Same payment, new lender - someone's making money, right?

August 18, 2007 3:06 PM

A matter of curiosity: What's a mortgage worth?

Most people who have a mortgage have at some time or another had their mortgage sold to another servicer. We've had ours sold a couple of times. Here in Tucson, a major mortgage company just collapsed; their business model was to sell every mortgage they originated, but with the current problems in the home loan industry, suddenly they couldn't find buyers any longer. (And they weren't heavy on subprime loans either.)

I'm not really knowledgeable about how these things go, but since it's all over the news now, I'm wondering how it all works. So what I'm wondering is - does a mortgage get sold for the current balance remaining, for more money, or maybe for less, since the interest is the thing that makes the money on it? Or is it something different altogether?

posted by azpenguin to work & money (11 comments total)

Not quite a direct answer to your question, but if you are further interested in the mortgage industry and what is done with the mortgages that are purchased, check out the Calculated Risk blog. Tons of discussion there, and the UberNerd posts go into astounding detail about it.
posted by procrastination at 3:16 PM on August 18, 2007

The buyer is buying the future stream of payments from you. So the price the buyer is willing to pay depends on the interest rate they want to get at the time of the transaction, not the interest rate when it was written. In other words, the value of the mortgage will rise and fall with the prevailing interest rates the same way that bond prices rise and fall -- higher interest rates mean the value falls, lower rates means it rises. So the value can be calculated if you know the remaining principal on the mortgage, the time left on the mortgage, and the interest rate the buyer wishes to earn. The current crisis comes from the fact that buyers of mortgages (packaged as securities), have not particularly paid attention to the credit-worthiness of the underlying mortgage, or the actual value of the underlying real estate. You can bet that going forward, these factors will also play into the value of a mortgage being sold.
posted by beagle at 3:23 PM on August 18, 2007

The technical or jargon-laden way of phrasing this question is: what is the present value (sometimes called Present Discounted Value) of my mortgage?

It's an interesting question. It's also worth noting that the simple equations in that Wikipedia article do not take into account things like price inflation and risk of default. But if you really want to know the present value of your mortgage - to give an honest answer to your question - those things have to be taken into account.

As far as I can tell, this question seems to be at the heart of this current subprime liquidity crisis. When large numbers of mortgages are packaged (securitized) together and sold en bloc, the individual mortgages have not been carefully scrutinized and valued. The reason that the credit market has ground to a halt is that there is a great deal of uncertainty about the correct valuation of the individual mortgages in these mortgage-backed securities. As one financier put it in an article I read recently, "There's a lot of unglamorous roll-up-your-sleeves type of work to be done valuing these mortgages."

As near as I can tell, to get this market in mortgage-backed securities going again, someone's going to have to, for example, go through a package of 10,000 mortgages and try to calculate their present value; the hard part is going to be estimating the risk of default; and this is going to have to be done for all these packages of mortgages before they can be traded again.
posted by ikkyu2 at 3:49 PM on August 18, 2007 [1 favorite]

There is a difference between the servicer of the mortgage and the owner of the mortgage. The actual ownership of your mortgage is pretty opaque (to you at least) and is probably actually owned by more then one person having been sliced and diced into different tranches of mortgage backed securities. You wouldn't know when/if these have changed hands.


The "thing" that has been sold along several times in your case is the right to service the mortgage- i.e. collect your payments. I'm not an expert on that business but my understanding is that the servicer is paid a very small fee (like a few hundredths of a % of your loan), so when a servicer buys your loan from another servicer he is valuing the service rights based on 1) his cost of taking in your payment and sending it out with a bunch of other payments to the structure created to securitize your loan 2) the likelihood that you will pre-pay your mortgage 3) the fee he earns for his service.
posted by JPD at 3:52 PM on August 18, 2007

many great minds are at this moment trying to really figure out what mortgages are worth!
posted by Salvatorparadise at 5:01 PM on August 18, 2007

Between ikkyu2 and JPD, your question is answered.

So what I'm wondering is - does a mortgage get sold for the current balance remaining, for more money, or maybe for less, since the interest is the thing that makes the money on it? Or is it something different altogether?

Your mortgage is never sold for the current balance remaining. That simply wouldn't make sense, as every time you make a payment toward the mortgage, the investor who "owns your cash flow" would likely reinvest that money at a similar rate (or risk level)...That's what ikkyu2 is referring to when he says present value. A long term investment like buying a mortgage (from the perspective of the bank) is only good if you can reinvest cash flows at a similar rate (or risk level)...otherwise, you'd pay much less for the investment or buy else. Just for the sake of an example, if you're interested, a treasury bill or zero coupon bond are examples of financial securities with no reinvestment risk (i.e. the interest you make on it is always rolled into itself and compounded until it finally pays you the final principal amount; thus, you never have to worry about reinvesting the money before you finally get the final payment)

Prime mortgages, jumbo primes, and supposedly really good quality mortgages became very popular in the early 90s and servicing those mortgages (i.e. collecting your payments toward your mortgage, managing your account, making sure that it goes to the proper investor, and harassing the shit out of you if you miss a couple of payments, or foreclosing on you if you're REALLY late) was a "right" that got sold to companies for a fee. Now, as more people entered the field in the 90s, the fee got smaller and smaller, until the very business became commoditized (i.e. razor thin profit margins).

So the new "in thing" was to originate and service subprime mortgages. Back in the day, banks that originated the subprime mortgage usually held the mortgage, because even if they could securitize it, Fannie Mae and Freddie Mac weren't allowed to buy it from them. But that all changed with the advent of CDOs (and related structured products) and the boom of hedge funds. Suddenly really rich private investors could use hedge funds to buy these "subprime mortgages", which were wrapped up in really complicated financial securitization and given AAA ratings by Moody's and S&P (who were in on the scheme, BTW) and pay servicers the fees necessary to collect from the home owner.

Servicers are complicated entities. They buy the right to service a large pool of mortgages (usually 0.75%-1.25% of total mortgage balance in the pool), and their price usually involves a number of factors, including prepayment speed of homeowners, default rates, future interest rate fluctuations, and of course the credit worthiness of the home owners. This is a ridiculously complicated math question and needs to be made in an extremely competitive environment (because Wells Fargo, Countrywide, and many other big banks service their own mortgages as well as purchasing rights of others).

I could probably go on for pages; sorry if I went astray. If you have any more specific questions, please feel free to ask.
posted by SeizeTheDay at 5:26 PM on August 18, 2007 [1 favorite]

I did a really crappy job of explaining how your mortgage is owned by a bunch of different people.

Basically the way it works is I buy a bunch of mortgages from the originator (in this case for example the Phoenix homelender you mentioned) I can then shove all of them into one company who's only asset is those loans. In order for that company to buy the loans it issues bonds - who are legally entitled to all of the cashflow related to those mortgages. The bonds are called Mortgage Backed Securites. People started figuring out how to do this in early 80's with a big pile of government subsidized GNMA loans. In that case you knew all the loans had certain loan to values, certain underlying credit scores, and you had a great history of default rates so you could value them just like on a single mortgage. Actually the bet was you would be willing to pay a tiny bit more (well - accept a slightly lower yield - but that is just a better way of saying the same thing) because you also now had the benefit of diversification. if someone defaulted on their loan then all of the securites suffered the same amount of capital loss. As a result the bonds themselves were as creditworthy as the underlying mortgages, and because these were vanilla mortgages the market had a pretty good sense of the creditworthiness.

What has gone on in the current craziness is that someone figured out that if certain people were only willing to buy debt securities that were considered low-risk (as judged by the rating agencies), there was money to be made turning parts of pools of risky assets into lower risk assets. The way they did this was by tweaking how those Mortgage Backed Securities worked. Instead of all the MBS' taking part equally in losses, they "tranched" them - for example we will create four different types of MBS. Type one gets paid first, this second, this third, and these guys get anything left over. The net effect of this meant that in order for slice one to suffer any capital loss slices 2-4 had to be worth zero. That meant that S&P and Moody's could be convinced Slice 1 was AAA rated - meaning the lowest credit risk possible other then treasuries. Which opened up an entirely new pool of investors for assets that had previously been seen as not safe enough. So if you a relatively new mortgage when you send in your payment pieces of it go to different piles (depending on how the other people in the structure are doing when it comes to paying their mortgages) - which is what I mean when I said your mortgage is actually split up between a bunch of people.

So what went on in all of this craziness that we now seem to be suffering the fallout from. Someone has to figure out how big each one of the slices needs to be to make sure the slices above it in the structure are as secure as advertised (usually as judged by the credit rating agencies - S&P, Moody's and Fitch). The way this gets done is lots of lots of modeling of default rates and prepayment rates of lots and lots of different mortgages with different characteristics. Lots and lots of assumptions of normal curves and stuff like that. That is supposed to tell the guy putting the mortgages in slices how big each slice needs to be. These models are also used to convince the rating agencies how big each pile should be.

All well and good - but this is where we start to run into all sorts of conflicts of interest. The guys who create these tranched MBS' are nearly always investment bankers who will get paid a fee based on how many of these deals they are able to complete - they usually do not get paid based on how close to forecast the securities they created perform. I.E if a AAA rated tranche ends up being worthless in 5 years no one goes back to the guys whom made up the tranche and asks for their money back. The best way to ensure you get lots and lots of these transactions is by offering the originating banks the highest price for their loans - or the lowest implied yield on a dollar of mortgage value. The originator wants this because because the lower the yield he can sell his mortgages for, the lower the yield he can offer borrowers - which means his origination volumes will go up. The originator makes his money on fees related to the loan origination so he just cares about volume. Also a lot of the bigger originators are also loan servicers so they guarantee themselves the earnings stream related to each origination.

So how do the investment banks offer the highest price/lowest yield - by maximizing the size of the higher rated tranches and minimizing the size of the lower tranches - because the price the are able to pay is the weighted average of the prices paid for each of the tranches, and the lower rated the tranche the lower the price/higher the yield paid. So it is in the bankers interest to convince the rating agencies (and everyone else) that the degree of collateralization of each successive tranche can be smaller and smaller.

What we've had happen in the last five years or so is that the bankers have used these models to create lots and lots of securities of all different types of mortgages, with the only difference being the size of each of the tranches being a function of the creditworthiness of the original loans. You can make X% of a pool AAA if they are FHA conforming, but only .5X% if they are mortgages with LTV's greater the 95% and FICO scores under 650. The problem with all this is that it is in the banker's interest to push the models to come up with the most competitive bid possible so they can earn their origination fee. The harder they pushed their models the better the price they could offer lenders, the better the price offered lenders, the more aggressive lenders could be in pursuing loans from people who previously would have been judged unacceptable credit risks. This meant that once interest rates went up and the weaker borrowers came under pressure all of these models were tested, and many of them failed. If you owned a a second or third tranche you thought had only a 1/50 chance of defaulting every year, you are know realizing it has a 1/25 chance of defaulting every year. Since most of the people owning this security own it on leverage the actual loss is much greater then implied by the decreased value of the asset.

On the flip side of all this there was also a massive trend towards decrease risk aversion, which meant not only were people willing to believe what the rating agencies said about the different tranches of the MBS', they were also willing to pay a higher price/accept a lower premium to treasuries then they would have historically for an asset with the same perceived creditworthiness. Which only exacerbated the mistakes made in the models.

I'm sure someone is going to come along and tell me I butchered this. There are all other sorts of complexities I sort of glossed over as well. I also don't think the MBS origination side is a free market-y as I made it sound, and I'm sure I made lots and lots of other mistakes so I apologize in advance for any necessary corrections that need to be made. Also it isn't just the bankers fault. The people who were willing to pay these prices for these securities are equally at fault.
posted by JPD at 5:27 PM on August 18, 2007 [2 favorites]

As far as I can tell, this question seems to be at the heart of this current subprime liquidity crisis.

You're partly correct. The liquidity crisis has had a number of factors that have caused it, including, but not limited to:

1) Hedge funds not knowing the real value of CDOs, and thus being very afraid to sell them...if too many got marked to market, it could cause a run on the hedge funds from investors who want their money back, and cause a financial panic not unlike the beginnings of the Great Depression;

2) Because of certain margin calls by investors, hedge funds are puking more "liquid investments" at significantly discounted prices, which basically halted a number of M&A deals from coming to the market, because people were afraid to invest;

3) Because of banks tightening their lending standards, less mortgages have been made, and because people are afraid to buy MBSs, banks have had less securities to sell to investors and investors have purchased less securities. This seizes up liquidity, as banks are stuck holding this crap AND the money supply (i.e. people tapping into equity of their homes in the form of home equity loans and mortgages) has dried up (which has been partially caused home prices to stall/fall);

4) Because of all of this, people are now repricing the rate at which is acceptable for a certain amount of risk...and waiting to get back in the market, or waiting to make moves, since no one really knows what the correct price for risk is right now, which again, has stalled investment.
posted by SeizeTheDay at 5:35 PM on August 18, 2007

SeizeTheDay, I would say that all four of these things have happened because of the fact that people (mostly bankers and institutional investors) trying to predict the value of a mortgage have either been wrong about the mortgage's value, or else they are becoming more uncertain about their ability to be right about the mortgage's value.

JPD's explanation of how mortgages have been valued, in aggregate, was very lucid. Indeed, statistics and models should in theory make it easier to value 1,000 mortgages together than it is to value one individual mortgage.

The way that I see it, and I speak under correction, is that an individual mortgage really does have a true value. Truly, in reality, either it's going to be paid off in installments according to the original agreement; or it's going to be prepaid on a specific date (either by refi, home sale, or out-of-pocket); or it's going to be defaulted on on a specific date. Whichever outcome is the true one, that outcome has a present discounted value that could be calculated if the outcome were known. It seems to me that the practice of betting on the probabilities of these unknowns - "pricing risk" as you Wall Streeters would have it - is what has really gotten the market into trouble, as the pricing models may not have been accurate.
posted by ikkyu2 at 9:55 AM on August 19, 2007

ikkyu2, from CNN.com, one of the smartest things I've read in a while, because the liquidity crisis has so many root causes (and can also be blamed on Greenspan and the Fed's lowering of rates to historic/near historic levels, which had the same effect of the Yen carry trade, but on a massively larger scale):

The most dangerous words on Wall Street
Wilbur Ross
Chairman and CEO, WL Ross & Co

I recently overheard two men arguing about who was better off. One boasted about his new car, the other about a plasma TV and so on, until one proclaimed, "I am better off because I owe more than you are worth." The second man conceded defeat. This anecdote summarizes the mortgage bubble. Americans spent more than they earned in 2005 and 2006 and borrowed the difference. The federal government did the same. Everyone secretly feared this was unsound but wanted immediate gratification, so there was applause for talking heads who said global liquidity would make these borrowings safe. Alan Greenspan went so far as to suggest that people take out adjustable-rate mortgages.

Liquidity, however, is not about physical cash; it is mainly a psychological state. Subprime problems have consumed only trivial amounts of global cash but already have burst bubbles by shocking lenders. Clever financial engineering effectively had convinced lenders to ignore risk, and not just in subprime. A major hedge fund participated in a loan to one of our companies, but sent no one to a due diligence meeting. So I called the senior partner to thank him and tell him about the non-attendance. He responded, "I know. For a $10 million commitment, it wasn't worth going to a meeting."

When subprime issues first surfaced this spring, many major institutions said they had none, but recent quarterly write-offs show they did. They weren't lying; they just didn't know what they had. Their embarrassment has brought risk control back into vogue. It was always silly to lend to weak credits at discounted interest rates, and without documenting income and balance sheets and without appraisals. No amount of model building should have enabled Wall Street to take $100 of such paper and alchemize it into securities sold for $103. Models inherently assume a future similar to the past and therefore they fail when multiple standard deviations occur. Subprime models also did not capture ever more lax credit standards nor that real estate might suffer severe and protracted price declines, again proving that the two most dangerous words in Wall Street vocabulary are "financial engineering."

Now that we have identified the cause of the disease, how severe and how contagious is it? The present $200 billion of delinquencies will grow to $400 billion or $500 billion next year because $570 billion more low, teaser-rate mortgages will reset to market and consume more than 50% of the borrowers' income. Therefore most of the loans will be foreclosed or restructured. Probably 1.5 million to two million families will lose their homes. Meanwhile, few lenders will put mortgages on the foreclosed houses, so the prices will plummet. Despite these tragedies, total losses will probably be less than 1% of household wealth and only 2% to 3% of one year's GDP, so this is not Armageddon. However, even prime jumbo mortgages will be more expensive and more difficult to obtain.

Similar excesses occurred in corporate debt markets. Leveraged buyouts were financed with few or no restrictive covenants and with some borrowers able to "toggle," or issue more bonds to pay interest in lieu of cash. The debt-to-cash-flow ratio hit record highs, and more than 60% of junk bonds issued are rated B or lower. Only 13% of high-yield issuance proceeds was for capital expenditures for expansion--87% went for sponsor dividends, stock buybacks, LBOs, or refinancings, none of which inherently advance credit worthiness. And this exotic lending paid only 2.5% to 3.0% more interest than Treasury bonds' 5.5%. Therefore investors received only 8% or 8.5% interest on bonds that had a 25% probability of defaulting, the same ignoring of risk as in subprime.

The cause was also the same. Wall Street made $100 of these credits into tranches of securities that sold for $102 or more. Again we had securitization pseudo-alchemy creating fool's gold. The weakest 5% or so of a $2 trillion universe of leveraged loans and high-yield bonds will crater. This is only 1% of GDP, but lending standards will tighten for a while, just as they did after the telecom bubble burst.

Because of this outlook, WL Ross portfolio companies raised $2 billion this year to eliminate outside financing needs. More recently, we provided a modest $50 million debtor-in-possession financing to American Home Mortgage, the tenth-largest subprime lender, as it entered bankruptcy. Ultimately, we will make a major move into mortgages, because lending to weak borrowers makes sense at premium rates with proper due diligence and appraisals. After Japan's real estate bubble burst, we used a similar strategy to rehabilitate Kansai Sawayake Bank. It was earning 17% a year on equity after one year, almost twice the return typical of a Japanese bank.
posted by SeizeTheDay at 3:52 PM on August 19, 2007

Good stuff here, all. Thanks. I'm still wrapping my head around these answers, but this is something I've always been curious about, ever since my parents first had their mortgage sold many years back.
posted by azpenguin at 8:32 AM on August 20, 2007