The subprime space will be very interesting from a structural perspective with loss mitigation. Actually, there is some expectation of bonds, the triple-As going pro-rata more quickly than expected. You see a lot of that getting priced into the late '06 and early '07 like front-pay bonds, for example. In the earlier vintage subprime credit, it's another interesting aspect where you have '04 and '05 stuff that might be failing triggers. If you have modifications, it won't hit your losses if there is principal reduction, but it will reduce your trigger possibly to the extent where you might be actually getting cash flow when it passes the triggers. It's probably a bond-by-bond specific thing. In option ARMs, loan modifications, generally speaking, will extend the duration and will reduce the principal that you are expecting that people are projecting on a month-by-month basis right now. If you are running losses through the deal in a no-mod scenario, you are expecting a certain amount of a chunk of cash in the next six to twelve months that you are not going to see now. And you are going to be given kind of an IOU at the end and them saying, 'Don't worry, these guys are better now than they were before.' Even though you have lower credit enhancement, this is eventually now going to be a money-good bond as they say. So the question with the pricing of those assets will be will there be a tightening of the risk-free rate that you need on these assets enough to mitigate the duration that you are taking on? One thing that investors love the most is certainty. And if you increase certainty on your bonds, technically you should get a tightening of rate on those bonds. The question remains if you do modify option ARMs, if you look at all these passthroughs, you are talking about Alt-A and fixed-rate or an ARM. If there is a greater certainty for you getting your assets, even though a longer duration, will your net net be higher in price or lower in price? You ask 10 people and they will give you 10 different answers.
ROSENTHAL: It depends on the liquidity of your investors.
KRONICK: In credit IOs, it also depends on the price you pay. It's a matter of how efficient you think modifications are going to be. If you think that they will be very efficient, you will probably lose money. If you think that they won't be very efficient, you will probably make money. It's not very sexy but you could make a little money there.
MACDOWELL: When you see loan mods coming through, are you seeing more servicers require trial periods, or are you seeing the majority of servicers report the loan as "current" upon modification? In other words, are you seeing more servicers require some level of performance under the modified terms for a period of time, or are more borrowers simply being reported as "current" upon modification - and to what extent have you seen that impacting triggers?
KRONICK: What I have seen in remittance reports is that once they make their payment, they are deemed current until they start not paying anymore. I don't think modifications have occurred in such a massive way that you can actually affect triggers significantly right now.
GREEN: The original older vintage bonds that were performing better have certain delinquency triggers that redirect cash flow to the bottom of the capital structure. To the extent that delinquencies hit a certain threshold, it alters the cash flow to the deal. This is not an issue with respect to the '06 or '07 vintages where the delinquencies are extremely high, but for some '04 and '05 vintage bonds, it is an issue. Have you seen any cases where a mod has altered delinquencies?
KRONICK: Not yet, but we haven't seen a mass modification environment yet.
ROSENTHAL: We are seeing more impact on REO liquidations affecting delinquency triggers more than the mods affecting a specific trust. You get this massive CDR prints - we are seeing 40s in a lot of deals and that cleans out a lot of delinquencies and affects the trigger question. The mods will affect it, but they are re-defaulting so fast that you are getting them back into the math.
KRONICK: That's if you believe they are going to re-default so fast.
HUNTER: Well, the question we don't know is what's next once they re-default; is there an additional severity, and because the market has continued to deteriorate, it will have additional severity.
ROSENTHAL: You might get cash in the interim to offset that incrementally.
GREEN: Just because they re-default, I think it might still be a good thing from a cash flow perspective. For example, if the severity goes up 1% but the investor gets six extra coupon payments, the modification still might be a good thing for the overall deal. But you need good analytics to figure it all out.
ASR: What about pooling and servicing agreements (PSAs)? Do they give certainty because they restrict the number of modifications that can be done?
CALABRESE: It's impossible or it's very difficult to get the PSA on most deals. In a number of deals that I have worked on, I have had clients that have said, well, this XYZ servicer is actually redirecting cash flows separate from what's actually put down in writing in the PSA. So I don't think that investors are really protected by anything out there. There is so much due-diligence that you need to do. Some deals you have to take with a grain of salt. But the bottom line here is I have found that if you are able to access the PSA, which has been almost as hard as getting a deal portfolio on a CDO, the lack of transparency makes it very difficult to use that in your favor.
KRONICK: Personally, I think that the housing issue is so big in the market right now, not just the mortgage market, but the overall economy, that you are fighting an uphill battle if you are going to start arguing as an investor to the trust if the PSA gets broken.
CALABRESE: Also trying to trace back to the past problems that we have had in mortgage-related sectors and trying to use manufactured housing as a proxy from that whole experience and the investor community that rallied around that whole issue, one of the most telling things about that whole process was that the servicer, which was really one large servicer, was able to negotiate to get a much larger servicing fee that was put at the top of the waterfall and the cash flow. That actually incentivized the servicer to get out there and knock on doors and really make a face-to-face type of appraisal as to what the current state of the borrower was in. This is a factor of 100, the difference in size between manufactured housing and where we are in all the scattered resi credit sectors.
ROSENTHAL: It would be interesting if they all of a sudden increased the servicing fee across all the deals. That would certainly make up for the fact that it's not economic to service right now. But that sector got saved also as Libor went down and got lifted with the whole housing sector. So we are in unchartered territories. My experience has always been that servicers have more leeway than most people think and that they are able to do a lot of things like modifying current loans by simply saying the borrower is going to default. And how do you prove that?
HUNTER: The housing act of this summer actually gave servicers a carte blanche to do whatever is necessary for the borrower's best interests. So, okay, there is a 5% limit on modifications. But then, of course, you have contracts law. So you have that issue where you have a Congressional act that says servicers can do what's best. But then you have issues like with the rating agencies. How are they going to treat it? I talked with people at GMAC last week and GMAC did some massive modifications last month and they said it was because they got a ruling from S&P that it wouldn't affect the rating. So they put through a whole lot of modifications. But then you have someone like Wells Fargo and [Litton Loan Servicing], which is now reporting future losses and they are passing through those. I don't know what part of the servicing agreement that falls under. But they are now passing through future losses when they do a modification. Servicers have an awful lot of leeway, and I think they are taking advantage of it.
ASR: What do you think is going to happen in 2009 - where will the opportunities be?
MACDOWELL: From my perspective, as an investor in the securities, it would be helpful on the data side to get more consistent reporting so we could actually analyze the impact of loan modification in a more efficient manner - which goes hand in hand with transparency and consistency. Right now, there just seems to be such a divergence between servicers in how they are approaching this aspect of loss mitigation that it's very difficult, as an investor, to figure out what is going to be the next surprise out there. However, this also presents opportunities for those who have the capability to do the work, maintain an active dialogue with servicers and identify trends at the servicer level. So just being able to get some of this information in a useable format would be helpful.
ROSENTHAL: As an investor, we do have a bottoms-up approach, and when we think about modeling, we take it very seriously. When you look at the modifications, I look at it two ways. One is who has already been modified, so that you are not fooled by the fact that the borrowers are current when they are really not current - they are modified current borrowers - versus trying to predict who is going to get modified in the future. Predicting the re-default rates on the guys you can identify is something that is, right now, unachievable. Since the bonds are cheap enough, we can just assume 80% re-default and buy the bonds at our yields. To try to predict, we used some of the similar approaches that were talked about - what's the WAC going to be and how much principal forgiveness or how much yield do I really need to get to take this risk, which is ultimately the answer to that question because all these things are from the bottoms-up approach still not achievable in the current market conditions. You have too much uncertainty about the behavior of modifications, about how many modifications will be done. It's something we obviously have to keep an eye on, and hopefully, we continue to find relative value based on things like seasoned bonds versus current bonds, where in the capital structure do you want to be, are you pricing the modification risk appropriately?
HUNTER: The biggest problem is the fact that everybody thinks that something is going to be done, so you have to look at both sides, the borrower versus the investor. From the investor's point of view, we modeled everything to the worst case, 95% of all the loans are going to default, all right, with 80% severities. The borrowers are out there saying where is my piece of the action? So that's actually creating the worst-case scenarios: We need to get some kind of floor in there. There have been examples of where a floor has been established, like in California, where property prices have dropped 40% and 50%. All of a sudden, people are coming back into the market and it's doing better, relatively speaking. In other parts of the country, you have to establish a floor and stop having people think that they are going to get something - whether it be this is what everybody is going to do or this what we have to do as opposed to everybody is going to get something. That is creating a bigger problem because that's creating more and more defaults and the moral hazard issues and all that kind of stuff.
CALABRESE: From an investor perspective, the main reason we are all here today is to try to figure out how loan modifications are going to affect the dollar prices on securities. Until we can get to a more conclusive answer to that question, the investor community is going to stay in the top part of the capital structure, which is relatively insulated to a certain degree. You can't paint the market with a very broad brush. Everything is a bond-per-bond type of analysis. To make a long story short, we are trying to figure out where dollar prices are going. If you look at the market, just in the month of December, we have seen a pretty significant uptick in dollar prices at the top part of the capital structure, and that's been driven by the announcement by the Fed and also the asset managers that were selected to participate in TARP and TALF and all these iterations. The interesting trait for a number of our investors at Guggenheim was the capacity. As bonds were being sold in massive liquidation lists and dollar prices began to get very dislocated, you were able to purchase super-senior securities at dollar prices below what the credit enhancement was underneath those securities, and there was added a slight discount to that. So there seemed to be a level of comfort at that entry point and since that time we have seen dollar prices over the last month increase by five to ten points in those types of securities: Alt-A, option ARM, prime, hybrid securities. And that equates to a 10% to 20% increase in the value of those securities given where the dollar prices are trading, in the $40 to $50 price range. I don't know of any other asset class that's had that kind of performance in the month of December, aside from maybe the stock market. The problem with what's happening in dollar prices is the fact that as these prices go up, it begins to take that level of security out of the yield profile that you are modeling for these securities. The reason why these securities at the top of the capital structure are trading in the face of these concerns about loan modifications and what-not is because they really don't have that much of an impact. As investors start to put their chess pieces on the board for 2009, they are starting with the safest part of the capital structure because you really don't have to go down that far in the capital structure to get pretty significant double-digit yields. It's kind of a waiting game at the beginning. Let the market react to what processes develop in the loan modification space, and then get some kind of real data to actually use as a means to forecast how these loan modifications are going to affect securities further down the capital structure. In summation, aside from any sizeable liquidations or other major blowups in the hedge fund community or the like, the current dollar price appreciation that we have seen for those types of securities will probably continue going into 2009. As that yield compression continues, it will force investors to begin to look further down the capital structure, and as such, use services like 1010data to kind of create a more granular type of approach to taking loan modifications into the equation from a modeling perspective.
GREEN: You can see from this panel all the analytical challenges that investors - and these are some of the most sophisticated investors out there - are thinking about. It's hard for them to get a handle on this. What we try to do is make sure that people are able to get at the most granular level of detail, the cleanest information, and the most accurate data because in this market, you can't look at data secondhand if you plan on owning bonds. You have to roll up your sleeves and do your own analysis. People look at roll rates: Are loans rolling from current to current or are they rolling to current and modified and which particular date? We are just trying to stay ahead of the game and make sure that our customers are able to effectively analyze loan modifications; so our customers see every loan mod, how loans were modified, by which servicer, and the date of the modification. As the market evolves, we are just trying to stay ahead of it and ensure that our customers are equipped to come up with an answer. It's still going to be difficult, but they are more equipped to really bid these securities.
KRONICK: For '09, what I'm looking for and hoping for is just some more clarity in the market. Some more certainty, especially with the loan modification issue. Once we get that kind of certainty - how fast they are happening, how successful they are, what kind of nature they are taking - these will increase the certainty, the investor base and the money that used to be in the sector will come back. I think that when everything is pricing at worst-case scenarios, the yields are attractive to long-term investors. As certainty comes back into the sector, so will the money.
MACAULEY: The entire loan modification process is still evolving. We talked about a lot of different modifications here. There are still some that have been proposed but haven't gone anywhere. For example, there is one where the FDIC/government might guarantee a certain percentage of the re-default risk and basically provide some insurance to servicers that decide to pursue modification, and that insurance would come in as far as re-defaulting is concerned. From an investor's perspective, the goal should be to look at, do some kind of a scenario analysis and be aware of all the different options that are out there. There is data out there to do some analytics and analysis of loan mod. Yet it is an evolving landscape, and there could be some new loan modification types that may very well change most of your assumptions. It's necessary to be aware of all the potential modifications that have been discussed, even though we may have not really seen any volume in them yet.
HURST: For '09, we will not see a complete evaporation of the uncertainty that exists around loan modifications, but just more certainty about what unemployment is going to be, what's going to happen to the car companies, what's going to happen to our economy and where we are going to go and how much government spending is going to be there to provide jobs to build up the infrastructure and how many people. As we get into this year and we see what happens to the programs that are put in place, it will be more certain what the cash flows are on the senior bonds and in the credit IO space, the A-B kind of bonds. I have to think back about when we first started looking at the mortgages and realized something was wrong with the 2006 and 2007 vintages. Then there was no trading in these bonds. They went to super-seniors. And then there wasn't a market for a while. The market came back. It came back at the top and it came back at the bottom. As we got closer to Lehman Brothers or TARP or whatever your point of division might be, that's when the market went away again. But at that point, we also had a market in the middle. We had 30- and 40-dollar bonds, and people were taking a view based on what they knew and what they thought was actually going to happen with this mortgage product that was still a little bit uncertain about what's going to happen in 2009. What we are seeing right now is a credit IO space pricing to the worst. As we get more information and more certainty about what's going to happen to our economy this year and we know what unemployment is going to be, I would think that the market will start to come back. We will have a market at the bottom. We will have people that will be making an assessment of the place to be: in the mezzanine or in the last pay or do I go back to the super-seniors because I know what's going to happen. So I would expect to see the market liquidity improve over this year.
ROUNDTABLE PARTICIPANTS
John C. Calabrese
Managing Director
Guggenheim Capital Markets
Jonah Green
Director of Mortgage Analytics
1010data
Robert Hunter
Vice President
Amherst Securities Group
Rusty Hurst
First Vice President
Stifel Nicolaus
Stuart Kronick
Managing Director
Jefferies & Co.
Desmond Macauley
Managing Director
RBS Greenwich Capital
Jack MacDowell
Portfolio Manager
Old Hill Partners
Marc Rosenthal
Portfolio Manager
FrontPoint Management
(c) 2009 Asset Securitization Report and SourceMedia, Inc. All Rights Reserved.
http://www.structuredfinancenews.com http://www.sourcemedia.com/
