There's an interesting post on Lansner on Real Estate today reporting Orange County's beach communities are suffering least from the current housing slump.
In an earlier post I discussed research which indicates markets with natural amenities (good weather, water, mountains) have better long term growth records than markets which lack these amenities (too hot/cold/humid, dry, flat). Orange County's beach communities are at the top of the scale when it comes to weather and water, and maybe the data referred to in Lansner's post indicate amenities mitigate against value loss in a down market. A broader study would have to adjust for the fact that many high amenity communities have more second homes and rental vacation properties, which probably have their own dynamic.
Wednesday, December 19, 2007
There's an interesting post on Lansner on Real Estate today reporting Orange County's beach communities are suffering least from the current housing slump.
Housing Wire reports few borrowers are getting FHA Secure loans. One problem (the primary focus of the post) is that few of the borrowers who apply have been approved (266 out of 3,200). As Housing Wire says, this could be because few qualify, but it seems more likely to me that there are probably the same kind of logistical problems with the FHA program as I anticipated for the Paulson modification program in one of my previous posts.
A bigger issue is the fact that only 3,200 (out of an anticipated 240,000) borrowers have applied. Low borrower anticipation rates are to be expected, see my previous post on this topic.
Tuesday, December 18, 2007
A lot of the analysis I do is based on home value data derived from either the S&P/Case-Shiller House Price Indexes or OFHEO's Housing Price Indexes. I also use the Bureau of Labor Statistics Local Area Employment data and the Commerce Department's C-40 Residential Permit Reports for supply and demand trends. These data sets are useful because they drill down to metropolitan areas (there is wide variation in performance between markets). Also, they are published frequently (monthly for S&P Case-Shiller, BLS, and Commerce Department data, quarterly for OFHEO). However, like all data there are issues with each data set.
Although the S&P/Case-Shiller and OFHEO indexes use a similar approach to estimating value changes, there are important differences which are explained Andrew Leventis (an OFHEO economist) here. The S&P/Case-Shiller data has less complete geographic coverage, so data may not be available for some markets, and the aggregate index may be skewed by the smaller set of markets covered. The OFHEO index relies on data from Fannie Mae and Freddie Mac loans which means homes using jumbo loan financing are not represented. If, for example, prices on high priced homes are falling faster than more affordable homes and the OFHEO sample is overweighted with affordable homes the OFHEO index will not fall as fast as an index including the higher priced homes. When possible I try to use both indexes and keep in mind their potential biases.
The primary issue with the BLS employment data is explained in this post from The Big Picture. The BLS attempts to estimate employment changes from new businesses just starting up and from businesses which have terminated using a birth/death model. The consensus is the model tends to understate job losses during a slowdown and understate job gains at the start of a recovery. Again, the best approach seems to be an awareness of the potential bias.
Finally, the permit data issues relate to potential reporting and sampling errors. The Commerce Department summarizes the issues here.
None of these data sets seem fatally flawed, especially if the data is looked at over a period of time and averaged over a number of periods to miminize the impact of errors or distortions in any one period.
Monday, December 17, 2007
Hindsight is 20-20, and I know I've made a few bad calls over the years. Still, I had occasion to look through Harvard University's Joint Center for Housing Studies' The State of the Nation's Housing 2006 last week, and it has some gems:
"The greatest threat to housing markets is a precipitous drop in house prices. Fortunately, sharp price declines of five percent or more seldom occur in the absence of severe overbuilding, dramatic employment losses, or a combination of the two...With building levels still in check and the economy expanding, large house price declines appear unlikely for now..."
Not that they didn't suspect something was a little wacky:
"Until 2000, national weighted average home prices rose closely in line with median household incomes and general price inflation. Since then, however, house price appreciation has shot ahead of those benchmarks, outstripping income growth more than six-fold from 2000 to 2005."
Not to worry:
"But, when and if house prices do fall, the so-called bubble is more likely to deflate slowly rather than burst suddenly..."
"Over the longer term, the outlook for housing markets is favorable... improvements in the mortgage finance sytem over the past several years, together with stricter inventory management in the home building industry, will help to dampen boom-bust cycles in the future."
Improvements in the mortgage finance system? What a difference a year makes.
Sunday, December 16, 2007
Calculated Risk has a recent post talking about Housing Inventory and Rental Units. The bottom line of the piece is that most people's estimates of excess housing inventory are too low because they don't consider vacant rental units. It also contains the assertion that home builders have built too many homes. I have a different viewpoint.
A housing unit, rental or ownership, is a housing unit. Most housing units are in 1-4 unit structures, and it is very easy for these units to move from rental to ownership and back again. In recent years large numbers of rental units in even 5+ unit multifamily structures converted to ownership as the buildings were converted to condos. When you talk about overall supply you really need to talk about the whole picture; trying to parse what's a rental and what's ownership is an exercise in futility.
"Too many" residential units happens when there are not enough households to occupy what's been built. It's a topic for a separate post, but I argue that housing demand is closely correlated with employment, and in recent years the number of residential units built has been generally in balance with employment growth. Yes, there are too many homes built in the sense that they're not selling and builders are continuing to add to supply as they work through their pipeline. However, this is a price issue created by the loose underwriting bubble popping. There are still plenty of households who need housing at the right rental rate or the right price and debt structure. Prices will adjust to the necessary levels.
The availability of cheap, loosely underwritten financing unquestionably drew a number of renters to home ownership from rental units (although not as many as you might think, because home values inflated rapidly to accommodate the increased demand such that many renters continued to be priced out of the market). You would expect this would result in more vacant rental units and reduced rental rates. To a certain extent this happened, but the effect was dampened because at the same time some rental units were converted to ownership units, reducing the rental unit supply.
Now, the reverse is occurring; a certain number of owners are losing their homes and becoming renters again. You would expect this would result in reduced rental vacancies and rising rental rates, and to a certain extent it is. However, the effect is dampened because ownership units are being converted to rental units, increasing the rental unit supply.
This interpretation is supported by the Census Median Asking Rent data. This data bounces around a lot and is seasonal, so I think the most helpful way to look at it is as percentage change over a rolling four quarter moving average. It looks like this:
Because this is a rolling average it lags, but what it shows is rent growth decelerated from 2001 through 2004. I believe this started out as a recession effect and continued as renters shifted to ownership. Keep in mind this was not a huge shift, rent growth was only negative three quarters of this period. Since then rent increases have accelerated as owners have shifted to renters, but again the effect is not huge. In fact, a large part of the increase is a reflection of a spike in 4th Quarter 2006 which looks like an anomaly.
My bottom line is the current situation is less about supply and demand and more about a price bubble created by leverage popping.
Saturday, December 15, 2007
Last night I dusted off my copy of Charles P. Kindleberger's classic Manias, Panics and Crashes (I have the 2000 edition which leaves off at the East Asian Financial Crisis in 1997; the link is to the 2005 edition which no doubt has something to say about dot-coms). Kindleberger outlines the life cycle of a bubble, and summarizes 40+ events dating from 1618. My edition has a blurb on the cover from Paul Samuelson which says, "Sometime in the next five years you may kick yourself for not reading and re-reading Kindleberger's Manias, Panics, and Crashes." This was good advice for residential real estate and RMBS investors.
Our current situation fits Kindleberger's taxonomy perfectly:
Object(s) of Speculation: Previous bubbles have related to tulip bulbs, canals, cotton, railroads, coffee, Argentine securities, bank stocks, and many other commodities and financial instruments. In the current case, it's homes and residential mortgage backed securities (RMBS) financing homes.
Exogenous Shock Setting off the Mania: Subject to argument, but my belief is the trigger was the rapid decline and extended period of very low short term interest rates between July, 2000 and July, 2004 compounded by flawed securitization models which did not appropriately price risk and investor misinterpretation of the risks underlying the securities.
Scandals and Defalcations: Subprime fraud revelations.
Turning Point: Week of June 18, 2007, collapse of Bear Stearns hedge funds.
Domestic Propagation: Falling home prices, mortgage securities indices.
Crisis Management Devices: SIV Superfund, Joint Term Auction Facility, Paulson Modification Plan.
To a certain extent it's comforting that current events fit a well established pattern.
Thursday, December 13, 2007
Most news coverage focuses on national housing trends, largely because there's more data available. However, there are huge differences in how home values have performed between markets, and one of the performance drivers is natural amenities.
The natural amenity concept was developed by David McGranahan, an economist with the United States Department of Agriculture. McGranahan was investigating why some rural counties grew while others lost population. His idea is very basic; people move to places that are pleasant. McGranahan developed an index which scored places based on weather (How cold in the winter? How hot and humid in the summer?), topography (hills and mountains are more interesting than flatlands), and water surface (coasts, lakes, ponds and rivers are more interesting than drylands). When he mapped the index and population change, this is what he got:
Wednesday, December 12, 2007
Tuesday, December 11, 2007
A post yesterday on The Big Picture suggests the current mess was caused by cheap money and slack underwriting allowing millions of renters access to home ownership which they couldn't afford. In the same vein, this weekend's Wall Street Journal (12/8/07) has six letters to the editor all decrying "Mortgage Bailouts." Here are some excerpts:
The Ants and the Pickups buy on these terms. The Grasshoppers don't have $75,000 to put down and they rent the house instead of buying it.
Monday, December 10, 2007
The Big Picture has a post this morning suggesting cheap money and slack underwriting put renters into home ownership, and as these owners are foreclosed on they will be returning to rental status. The residential rental market is already fairly tight, and if this shift occurs here is the feedback cycle:
1) More rental demand equals higher rents.
2) Higher rents mean an increase in the CPI. Housing is by far the biggest component in the CPI, and the CPI calculation is based on rent levels, not house values. The CPI did not pick up the escalation in house prices because it focuses on rents and rent increases didn't keep pace with escalating house values. As rents increase the CPI will not pick up the falling house prices.
3) Higher inflation will put upward pressure on interest rates, exacerbating the housing crisis.There's a scary feedback loop for you.
Ultimately, an equilibrium will be reached as investors buy foreclosed houses at prices which can be supported at market rent levels. That equilibrium is probably years away.
Sunday, December 9, 2007
The Wall Street Examiner has a post which draws the correct conclusion about the Bush Administration modification plan and which identifies a great source of data (the GAO's "Home Mortgage Defaults and Foreclosures briefing). But, the payoff line of the post is, "The main driver of foreclosures is the change in real estate prices." I think that's wrong. The main driver of foreclosures is income curtailment. Falling real estate prices are a condition which sometimes leads to foreclosure when the driver is in effect.
I've written a previous post discussing research identifying causes of foreclosure, with Income Curtailment (most commonly loss of employment) #1 on the list. I do think there is some fuzzy thinking in the research, the surveys, and/or the minds of the people answering the surveys when it comes to other items on most of the lists. For example, #2 is typically Illness/Medical. I doubt in and of itself illness causes a whole lot of defaults ("Sorry, lender, I'm too sick to mail in my payment"). I think it's more likely people got sick and lost their job (income curtailment), had to get by on reduced disability payments (income curtailment), or they or a dependent incurred major medical expenses (income curtailment again, in the broader sense of less net income available to service debt). Similarly, #3 on most lists is Divorce. The most likely scenario here is there were two incomes available to service the debt, and post-split there's only one (income curtailment). These three causes account for 80% of defaults. Rate Resets in the research referred to in my previous post are #7 (clearly not a driver).
Where do falling home values come in? During the happy days if your income was curtailed and the value of your house was up you could refinance and pull out the cash you needed (those no doc loans were a boon for the unemployed borrower), or you could sell your house. Either way, foreclosure was avoided. As teaser rates climbed and underwriting standards tightened, the refinance option went away, but you could still sell. As home values fall to the point there's no equity the sale option goes away for more and more people, which leaves foreclosure. Two additional points:
1) Income curtailment happens to everyone in the socioeconomic spectrum; what distinguishes those who default from those who don't is whether or not they have resources (savings, insurance, etc.) to tide them over. It's a lot easier and cheaper to tweak the rate reset features of subprime loans than it is to provide an income safety net to support those whose income has been curtailed. But, don't expect big results from the tweak when the real driver is more fundamental.
2) Income curtailment is more widespread and prolonged during a recession. If we slip into a recession foreclosures will go up significantly and home values will fall faster and farther.
Saturday, December 8, 2007
The Big Picture correctly describes the BLS emloyment release on Friday as continuing evidence of a slow motion slowdown. That conclusion isn't all that clear from the accompanying graphs, but looking at the rolling average 12 month employment change does make it clear:
Friday, December 7, 2007
Here are a couple of different ways to look at the same housing value data and draw some radically different conclusions. The data are OFHEO's US Housing Price Index. When you graph the index itself it looks like this:Almost peaceful looking, isn't it? The most recent quarters clearly show prices are flattening, but all this talk of plummeting values seems overblown.
Another way of looking at the same numbers is to compare the percentage change in the index over a rolling 12 month period (the first data point is the percentage change between Q1 1975 and Q1 1976, the second data point is the percentage change between Q2 1975 and Q2 1976, and so on). When you graph that, you get a much different picture:
The only nice thing you can say about this chart is the deceleration today is not as bad (yet) as the period from Q1 1979 and Q3 1982. Of course, those of us who were around then remember that period as pretty dark days for real estate.
These charts are good examples of why rates of change almost always give a clearer picture of what's going on than looking at absolute levels of almost any index. For a great discussion of this point, see Ahead of the Curve by Joseph Ellis.
Thursday, December 6, 2007
What causes home foreclosures? You don't get the answer from current mainstream journalism headlines, which are fixated on adjustable rate resets. The Peridot Capitalist posted a chart from a Countrywide investor presentation reporting more than 80% of foreclosures are attributable to loss of income, illness, and divorce (payment adjustment ranked 7th at 1.4%). A post on Naked Capitalism has some good commentary on what this means for the mortgage modification plan.
I have a couple of additional observations:
1) These are the causes of foreclosure for all borrowers. They are not unique to subprime borrowers.
2) Given current economic conditions, why are foreclosures increasing now? Loss of income generally means loss of employment, but unemployment is not going up, and illness and divorce are both steady state conditions that don't fluctuate much. The answer is an underlying unstated condition to all these reasons: the borrower couldn't sell the house for more than the mortgage amount. A foreclosure requires both inability to pay the current debt and the inability to liquidate the asset for more than the amount of the debt. The bigger question is, why did houses stop appreciating? That's a topic for a separate post, but in a nutshell I believe the huge escalation in house values between 2003 and 2005 was driven largely by a demand spike created by easy financing, and when the pool of borrowers taking advantage of this financing was saturated there was nothing left to drive prices higher.
3) Loss of income and divorce both spike during recessions. If we slip into a recession, the foreclosure situation will get much, much worse.
Tuesday, December 4, 2007
Which direction are home values going to go? Is net operating income from residential rental properties going to go up or down? Subprime foreclosures, interest rate changes, tightening underwriting standards, a weak dollar, potential recession, and overbuilding are all variables being used to construct endlessly varying scenarios to explain what's going to happen next. Most of these scenarios have elements that make sense, but which ones make the most sense?
The simplest way to go is to simply listen to the opinions of experienced real estate professionals. There are a lot of smart people who have been around through multiple real estate cycles, and many of them are quite vocal. Although you will hear a lot of differing opinions, you can pick the one that makes the most sense to you.
I personally like fact based decisions. One fact based approach is to take available historic information, develop a model which explains the history, and use that model to forecast future results based on current data. Such models typically have dozens of variables, and the math becomes very complex. For example, a recent paper assessing the likelihood and factors underlying subprime defaults contains the following paragraphs:
I was at a dinner the night before an Urban Land Institute forecasting conference last month, and the keynote speaker (an economist with a major accounting firm) mentioned their commercial property forecasting model had more than 60 variables. Almost always these models are proprietary and we never get to see how they work, and even if we did get to see them, most of us are not going to question the math. Instead, we decide whether or not to believe the forecast based on the reputation of the person or firm presenting the forecast. In, the end, this is not all that different than just relying on the opinion of the professional that seems to make the most sense.
The approach I like is to rely on some simple rules of thumb. Simple rules of thumb sound like a country boy, backwoods approach to decision making, but academics have demonstrated they can be a powerful decision making approach which rival complicated models. Academics don't call them simple rules of thumb, they call them simple heuristics (hyur-is-tics). I will too, because it sound more sophisticated.
What kind of rules apply to residential property net operating income and values? Here are a few which I will discuss in coming posts:
- Residential property values go up when interest rates go down.
- Residential property values appreciate and net operating income growth is strong when the ratio between new jobs and residential permits is greater than 1:1.
- When values escalate more than what is indicated by the first two rules they revert to the mean.
- Markets with a combination of good weather, interesting landscapes, and water do better than markets that have bad weather and are flat and dry.
How can simple heuristics we all can understand compete with complex models? Because in any complex situation there are always a few factors which have a strong effect and many other factors which tend to cancel each other out. Complex models try to take all the variables into account by looking at historical results, but when they do that they sometimes end up with a model that only reflects that history and is not a good predictor of the future (a result called overfitting).
An excellent discussion of these principles can be found in a book by Gerd Gigerenzer titled Simple Heuristics That Make Us Smart.
Parts I and II of this series have discussed the logistical and borrower related impediments to making Paulsen’s subprime modification plan work. This installment discusses the issues on the lender side.
To keep it simple for now, let's forget that the lenders on these deals are not single entities, they are multiple-personality schizophrenics consisting of servicers, subordinate tranche holders, senior tranche holders, and sometimes insurers. Let's also forget that even if these lender components wanted to behave rationally they are tied together in a web of contracts and fiduciary responsibilities that their respective legal counsels will not lightly bl0w off. Ignoring all this complexity, most commentary seems to assume that on the lender side a modification will result in a better recovery than a foreclosure. That's a pretty big assumption.
Certainly it does not take much imagination to envision how lenders foreclosing on houses dump supply on an already distressed market, creating a downward spiral in prices which contribute to more foreclosures. Yes, the way to avoid the spiral is to stop foreclosures through modifications. But, whether or not your losses are minimized by modifications depends on market conditions when the modification ends and what happens to your borrower in the meantime. If prices are even lower at the end of the modification or your borrower defaults during the term of your modification before prices have recovered you will be worse off than if you had foreclosed immediately. This is a real possibility, and if we actually slide into a recession I would argue it's a probability over the next 3 to 5 years.
As Yogi Berra (may have) said, "It's tough to make predictions, especially about the future." Jim Cramer's second investment commandment is "Your first loss is your best loss" (not an original idea of Jim's). Many lenders firmly believe this is true, and that's why in previous real estate recessions you saw lenders selling pools of debt at large discounts to bottom feeders willing to work through individual deals to maximize value. Some of the parties that need to buy into the plan on the lender side to make it work will have this philosophy and won't play ball.
Monday, December 3, 2007
Felix Salmon has a post discussing the low likelihood of investor litigation arising out of the proposed mortgage relief plan (Portfolio Market Movers). I think he's right about the poor economics of pursuing litigation and the fact the investor profile is not litigious, but I do disagree on a couple of points:
1) Depending on what happens, it could be pretty easy to establish damages on any particular modification. Let's say the servicer agrees to a seven year modification at a below market interest rate on a nice house in a good neighborhood. Let's say that market bottoms in 2008 and comes back to current levels in 2010. It's not hard to calculate in hindsight how much the investor lost as a result of the servicer entering into a long term modification when a short term modification would have worked just as well. Whether or not something like that will happen and whether or not the investor will take the trouble to pursue it is unknowable at this point, but the possibilities are likely to be on the minds of legal counsel for the servicers.
2) Even if we agree it's unlikely there will be investor litigation, I don't think that will matter much to servicers. I can't see counsel for the servicer saying, "Go ahead, sign that modification - there's not much risk" when the servicer can remove the risk by getting investor consent. Getting the consent will undoubtedly slow the process down, possibly to a halt, but I think that's what servicers will do.
Sunday, December 2, 2007
The first obstacle to getting a lot of subprime loan modifications done is Logistics. The second problem is the borrowers themselves. The Wall Street Journal reported on Monday (Citigroup Feels Heat to Modify Mortgages 11/26/07) ACORN and Citigroup got together and sent letters to 340 Los Angeles homeowners inviting them to a workshop to help them keep their homes, and got a dozen responses. A 3.5% response rate does not bode well for modifications as a route out of this mess, and while there could be plenty of reasons why this particular event did not come off, the reality is a lot of borrowers are not going to ask for modifications. Some won't be aware of the modification options, no matter how much effort is made to communicate it. Some will analyze the available modification options and make a rationale conclusion that based on their particular situation they are better off walking away. For those who don't participate, the biggest group will be those who just want to move on. This is not a crazy decision if you stretched to get into a house whose value has fallen and which you don't see bouncing back anytime soon.
Post Northridge earthquake (1995) I was a consultant with the City of Los Angeles trying to facilitate the reconstruction of earthquake damaged condominium projects. These were deals which had everything going for them to make a workout happen; an Act of God (so no blame game to cloud the picture), insurance proceeds sufficient to reconstruct in almost every case, and 0% 30 year loans from the City (courtesy of a Federal HOME loan grant) to cover any gaps. The borrower participation rate? Less than 50%, and almost all of them were people who had substantial equity before the earthquake. It's hard for me to see how in a situation where people almost by definition have no equity we can expect large numbers to buy in.
This is not to say a program like this is a waste of time; appropriate modifications can be good for all parties, they can create a happy ending in individual cases, and the option should be available. I am saying even if a program like this gets off the ground there will be a whole lot more foreclosures than modifications, and low borrower participation rates will be one of the reasons.
Saturday, December 1, 2007
There are a lot of reasons why few borrowers are going to get modifications of their subprime loans. Although Friday's Wall Street Journal (U.S., Banks Near a Plan to Freeze Subprime Rates 11/30/07) has a front page article reporting that the Treasury Department and a number of large residential mortgage servicers are close to agreeing on a plan to freeze subprime home loans, the truth is more along the lines that they are agreeing on how to look like they're doing something to address the problem.
Problem I: Logistics. The Journal quotes Treasury Secretary Henry Paulson as saying [That it would be impossible to] "process the number of workouts and modifications that are going to be necessary doing it just sort of one-off." He's right, for reasons eloquently outlined by others (for example, Tanta: "Dear Mr. Paulson"). But, since they're working on a plan anyway, he must be thinking there's a way to modify loans in bulk. According to the Journal, the plan will be to divide the modification candidates into three groups:
1) Those who can make their payments when their rates go up. This group would not get modifications, presumably because they don't need a modification.
2) Those who can't make their payments even if their rate doesn't go up. No modifications for this group either, presumably because it wouldn't do them any good.
3) Those who can make their payments if their rate doesn't go up. These folks would get their rates frozen for a period to be determined by criteria to be worked out.
That's all well and good, except I think maybe we should exclude from Group 3 the borrowers who committed fraud to get their loans in the first place. It's hard to know exactly how big this group is, but apparently there could be quite a few members and the original loan files helpfully already identify many of them (Tanta again). Culling out these borrowers will spare Mr. Paulsen the embarrassment of the inevitable Wall Street Journal follow up story, "Treasury Secretary's Plan Aids and Abets Subprime Fraudsters."
My question is, how do you sort borrowers into these groups without resorting to "sort of one-off" analysis? By relying on the same crappy data tapes used to securitize the loans in the first place? Doing the modifications requires reunderwriting the borrowers, and the servicers don't have and are not likely to hire the people necessary to do that.