Wednesday, August 15, 2007
The havoc among mortgage originating firms has been so severe that it has generated two schools of thought. The first view is that classes of mortgages that are not in peril are being devalued unfairly, and thus that this panic is temporary and will readjust to normalcy within a few months. The second camp is that reality is returning in a vicious sudden attack, and that only the strongest will survive.
[Note that the point of this post is that the bad underwriting has created market conditions in which careful underwriting no longer has the benefit it should carry. There ought to be a much greater return for due diligence, doing old-fashioned things like verifying income and assets, securing real appraisals, and in some cases more than one appraisal, conducting real due diligence on loans bought or originated by delegated underwriting, etc. But there isn't turning out to be that much benefit in the market these days. Consider the case of a good and conservative borrower, who bought a home with an old-fashioned fixed-rate mortgage that is quite affordable to that borrower. But the borrower's neighbors didn't - and now the borrower is sitting on a major loss, because the borrower's neighbors went for those neg-ams, or ARM no-docs, or stated, and are now selling before default. It's a rotten situation to be in. When market risk (especially geographic risk) overwhelms the efforts of good underwriting and careful buying, there is something badly wrong with the entire system. This is why we need a nationwide system of mortgage regulation. Thornburg wasn't a bank, but it performed due diligence and followed strategies that should have prevented it from ever being in this situation.]
I hold generally to the second point of view. Let's take, for example, Thornburg Mortgage. Housing Wire cites an AP story which includes this:
The flight from risky mortgage debt has now spread to loans carrying little actual credit risk. With a $56.4 billion portfolio, Thornburg is the largest mortgage-related security REIT and owns primarily prime loans. But the absence of liquidity for home loans means their market prices have tumbled this year …And Tanta at CR commented:
In my view, outfits like AHM and LUM were accidents looking for a place to happen. Thornburg makes those pretty, competitively-priced, conservatively-underwritten jumbo loans whose rate spread over conforming has tripled over the last few months.Both Tanta and Paul Jackson of Housing Wire are industry insiders. They know their stuff really, really well, and ordinarily I would merely genuflect at their altars and feel thankful for their efforts. Now I must disagree to some extent.
I will concede that the news about Thornburg is shocking, and the reaction genuinely is somewhat overblown. However, this doesn't mean that Thornburg's underwriting and rates won't have to adjust to account for higher risk in the marketplace, and their book value has taken a real loss.
First, Thornburg probably has losses in the loans it has written over the last few months, and it certainly has losses in its holdings (it estimated that its book value dropped from $19.38 to $14.28 during August). Second, it's a REIT, and one of the REIT features is that they must distribute the great bulk of their income each year. This leaves them in a poor position to compete during disruptions, because they don't have the retained earnings to carry them for six months or a year. I am strongly negative upon the use of the REIT structure for this sort of business enterprise, because I believe that being unable to retain more income is a fatal flaw in the mortgage business. To compound the problem, Thornburg doesn't seem to have retained the 10% it could retain. Third, jumbo mortgages do carry higher risk in a situation like this, because the fundamental problem is not "illiquidity", but a discontinuity between home prices and effective incomes of the would-be buyers, plus overbuilding in far too many markets. The situation is not going to improve in the short term. In many markets, any real recovery will arrive after 2010.
The problem with jumbos is that bracket compression dictates that on average, housing priced at the 75th percentile or up is doomed to lose more value than homes priced around the median in a significant downturn. In many markets, these homes can drop 25% to 35% of their value unless a would-be buyer can afford to take a year or two to sell them. Worse yet, there are markets in which a disproportionate number of owners of these homes are involved in either real estate in some capacity (agents, brokers, builders, or financing), and therefore will be disproportionately suffering economically in this downturn. Underwriting on these loans has been exceptionally lax during the run-up of this bubble, and this produce a sharper downturn. Accentuating the normal trend, homebuilders have been concentrating on building these higher-priced homes, so the oversupply and the consequent undercutting of prices on these homes bought in the last few years has been jump-started. (Because builders have been selling at steep discounts, recently built developments are often seeing the sharpest price drops.) And then there are the three D's - divorce, death and disability - which dictate that no matter how good the underwriting, some perfectly stable homebuyers always find themselves having to sell long before their expectation.
Let's take the situation as it stands. Yes, most would-be buyers using jumbos are getting charged higher rates, but the real constriction is most markets is that it is hard to find 100% financing for these homes again, whereas for a few years a buyer with a high credit score would have no problem buying without a downpayment. In states in which the liability for a purchase-money mortgage is generally restricted to foreclosing on the home, right now it is a pretty risky proposition to be handing out 100% mortgages on these things. Even if the borrower continues to be able to pay, the borrower might well choose to walk away to limit his or her loss. (The standard way to do it is to first buy another house, then default on the losing proposition, thus sticking the creditor on the first one with a 20% loss or so). So right now, getting a 10% downpayment is a risk-hedging affair on these homes, but needless to say, it knocks some borrowers out of the market. This, of course, tends to suppress home prices further....
So in this environment, Thornburg's conservatism in writing loans generates less benefit than in a normal environment. I.E., there is far more area risk under current conditions than normal, so the difference between conservatively underwritten loans and laxly underwritten loans yields less in the way of lower defaults and losses than it normally does.
The rate environment continues to eliminate buyers out of the market also. It is axiomatic in the mortgage business that falling mortgage rates lead to housing appreciation, whereas rising mortgage rates dampen appreciation. Bill Poole (yes, Cramer's target), referred to this in his speech of July 20th:
It is important to emphasize that what is odd is not that there was a risk of rising short-term interest rates, as there always is, but that the market clearly expected an increase, as indicated by the shape of the yield curve. This expectation, in turn, was encouraged by the Fed’s Open Market Committee.In other words, don't come crying to Fed, you jerks. We told you, you knew it, and then you acted as if it weren't going to happen. Not expecting losses on portfolios was tantamount to pretending that water was going to suddenly start running up hill, and no one in the business can deny it. See CR on the topic.
Given these widely held expectations of rising interest rates, it is difficult to avoid the judgment that these ARM loans were poorly underwritten at the outset. It was imprudent for mortgage brokers and lenders to approve borrowers who likely could not service the loans when rates rose, and it is surprising to me that sophisticated capital-market investors willingly purchased securities backed by such poorly underwritten mortgages.
Many members of the Fed have produced part of this problem by consistently maintaining that the problem was subprime loans and that the problem would stay with subprime loans. The ratings firms have produced part of this problem, because they have ignored real risk factors such as those I describe here and have rated paper for sale above its proper level. Therefore it is not surprising that investors are now exceptionally leery of anything that is not vanilla all the way. They don't want this Ben 'N Jerry stuff; they want vanilla, because they know the market for vanilla will be there regardless. And vanilla, in this market, is agency-guaranteed.
Thornburg has jumbo ARMs that include an average of 35% down payment. Their provision for loan losses is in the mid-hundred thousands, and their actual loan losses have been miniscule, less than $100,000. This is from a portfolio of some 50+ Billion dollars.
The losses in book value are not real. They are accrued based on Mark to market values of CDO's that they have in inventory. The disappearance of value is the result of panic spilling over from the sub prime debacle. Given some rationality returning to the market, these securities will revert back to their resonably normal values. However, GAAP mandates that the securities be marked down even though there is no market for all intents and purposes.
When people make their mortgage payments in August, and the markets settle down after the screamers run out of breath, a lot of, if not all the loss in book value will come back. It's not like there are any real losses. These are strictly accrual (estimates) of values. That can change dramatically, as we have seen on the downside.
Yes the price to stay in the game will change, and likely rates will adjust upward. This is nothing at all like an Enron.
I am not trying to intimidate you on these facts. Go to WWW.Thornburgmortgage.com and get a copy of their 10Q for June 30. It's all there. However, it also is very complex finance. But take your time, and you should be able to puzzle through the facts presented.
If you need help, contact me: firstname.lastname@example.org. P.S. I have a doctorate in finance.
My personal guess was maybe that they'd get back about $3.00 of that over some months.
Don't get me wrong. Thornburg is a company that has been far more responsible than most such companies - the wretchedness of the current situation is that the responsible get hit badly along with the irresponsible! Thornburg is in a very similar situation to a borrower who put 25% down on his new home and got a fixed-rate mortgage, only to watch his home price fall by 15% - 20% in eighteen months because his neighbors weren't qualified and so are bagging out right and left. Any time you have a market in which rational behavior is punished almost as severely as irrational behavior, you have a wrecked market.
This is why we need more regulation.
There is considerable panic, but the risk factors in some localities are pretty extreme.
Uh, I'm not impressed with the doctorate in finance, because unfortunately a lot of those have been used to ornament nonsensical claims. My guess is that you are not one of them, but honestly, right now it's not the best tactic for public relations.
Thank you for your comments, because I see by them that the point I was trying to make did not come through clearly. I guess I'll put it at the top of the post.
As for "losses", losses in portfolio value of underwritten loas are universal right now. They aren't default losses, but current-value losses. FWIW, I have some customers in positions very similar to Thornburg's right now with regard to loan portfolios. I know exactly how careful they have been, so it's not like I'm criticizing Thornburg's underwriting practices. Or, for that matter, its due diligence on purchases, which far exceeds the recent industry standards.
One of things making investors wary is the concentration on ARMs which you didn't mention, btw, and which I didn't mention precisely because I am not out to harass Thornburg. But honestly, when their 2006 annual report contains items like the goal for 2007 being to originate 25% of volume from brokers, plus 7.2 billion worth of hybrid ARMs resetting from 4.63% to a "current market rate", that is going to cause a bit of uneasiness in the current situation. Nor does their geo concentration help them.
The reason why there is unreasonable fear in this marketplace is because the ratings firms screwed up. This makes companies that actually do due diligence and careful underwriting lose the investor confidence that they rightfully deserve, because no one has no independent confirmation that anyone trusts.
Furthermore, you are being a bit misleading when quoting CLTV at origination, because Thornburg did have a small amount of neg-ams (about 2 billion out of 52 billion), plus they offer very flexible refis, program swaps, etc.
And I hate to tell you this, but 35% losses are all over. I just had to break it to a very elderly person in the NE that the market value of her home has declined so severely that it must be listed at about 34% less than last year's appraisal, and in order to sell will probably go to 60%. This is not even in what is normally considered a bubble area. It is jumbo, but low jumbo.
I'm reminded of the time in grad school where I wanted to take an options pricing class that required the prof's permission. Stopped by the guys office and he asked me a couple of calc. questions, which of course I couldn't answer. At that point is was told that options pricing was beyond my abilities.
My wife and I laugh at that every time we step into our Lear 31
First I think Peter Gabriels "I don't remember" would be more in tune with the bankings folks posture at the moment.
Second I think Maxed-out-Mama is right about the sharp drop in values. If JUMBO (over 417,000) mortgages are going to require 20 percent down,with higher rates(.4~5 percent extra) how many folks have the required cash to buy such homes? Sure there are some, but in areas such as CA where most homes are over 450,000 dollars it means the end to almost all sales, until the prices drop.
Now you might say these folks will get the cash from the folks buying theri home, but the bottom end homes are so expensive that few of the first time buyers can come up with 5 percent, and 20 percent will beyond their means for years. If the entry level folks are cut out, then no one can move up.
More to the point, I note that the thoughful folks that have piles of cash and could buy are not buying. For years they have seen and heard the dumb and those with a limited event horizon prate on and on about their "investment homes" when the fundementals dictaed it had to end pooerly for most of the Alt A and subprime buyers. In the Northeast is seems the stupid money outbid the more prudent home buyer who had some foresight. These folks have hunkered down and are not about to rush back in to save the stupid, the banks, the brokers etc. I think these potential buyers, who have had their dreams put on hold by the fols that rushed in, are now going to sit back and not purchase until the prices are very low. They are semi-enjoying the "blood in the water", so to speak. I say this as a home owner who will see a hit, so it is not personal, just an observation.
The financial insitutions made such obvious mistakes, with the foresight of a nearsighted mole, that I do nto think there is a similar situation to this in the housing market since the 1930's.
About the wine...if things keep going like this, the price might drop enough that I could drink some of it.
And yes, some of these mistakes were obvious. That is Poole's precise point. Who the heck doesn't account for the most basic relationship of them all? Rising rates = higher monthly cost = lower house price.
CF - I have a very strong suspicion that you knew exactly how that sentence would read to someone who is not an economist. That was brilliant. Drink the wine now with my blessing.
It is the same "safehaven" thinking that is supporting the GSE's. Sure the CLTV's in their book are low, but so are the capital/reserves they have to offset losses in that book. All it takes are for losses to exceed the few standard deviations they have modeled, and the capital evaporates. As you note, the probability of the 4+sigma event is not a function of GSE underwriting, but a function of everyone else's.
It's also why to write or speak of "containment" to subprime was always a farce.
Links to this post: