Debacle at 250 Montgomery Street: Other People’s Money

GlobeSt.com has a story about the debacle at 250 Montgomery Street in San Francisco:

Realty Finance Corp. of Connecticut has sold its original $47-million loan on a class A office building here for approximately $25 million or $200 per square foot, according to a source familiar with the transaction. The building is 250 Montgomery St., a 15-story, 126,736-square-foot office building completed in 1989 at a cost of about $41 million.

The borrower, Lincoln Property Co., paid approximately $47 million or $405 per square foot for the building in late 2006 and defaulted on the loan in late 2008. Prior to the note sale Lincoln agreed to hand over the property to its new creditor in lieu of foreclosure…

In its first quarter filing with the SEC in March, Realty Finance said the loan matured in March 2009 without payment, pushing it into default. At the time, Realty Finance expected to lose between $0 and $11 million on the sale. The actual loss appears to be closer to $22 million. Whitehall Street Real Estate Funds reportedly had an additional equity position in the building that has been completely wiped out.

So Lincoln paid $47 million in 2006, Realty Finance loaned $47 million, and Whitehall had an equity position? That would suggest Lincoln had little if anything in the deal at any point. Call me old fashioned, but when a major investor like Lincoln (which at the time was perfectly capable of raising cheap equity or borrowing at a low cost of funds) brings in an equity partner like Whitehall, the only conceivable reason is to eliminate it’s risk in the deal. Red flags should go up under these circumstances – I’d love to know what Whitehall and Realty Finance were thinking.

After the Honeymoon: Trusting Loan Brokers

Should a loan broker who has established a successful relationship with a lender be trusted by that lender? Not according to research by Mark Garmaise, a finance professor at UCLA Anderson (working paper “After the Honeymoon: Relationship Dynamics Between Mortgage Brokers and Banks”). From a July 6, 2009 Financial Times story on the research:

The financial industry’s vaunted belief in trust and long-term relationships is being challenged by research showing that before the crisis US mortgage brokers fed loans of deteriorating quality to the banks they did most business with.

By questioning the prevailing wisdom that dealing with well-known counterparties is more fruitful and less risky than venturing into new relationships, the academic study puts in doubt one of the banking sector’s most enduring beliefs.

The key findings of the study:

• The quality of the loans submitted by the broker deteriorates over the course of the relationship
• The volume of loans submitted grows even as the quality deteriorates
• The effect is stronger for geographically distant brokers
• Even though the bank’s ability to evaluate the quality of the broker’s loans increases over time, the bank is increasingly reluctant to terminate the relationship.

It’s easy to dismiss this as a problem unique to loan brokers, but what if it’s true in other situations where initial monitoring is high and then relaxed over time? For example, the first few times you use a new appraiser you might carefully scrutinize the work. Do you need to do that every time, or can you relax? It’s a big enough topic for a separate post, but I think the answer (for commercial real estate, at least), is to check the key elements every time, no matter who you’re dealing with. Finley Peter Dunne had the right idea: “Trust everybody, but cut the cards.”

Mortgage Modification Blues

The New York Times article "Paper Avalanche Buries Plan to Stem Foreclosures" documents the logistical nightmare of processing single family mortgage modifications. An excerpt:

A note in the system shows that the bank confirmed receiving documents on April 29 — pay stubs, tax returns, a letter disclosing her hardship, bank statements. Since then, the company has been waiting for WaMu to review the file.

But when Mr. Lavi calls, a representative coolly discloses that the application has been rejected because one document, a proof-of-insurance form, is missing. He must start over.

“The file had been submitted properly, and you didn’t put the pieces together,” Mr. Lavi says, his body quivering with anger. “I’m not going to stand in line again for another six months.”

He demands to speak to a supervisor, but the representative says none is free. He hangs up and redials, hoping to land in a different call center. Eventually, he reaches Chase’s executive offices, where Becky takes over the call.

“We’re not taking cases now,” she says calmly.

“Why was I transferred to you?” Mr. Lavi asks. Becky does not know. He implores her to keep the file open while he faxes in the lone missing document.

“Impossible,” she says, warning of “the sheer amount of papers coming in.”

So, to get a modification on a WAMU (now Chase) loan, you need pay stubs, tax returns, and bank statements? Contrast that with the process of getting the loan in the first place, as reported in the New York Times piece, “Saying Yes, WAMU Built Empire on Shaky Loans.” An excerpt:

As a supervisor at a Washington Mutual mortgage processing center, John D. Parsons was accustomed to seeing baby sitters claiming salaries worthy of college presidents, and schoolteachers with incomes rivaling stockbrokers’. He rarely questioned them. A real estate frenzy was under way and WaMu, as his bank was known, was all about saying yes.

Yet even by WaMu’s relaxed standards, one mortgage four years ago raised eyebrows. The borrower was claiming a six-figure income and an unusual profession: mariachi singer.

Mr. Parsons could not verify the singer’s income, so he had him photographed in front of his home dressed in his mariachi outfit. The photo went into a WaMu file. Approved.

Proper underwriting (of new loans and modifications) is labor intensive. Most servicers never had the proper underwriting infrastructure in place to originate the loans, and they certainly don’t have it now that those deals need modifications.

More at my post, “Why Did WAMU Abandon Underwriting Standards?”

Five Underwriting Issues Which Kill CRE Deals

I have an article in the July, 2009 commercial edition of Scotsman Guide which talks about five underwriting issues CRE lenders are focusing on, and which frequently kill deals in this environment:

• Upcoming loan maturities on the Sponsor’s other deals
• Sponsor liquidity
• Sponsor exposure to distressed loan types (e.g. condo construction loans)
• Lack of Sponsor experience in the market and/or property type
• Project dependence on tenants in a weak industry.

The link may take you to a free registration page…

Willingness to Pay, and Crescent Resources

Generally, borrowers pay until they can’t (a theory I discuss in more detail in “Does Recourse Matter on Income Property Loans?"). There are exceptions, of course; a recent example is Crescent Resources, LLC, which (along with more than a hundred subsidiaries involved in separate developments) filed bankruptcy last week. From Pensions & Investments:

Crescent Resources LLC, a joint venture between Morgan Stanley Real Estate Fund V U.S. and Duke Energy Corp., filed for Chapter 11 bankruptcy protection to reduce the debt level and improve the capital structure. Investors in Fund V include the $40 billion Pennsylvania Public School Employees' Retirement System, $119 billion California State Teachers' Retirement System and $6.2 billion San Bernardino County (Calif.) Employees' Retirement Association.

Although Morgan Stanley and Duke Energy (as well as their pension fund partners) are down, they’re certainly not out, and if they chose too they have the ability to write whatever checks were necessary to pay their debts as agreed.

The lesson is, although ability to pay is a necessary condition for a good CRE loan, it’s not a sufficient condition.

Why Are CMBS Multifamily Delinquency Rates So High?

The 60 day delinquency rate for multifamily CMBS loans is skyrocketing. From a Fitch release:

Declining performance, particularly in oversupplied markets, as well as in secondary and tertiary markets, has pushed the multifamily delinquency rate to 4.55%, the highest of all property types. Multifamily properties have been highly susceptible to default in CMBS during the current economic downturn.

Fitch seems to suggest the problem is the asset class, but there’s something else at work – delinquency rates on Fannie and Freddie multifamily loans are less than a tenth of the CMBS figure. From an MBA release on June 2:

Fannie Mae: 0.34 percent (60 or more days delinquent)
Freddie Mac: 0.09 percent (90 or more days delinquent)

Why are the agency loans performing so much better? I think there are several factors at work, but the main reason is the originators of Fannie Mae and Freddie Mac loans had much to lose by selling bad loans to the agencies.

Most of Fannie’s multifamily business has been originated through their Delegated Underwriting and Servicing program. Fannie agreed to buy multifamily loans which were within their underwriting parameters without prior review. The originating lenders retained the top 5% loss exposure, and shared losses after that to a maximum of 20%. A very limited number of lenders were allowed to participate (never more than 30 nationwide). Sell a bad multifamily loan to Fannie under the DUS program, and you not only shared in the loss, you risked losing a valuable franchise.

Freddie took a different approach. They didn’t require originating lenders to share in the loss, but the ability to sell to Freddie was if anything even more tightly controlled, with a limited number of lenders restricted to specific geographic areas (see current list here). Again, sell a bad loan to Freddie, and you risk losing your franchise.

By contrast, CMBS origination was wide open. But, that may be changing. The lead story in yesterday’s Financial Times:

Treasury plans strict rules for securitisation

The US Treasury is planning a sweeping overhaul of securitisation markets with tough new rules designed to restore confidence by reducing the incentive for lenders to originate bad loans and flip them on to investors…

The Treasury plans to force lenders to retain at least 5 per cent of the credit risk of loans that are securitised, ensuring that they have what investors call “skin in the game”. The 5 per cent rule – which looks set to be applied in Europe as well – is less draconian than some bankers feared.

Would such a rule have prevented bad CMBS loans? Probably not; I believe the risk of franchise loss was a much more important determinant of lender behavior. But, it’s a start.

Loan Paydowns from the Borrower’s Perspective

I previously posted comparing CRE underwriting in 2006 and today (bottom line, even if your project income is unchanged, loans are 15-20% smaller, mostly because cap rates have increased). Suppose you have one of those 2006 loans and it’s maturing. What should you do? A look at the numbers reveals borrowers are much better off if they can negotiate an extension.

Here’s an example drawn from an actual deal done in 2006. The original underwriting and today’s underwriting is summarized in the table below:

Key points to note:

• The value of the property is a little less than the current loan as a result of the NOI decrease and the higher cap rate. In other words, the original $2.2M cash invested is gone.
• The property now supports a loan of only $4.5M. In other words, to refinance the current loan, the borrower will have to put in an additional $1,551,328. The new debt and borrower cash investment total $8.2M on a property worth $6M.

Now, there are whole sets of cognitive biases which predispose people to overvalue what they own (endowment effect, post-purchase rationalization), continue to do what they've done in the past (status quo bias, sunk cost effects, loss aversion), and expect a positive outcome to their choices (optimism bias, and valence effects). We know these biases exist, and their existence helps explain why borrowers continue to perform on loans when it makes economic sense to walk away. However, when it comes to writing seven figure checks, people get rational in a hurry. We are not going to see many people contributing large amounts of money to refinance properties which do not have equity.

So, what are the borrower’s options? One is to walk away from the original $2.2M investment and default on the loan. That would make sense if the borrower sees no possibility of a value recovery on the horizon. However, almost all borrowers do foresee a recovery, want to stay in the game, and will request an extension of the loan. The most common requests are an extension at the existing contract rate, or an extension at the current market rate. The table below summarizes the economics of those scenarios, plus a third option:

Note that although nothing solves the value problem (it takes higher NOI and/or lower cap rates to do that), there is cash flow under each scenario which is a reason for the borrower to stay with the game. To make an extension more attractive to the lender, the borrower could offer to apply some or all of that cash flow to pay down the loan, or sweep it into a reserve account as a hedge against further declines in NOI.

The third scenario (Till) represents how the loan could be restructured in a bankruptcy (for more on Till, Lee, et ux. v. SCS Credit Corp, see my post Getting Tilled: How a $6,425 Truck Loan May Decide the Fate of General Growth Properties). Since this is clearly the worst case for the lender, you might think lenders would avoid the risk and extend loans without a lot of argument. I identify some of the reasons lenders may fight it out in the post What Should Lenders Do With Maturing CRE Loans?

CRE Interest Only Revisited

Barry Ritholtz at The Big Picture has a good post today on  interest only CRE mortgages (although I think he got one thing wrong, as discussed below). Interest only structures were very common during the boom. Sometimes loans were IO for the full term, but more often the loan was IO for a two, three, or five year period, so many loans made at the peak are seeing 15% –20% increases in payments now as the IO period ends. From Barry’s post:

“Investors in bonds that packaged $62 billion of debt for U.S. offices, hotels and shopping malls are bracing for more loan defaults through 2010 as Bank of America Merrill Lynch says landlords’ monthly payments may jump 20 percent or more.

Principal is coming due on the so-called partial interest- only loans as an 18-month-old recession saps demand for commercial real estate. About $179 billion of such loans were written between 2005 and 2007 and bundled into bonds, according to data from Bank of America Merrill Lynch.

With soaring vacancies and falling rents, some cash- strapped borrowers will fail to cover the higher costs, said Andy Day, a commercial mortgage-backed securities analyst at Morgan Stanley in New York. About 87 percent of mortgages sold as securities in 2007 allowed owners to put off paying principal for several years or until maturity, compared with 48 percent in 2004, Morgan Stanley data show.”

I think this is where Barry goes wrong:

Almost by definition, when a borrower users I/O financing, it suggests they cannot afford to make the actual purchase, and were unable to arrange other forms of financing.  Otherwise, the buyer would have arranged for to a less risky structure that is not dependent upon subsequent credit availability.

IO in CRE was not about maximizing affordability or leverage – although the actual payment was interest only, loans were still underwritten assuming amortization, so the loan amount was the same for IO and amortizing structures. Partial term IO structures were about boosting cash on cash returns in the early years of the deal. Here’s an example:

Most of us think of amortization as a small piece of the payment, but when rates are very low (like today) amortization is a very big expense component:

By deferring this expense for a few years a spreadsheet jockey could show a much better return in the early years. Combine that with rosy income projections in later years, and buying CRE at a 5% cap rate starts to look like a good idea.

I discussed IO in much more detail in my post “The Problem with Interest Only” last December. I don’t see a lot of defaults triggered solely by amortization kicking in on these deals – the easiest modification in the world is to extend the IO period, and I think we will see a lot of that going on. The underlying deterioration in cash flow and values is the much larger issue.

CRE Miniperm Underwriting Today and Yesterday

Deal Junkie suggests in this post that although the CMBS market is frozen, balance sheet lenders continue to lend on CRE. Traffic Court counters here that, although balance sheet lenders are making loans, the underwriting is much more conservative.

I took the current loan terms and underwriting parameters from one of the banks mentioned in the Deal Junkie post and compared them to the terms and underwriting on an actual deal done in 2006. The loan is a 5 year term with the first 3 years fixed. The bottom line, 17% fewer loan dollars. Here are the numbers:

Lender Groupthink

Here’s an excerpt from Michael Skapinker’s opinion piece in the Financial Times, “Diversity Fails to End Boardroom Groupthink”:

Disagreeing with the company’s direction is hard enough. Doing so when an entire industry is going in the same direction is harder still. It is not just boards that suffer from groupthink; entire sectors do. The banking industry did.

Any investment banking chief executive who had listened to a director’s warning that complex financial instruments spelt trouble would have been in trouble himself. As Peter Hahn, a fellow at Cass Business School, told the Treasury committee: “If one of those banks in 2005 decided to be more conservative and hold back in their activity, they more than likely would have had their CEO and board replaced in 2006 for failing to take advantage of the opportunities.”

The implication is that we should heed the advice of dissidents, but real life is not so simple. In the 1980’s a lender I worked for had losses in Las Vegas, and as a result of that experience and my general distrust of low constraint markets, I believed Las Vegas was a dangerous place to lend. Today, I’m right – lenders who made loans in Las Vegas after 2005 are going to take losses. But, I was wrong for 20 years. 12 month change in employment growth is a good proxy for the health of CRE in a market, and the chart below shows went went on in Vegas:

CRE loans in general went through an extended period of virtually no losses, and the lenders making speculative land development, condo, and aggressively underwritten loans enjoyed an extended run of success. In at least some cases more conservative lenders decided to join the party at the end, and are now paying the price.

I discuss how difficult it is for credit officers to go against the flow in the post below:

Fox Guarding the Henhouse: Bear Stearns Risk Manager Now at the Federal Reserve

The Commercial Real Estate Risk Culture at Deutsche Bank

Zero Hedge has published a letter from a former risk manager at Deutsche Bank which speaks to the difficulties of being a risk manager in a lending institution. Some excerpts:

For more than two years, I have been working internally to improve the inadequate governance structures and lax internal controls within Deutsche Bank. I joined the firm in 2006 in one of its foreign subsidiaries, and my due diligence revealed management failures as well as inconsistencies between our internal actions and our external statements.
Beginning in late 2006, my conclusions were disseminated internally on a number of occasions, and while not always eloquently stated, my concerns were honest. Unfortunately, raising concerns internally is like trying to clap with one hand. The firm retaliated, and this raises the question: Is it possible to question management’s performance without being marginalized, even when this marginalization might be a violation of law? Two years later, our mounting losses are gaining attention, and I offer my experiences and my thoughts in the hopes of contributing to the shareholder and public policy debate…

I joined Deutsche Bank in 2006 to build an investment business within its commercial real estate lending operation, and I was generally surprised by the aggressive sales culture within our firm. While many people consider the banking sector’s problems to be caused by residential lending, I witnessed multibillion-dollar loan proposals for commercial property.
With funds provided at more than 90 percent loan-to-value, these loans were “priced to perfection” and assumed that property prices and rental rates would continue to rise. For perspective, a single billion-dollar commercial real estate loan is equivalent to 2,000 residential loans of $500,000.
In general, my colleagues are hard-working, decent people, but the system of incentives encourages people to take risks. I have seen honest, high-integrity people lose themselves in this cowboy culture, because more risk-taking generally means better pay. Bizarrely, this risk comes with virtually no liability, and this system of O.P.M. (Other People’s Money) insures that the firm absorbs any losses from bad trades…

There’s much more at this follow up Zero Hedge post.

Related Post: Fox Guarding the Henhouse?  Bear Stearns Risk Manager Now at the Federal Reserve

Indications of a Credit Bubble

From Socializing Finance’s post Flashback: The Quality of Credit in Booms and Depressions, some commentary from 53 years ago:

In the past few years important new historical evidence has been developed on the cumulating deterioration in the quality of credit during the period of prosperity that precedes severe depression. […] With respect to the current situation we must concern ourselves with the fact that some, at least, of the economic conditions are in evidence today. What are these conditions? First and foremost is a rapid increase in the volume of credit or debt. Second, a rapid, speculative increase in the prices of the assets that are brought with the rapidly increasing credit, such as real estate, common stocks, or commodity inventories. Third, vigorous competition among leaders for new business. Fourth, relaxation of credit terms and lending standards. Fifth, a reduction in the risk premiums sought or obtained by lenders.” – Moore, G.H. (1956). The Quality of Credit in Booms and Depressions. Journal of Finance 11, 288-300.

How accurately did these conditions predict the current CRE bubble, and where are we today?

1. Rapid Increase in the Volume of Credit or Debt. This clearly occurred during the bubble. As of today, the amount of debt outstanding hasn’t really declined, because few assets have retraded at reduced value levels.

2. Rapid, Speculative Increase in the Price of Assets. Again, this obviously happened. Some distressed sales are starting to occur, but for the most part values have not been marked to market yet.

3. Vigorous Competition for New Business Among Lenders. That clearly went on. Today, there is very little competition occurring; the few lenders that are making loans can pick and choose.

4. Relaxation of Credit Terms and Lending Standards. Terms and lending standards were clearly relaxed during the bubble (Loan to Value, Debt Service Coverage, Interest Only payment structures, etc.). For the most part these standards have tightened, although arguably LTVs are still based on cap rates which are too low, and DSCs calculated on historically low interest rates may not be high enough to ensure an exit if rates return to historical averages.

5. Reduction in Risk Premiums. Again, this obviously occurred during the bubble, with spreads over Treasuries in the 100bp to 200bp range. Today, spreads are much wider, but again maybe not enough in light of the historically low Treasury rates.

So, it appears lenders in 2006 were not attuned to the risks publicized by this article 50 years earlier. And, it appears we are only part way to establishing a normal lending environment.

Seeing Patterns Where There Are None: Geography

Humans are wired to detect patterns, but sometimes there isn’t one. For example, what distinguishes the best and worst performing submarkets in Orange County?

You might focus on geography first; the real estate mantra is location, location, location. Are the best and worst performing submarkets concentrated in a particular area?

Here’s a map, with the five best performing markets (as measured by combined occupancy and rent change) highlighted in green, and the worst ones in red:

Looking at this, you would have to conclude there’s not a pattern; the best performing and the worst performing markets are pretty will mixed up.

The data is for the first quarter 2009 from RealFacts, as reported by Lansner on Real Estate.  Here’s the chart accompanying the story; can you find a pattern in the occupancy and rent changes?

Successful, Until You Aren’t

Lansner on Real Estate reports Pacific Property Assets has defaulted on the interest payment due on $90,000,000 in notes held by its investors. PPA has a 2,400 unit multifamily portfolio in Southern California and Arizona. Some excerpts:

Company CEO Michael Stewart said interest payments on about $90 million in notes would be suspended for an undetermined period, adding that he’s hoping investors will bear with the firm to give it “breathing room…”

“We’ve never missed a payment in over 10 years. It’s probably the toughest decision (we’ve made),” Stewart told the Register.

The fact that no payments were missed for ten years doesn’t mean much. I was Chief Credit Officer at ARCS Commercial Mortgage from 1997 to 2006, during which time we originated about $2B a year in multifamily loans with virtually no delinquencies, foreclosures, or losses. I would love to believe that was a result of my stellar judgment, and maybe it was. But, I’ll never know for sure, because during the time I was there any bad decisions I made were bailed out by declining interest and cap rates. Periodically, someone would complain we should do a risky deal I had turned down, because the fact we had no defaults indicated we weren’t taking enough risk. My response was that if the average CRE default rate was 2%, that was arrived at by 9 years of no defaults and one year of 20% defaults.

Commercial real estate performance, to paraphrase the quotes about airline travel and war, is years of boredom punctuated by periods of terror. If you’re a CRE investor who never missed a payment between 1995 and 2008, that puts you in the same class as 99% of all CRE investors. If you never missed a payment between 1979 and 1982 or between 1990 and 1994, I’m impressed. I expect 2009 to 2012 will be another period where never missing a payment will be something to brag about.

One of my grandmother’s sayings was “You don’t know if your roof leaks until it rains.” It hasn’t rained hard in the CRE world since the early 1990’s, and many lenders and owners (like Mr. Stewart) have assumed that, because they weren’t getting wet, they had a good roof.

Here’s a link to another story about Mr. Stewart during happier days just 8 months ago, in which he explains PPA’s decision to diversify into the Phoenix market (oops!), and how risks were lower in October 2008 than when he started PPA in 1999.

Roads Before Roofs, Roofs Before Retail

The stories and video of new houses being demolished in Victorville are continuing to pop up in blogs and other news sources (see Calculated Risk, the LA Times, and the Wall Street Journal, for example). It’s a compelling story, but the way it’s being presented almost everywhere is misleading.

First, here’s the video if you haven’t already seen it:

The video, and every story I’ve seen referencing it except one, gives the clear impression the bank thinks it makes economic sense to demolish completed and virtually completed but unsold houses because the market is so bad. However, the original source of the story (see this post) interviewed an officer at the bank, who makes clear the real issue is the homes were built before the roads and other site improvements were completed. Completed homes could have been sold at some price, but if there’s no road to the home you can’t sell it.

This is obviously bad construction lending practice; you should complete site improvements first (or make sure you’ve held back enough money to do so). Hence the headline, roads before roofs. This seems obvious, but it happens more often than you might think. When I was at Capmark a few years ago one of our workout deals was a project where we funded the equity portion of a purchase of a multifamily land parcel, and then discovered the access road we needed couldn’t be built because it would cross a stream which was the home of an endangered fish species. That investment was a total loss.

It’s also obvious it takes more than a few mistakes to bring down a lender, but when you have a major due diligence breakdown like this, you have to wonder if it’s not the tip of an iceberg of bad decisions. From the WSJ story linked above:

Guaranty Bank has significant exposure to construction loans to home builders. Last month, its parent company, Guaranty Financial Group, was issued a "cease and desist" order by the federal Office of Thrift Supervision, citing the firm's "unsafe and unsound banking practices."

I’ve previously posted about Capmark’s problems here. Since then, they reported a $1B loss in the first quarter.

The second part of the headline is roofs before retail. Before you develop a retail project, you want to make sure there are enough people living in the market area to support it. Because subdivisions were being developed at such a rapid rate, this rule was frequently violated, and when the music stopped on the residential side many retail projects were left without a customer base. Between the two retail sites indicated below, which do you think is doing better?

The lesson is, it’s important to develop in the right order; infrastructure, then residential, then retail.

Humans Are Wired to See Patterns Where There Are None

From Jonah Lehrer’s post on Frontal Cortex, Patterns and the Stock Market:

Alas, the human mind can't resist the allure of explanations, even if they make no sense. We're so eager to find correlations and causation that, when confronted with an inherently stochastic process - like the DJIA, or a slot machine - we invent factors to fixate on. The end result is a blinkered sort of overconfidence, in which we're convinced we've solved a system that has no solution.

Look, for example, at this elegant little experiment. A rat was put in a T-shaped maze with a few morsels of food placed on either the far right or left side of the enclosure. The placement of the food is randomly determined, but the dice is rigged: over the long run, the food was placed on the left side sixty per cent of the time. How did the rat respond? It quickly realized that the left side was more rewarding. As a result, it always went to the left, which resulted in a sixty percent success rate. The rat didn't strive for perfection. It didn't search for a Unified Theory of the T-shaped maze, or try to decipher the disorder. Instead, it accepted the inherent uncertainty of the reward and learned to settle for the best possible alternative.

The experiment was then repeated with Yale undergraduates. Unlike the rat, their swollen brains stubbornly searched for the elusive pattern that determined the placement of the reward. They made predictions and then tried to learn from their prediction errors. The problem was that there was nothing to predict: the randomness was real. Because the students refused to settle for a 60 percent success rate, they ended up with a 52 percent success rate. Although most of the students were convinced they were making progress towards identifying the underlying algorithm, they were actually being outsmarted by a rat.

Loan underwriters and credit officers are constantly searching for patterns that aren’t there. This is the first in a series of posts that will look at this problem.

I highly recommend Jonah’s book, How We Decide.

Problems Mounting in Orange County Multifamily

Lansner on Real Estate reports it’s taking twice as long to rent vacant units in Orange County, rents are falling, vacancies are rising, and landlords are looking the other way on tenant credit issues and cutting back on maintenance.

None of this is surprising; all these things go together in a softening market. But, it’s nice to see an article which puts all the symptoms of a soft market in one place. For more on the relationship between rents, vacancy, and turnover time, see Multifamily Occupancy Rates: Four Things to Think About. For a discussion of the nasty feedback loop cutting tenant credit standards and maintenance creates, see The Slippery Slope to Default.

Innovation versus Old School, Big Lenders versus Small Lenders

Matt Yglesias is not comfortable with how easy it was for him to get a mortgage, and talks about it in his post, “Financial Innovation Takes the Homework Out of Banking.” An excerpt:

An old-school local bank can expect the people supervising loan applications to have specific knowledge about situations. And perhaps more importantly, the head of a small institution can directly monitor what his subordinates are doing. And while he perhaps can’t have detailed information about everything that’s going on, he can have general knowledge of the local economic situation.

But when I got my mortgage from Bank of America, it’s not like there was some plausible worry that Ken Lewis was going to knock on the guy’s door unexpectedly and make sure that everything was being done right. You can’t really have a homework-based system at a giant institution. Things need to be handled through bureaucratic processes and rules and formulae.

I like the post, but what I really like is the quality of the comments, most of which are on point and contribute to the discussion. The comments include discussions of the expense of good due diligence, the role of rules in preventing discrimination, whether or not due diligence makes a difference in a severe recession, whether or not predicting someone’s employment prospects is possible, the scalability of underwriting supervision, the failure rate of large versus small lenders, the role of securitization. There are some of the usual “CRA/Fannie Mae/minority lending are to blame” commentators, but on the whole Matt’s got a good group of readers.

I have two contributions to the discussion. The first is that even if big lenders and small lenders want to do their homework, they don’t know what to study (see Why What You Know About Income Property Performance is Probably Wrong). My second observation is that by cutting underwriting steps the lender can reduce costs, decrease the time spent from application, and reduce uncertainty in delivery, all of which improves their competitive position (see Why Did WAMU Abandon Underwriting Standards?).

Maturity Kills: Operating Statement Defaults Versus Balance Sheet Defaults

No question CRE rents are falling and vacancy rates are rising, and these trends are getting a lot of attention (see, for example, Calculated Risk posts here, here, and here, and Zero Hedge posts here, and here). However, this threat is minor compared to what’s happening on the balance sheet side of the business.

There are two ways a CRE loan defaults; an operating statement default, or a balance sheet default. Here is a typical CRE deal illustrating an operating statement default:

The assumptions are an initial interest rate of LIBOR+2.25% with a 30 year amortization, no changes in interest rates or cap rates, but a 25% decline in NOI. This results in negative cash flow, which could lead to a default (one would hope on a $10,000,000 deal the sponsor could cover a shortfall this small, but that capability is not something CRE lenders focused on). The takeaway point is, even with a major decline in NOI the shortfall is not huge, and because there is equity on the balance sheet the problem can be solved with a sale of the property.

Here is an example of a balance sheet default with the same structure, but a smaller decline in NOI coupled with an increase in cap rates:

Note that the operating statement side of the equation is fine; the borrower can still make the payments. However, the increase in cap rates has wiped out the equity in the property, and if the loan matures the borrower can’t repay it. The takeaway here is that cap rate changes have a much bigger impact than operating statement changes (for a more thorough analysis of this point, here’s a link to Philip Conner’s and Youguo Liang’s Income and Cap Rate Effects on Property Appreciation).

Here is what things are actually looking like for 2010 – a substantial decline in NOI and an increase in cap rates, combined with a substantial decline in interest rates:

Note that the operating statement is fine; the decline in interest rates more than offsets the decline in NOI, and cash flow has actually improved since origination. However, the decline in NOI combined with the increase in cap rates creates a huge balance sheet problem, and if the loan matures the problem can’t be solved with a refinance or sale of the property.

This is why there is so much concern over upcoming loan maturities. Here’s a link to a Deutsche Bank CRE presentation which goes into more depth (the maturity discussion begins on page 25).

CRE Performance by Property Type: It’s the Tenants

Zero Hedge has a chart this morning showing CRE loan performance by property type. The information supports the idea that loan performance in a downturn is largely driven by tenant type, and specifically the term of the tenant lease.

Here’s the chart:

(Click on image for a larger version in a new window)

The Zero Hedge post has an attribution to Lehman, but not enough information to identify what exactly we’re looking at. About all I can say is it’s clearly CMBS data.

I’ve taken the data, excluded Credit Tenant Leases and Health Care (given the small balances involved, the performance of a few deals could skew the result), and sorted by worst-to-best performance:

Why does it stack up like this? In a market with declining fundamentals, the deterioration in cash flow is largely a result of tenant turnover (more on that here). When a tenant leaves, the new tenant comes in at the new, lower market rate, while the tenants still in occupancy continue to pay at the higher rate. The higher the turnover rate, the faster the reset to the lower market rate. Also, if there are fewer tenants looking for space, when a tenant leaves the vacated space stays vacant longer.

Hence the rankings above. Hotels, obviously, have the highest turnover. In most multifamily projects more than half the tenants move every year, while retail, industrial, manufactured housing, and office tenants move much less often.

The only real surprises here are full service hotels and self storage. Both are performing around 2% better than I would have predicted. The outlook for full service hotels in this recession is not good (see for example, this news release from PKF Consulting), and I would have expected performance more in line with other hotel types. The average self storage tenant rents space longer than you might think (15 months, according to this article), but I’m still surprised by how well that property type is performing.