Zombie Banks’ Distressed Assets

John Reeder’s post Distressed Assets Market and FDIC Closures on Real Property Alpha is a must read for those that want to understand what’s going on with regional banks. An excerpt:

Our business working in the commercial real estate industry (see the Deal Breaker site, or upcoming Sperry Van Ness auction) puts us on the front lines of the current blow-up that is going on in the banking industry.  Capitalization levels in financial institutions have a large impact on whether they are willing or able to dispose of distressed construction loans, commercial REO, or A&D loans.  The general rule of thumb is that the more distressed the bank, the less potential that you are going to be able to make a deal with that Bank on their non-performing assets.  It’s difficult to digest this reality as the potential that a distressed bank offers in the way of inventory can be enticing.    However, the chances are that the bank has not written down the value of the asset to real current market, so selling at today’s prices means that the bank has to take an additional hit to their capital and the really distressed banks can ill afford the additional hit.

Read the whole post, there’s much more. I have two small contributions to John’s points:

• Even if a bank conscientiously marks its bad assets to market, it will still probably incur smaller losses at any given point in time if it holds the asset instead of disposing it. The marks are based on appraisals less a discount for sales costs. This number will almost always be higher than what a bank actually realizes on a sale, because appraisal values tend to lag actual market trends (more on that in the Lansner on Real Estate post “Were Appraiser’s Late to the Price Collapse?”). So, a bank can adopt a hold strategy and still be in regulatory and accounting compliance. The risk, of course, is that by hanging on to the asset, the bank continues to be exposed to further value losses if the market continues to deteriorate, and may ultimately incur an even bigger loss.
• In most cases the management and staff working on the problem assets at the smaller banks are the same people who originated the deals. 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). The consequence is the management at these banks may genuinely believe these assets can be salvaged given time, while someone with less involvement would say it’s time to take the loss.

My point is that, while I am sure some banks are consciously manipulating their accounting, I am also sure many banks believe they are doing the right thing.

Valuing Note Purchases

There is a common misconception that lenders know the value of their CRE loans. For example, here’s an excerpt from a Naked Capitalism post (Yves is talking about the PPIP Legacy Loan program):

The problem isn't, contrary to PR designed to mislead the public, that the assets are hard to value. That holds only for an itty bitty percentage of the total. The real problem is that the banks are carrying them at above market values, and above any reasonable long term value too (their protests to the contrary). The problem is not the saleabilty of said assets, it's that they don't like the prices.

I can tell you with absolute certainty that lenders do not know the value of the CRE debt they are holding – there is just too much uncertainty around the key variables. Here are some of the problems:

Uncertainty around the current value of the underlying collateral. CRE is always a thinly traded market, and that is especially true now (more on this topic at Why There Are Very Few CRE Sales). To state the obvious, when there are few sales it’s difficult to establish values.

Uncertainty around the future value of the underlying collateral. Legacy debt service is a big component of the cost structure of existing CRE. When CRE is foreclosed and sold, the debt burden of the new owner will be much lighter, enabling them to cut rents and attract the best tenants (think Detroit automakers with huge pension obligations trying to compete with manufacturers that don’t have this burden). When this happens more of the old deals default and are sold as REO, which creates additional rent reductions, and so on in a vicious spiral down (more on this at CRE Loans and the Death Spiral of Doom). Even if you think you have a good handle on the current collateral value, there is no way to predict how far down this spiral will drive values.

Uncertainty over borrower actions. Apart from the collateral value issues, there is the fact that until you own the real estate the current borrower is still a factor to be dealt with. Depending on the jurisdiction, it can literally take years to get control of a property. And, given today’s low interest rate environment there is unprecedented risk that the debt will be restructured in bankruptcy at a very low interest rate (more on that at Getting Tilled: How a $6,425 Truck Loan May Decide the Fate of General Growth Properties).

All these are uncertainties affecting the current lender. Now imagine the position of the note buyer. REIT Wrecks describes the note purchase due diligence process in “What Hypocrisy? FDIC Loan Sales are a Total Black Hole”. An excerpt:

So what happens when you bid on one of these loans? The FDIC does not allow property inspections of any sort. Buyers are afforded the opportunity to review the original loan files, which contain such helpful information as the original, hopelessly out of date appraisal. Assuming you have enough local market knowledge to formulate a bid and "win", you'll have just 7 days to close. There is no futzing around with surveys, title reports and good standing opinions - we're talking an all-cash close on a 7-day fuse.

Arguably non-FDIC note sales afford better due diligence opportunities. However, I can tell you from experience it is very difficult to pick up an unfamiliar loan file and figure out the deal and it’s current status in the best of circumstances. Imagine trying to do that for a pool of deals, with limited due diligence time, no access to the borrower, and probably only partial access to the files.

It’s no wonder there are few note sales going on given the difficulties of establishing value.

Home Court Advantage in Real Estate

Home court advantage is a huge factor in sports; for example, historically the home team in deciding games has won 78 of 97 games up until the second round of the 2007 NBA Playoffs. There is a comparable effect in commercial real estate.

I learned about real estate home court advantage from Gus Williams, the Seattle-based basketball star that led the Sonics to their 1979 championship. Somehow Gus ended up as the primary investor in a strip retail center in Selma, California. Selma is a town about 20 miles south of Fresno on Highway 99. You’re probably heard of tertiary markets – Selma is a quaternary, or maybe even a quinary market. I’m not sure how Gus’s money got into the deal, but I can tell you it never got out, because the Los Angeles lender I worked for foreclosed on the center in the early 1990’s.

It’s not noteworthy when a professional athlete loses money in real estate. What distinguished this piece of REO was that fact that absolutely no one would buy it. Months passed, the listing price was reduced again and again, but nothing. Finally, the local businessman who sold the property to Gus came forward and put us out of our misery with an offer which was a small fraction of what he got from Gus five years before. We (and Gus) were the away team, and the home team blew us out.

Local investors are starting to step up this time around too. From Zero Hedge:

The Buffalo News reports that REIT Developers Diversified Realty is selling back 11 upstate New York shopping malls to the entity it originally purchased them from, Benderson Development Co., at a 30% discount to their 2004 purchase price…“It’s good that the ownership is going in the direction that it is,” said Michael C. Clark, director of retail tenant services at CB Richard Ellis in Buffalo. “There’s going to be a lot of markets in other parts of the country where they have portfolios for sale by different REITs and they don’t have someone like Benderson to step up.
“We’re pretty fortunate in terms of the market, in regard to that. How much better can you get than the folks that developed them and are intimately familiar with them and live and breathe here? They certainly know what they’re doing,” Clark said.

The Zero Hedge spin is that CRE values have fallen, but that misses the real point of the story – a REIT based in Ohio is not going to do a good job pricing and operating malls in upstate New York.

Another example is from the Portland Oregonian, via Portland Housing Blog:

Portland condo king Homer Williams is pursuing a surprising new business.

With the residential real estate market struggling, Williams has turned to a newly hot commodity: failed bank loans.

Williams confirmed that he's the man behind BCC Fund I Limited Partnership, which the FDIC identified this week as the successful bidder for two packages of loans from the defunct Bank of Clark County.

The FDIC auctioned the loans last month from the Vancouver bank that failed in January.

Williams declined further comment. But according to the FDIC, BCC Fund 1 paid just more than $2 million for one bunch of loans with an outstanding balance of $6.1 million. BCC also successfully bid $3.3 million for a group of 53 other loans with an outstanding balance of $10.3 million.

That means BCC paid about a third of the outstanding balance of the loans.

Buying a loan from the FDIC is buying a pig in a poke (REIT Wrecks has a great post on that here), but I have to believe a Portland developer buying loans from a failed Portland bank is going to do better than a hedge fund out of New York.

Moral of the stories: keep the home court advantage.

Legacy Loan PPIP versus FDIC Note Sales

It’s difficult to work up much enthusiasm for the Legacy Loan segment of PPIP, a program which will reduce losses for banks by goosing returns for private investors with low cost public leverage. Most people (other than the banks themselves) think banks should be punished with big losses, and most people are not keen on helping the investors who will get richer as a result of the mess do even better. I totally get that. However, I think it’s important to point out that the most commonly expressed alternative to PPIP (just let the banks fail and let the FDIC clean up the mess) will be tremendously expensive to taxpayers.

The argument against PPIP is cogently summarized in this Naked Capitalism post. An excerpt:

As readers may recall, we had been skeptical (and critical) of the Public Private Investment Partnership from the outset. It was the third effort at a program that had failed twice under Hank Paulson, namely, to have banks get dud assets off their balance sheets by selling them to a sucker.
That's why this program has never gotten airborne. It requires a bagholder.
The problem isn't, contrary to PR designed to mislead the public, that the assets are hard to value. That holds only for an itty bitty percentage of the total. The real problem is that the banks are carrying them at above market values, and above any reasonable long term value too (their protests to the contrary). The problem is not the saleabilty of said assets, it's that they don't like the prices. Selling them at below the marked value leads to losses, which in turn would reduce their equity at a time when they have been told, in no uncertain terms, to get more.
So the only way the plan works is if someone overpays. The only party that might have reason to is Uncle Sam. The whole point of the "public private investment" part of this is to disguise the overpayment. So the plan is an opaque subsidy to the banks.

Yes this program is a subsidy to banks. It’s not even opaque; it’s transparent to anyone with a spreadsheet. But it’s wrong to say “the only way the plan works is if someone overpays”. The plan works because someone will pay more if an investment can be leveraged with low cost funds. Imagine a housing market with no mortgage debt; fewer houses would sell, and they would sell for much less.

Here is an example I used in my post, Investor Returns on FDIC Discounted Notes. Let’s say this is a subperforming CRE loan which is still making payments:

Here is the same note sale under PPIP:

Note the low cost leverage allows the bank to get a better price (85) and the investor to get a better yield (12% versus 19%). To state the obvious, more banks will sell assets at 85 than 50, and more investors will invest if they can get 19% instead of 12%. Also, assuming a finite amount of investor money, it will go a lot further with the PPIP program (in this example, $5,000,000 without PPIP, $850,000 with the program). If the loan continues to perform and pays off, the investor is the big winner; they get their yield from the payments, and a nice pop when the loan is repaid at par. But, the Treasury wins too, because under PPIP the Treasury is the 50% equity partner.

Of course, the loan may not perform. If after liquidation costs the underlying collateral value is more than the purchase price, the equity investor will still get a return and the FDIC will get its PPIP loan repaid. If the recovery is less than 80% of the discounted purchase price, the equity is wiped out and the FDIC takes the remaining loss on its PPIP loan. The check against this happening is the fact that the private part of the equity does not want to lose its money. It could happen, but absent collusion with the loan sellers there’s no reason why private equity would intentionally overbid. Avoiding collusion is extremely important. Option Armegeddon gives a good explanation of the risk in this post. However, I think this concern is manageable as long as regulators follow the money trail and severely penalize infractions.

Assuming investors don’t overbid, the only “loser” in this scenario is the FDIC, which only collects a 4% interest rate. Is making this loan the best use of FDIC funding capability? Maybe not, but making too small a return is a lot different than characterizing the FDIC as a bagholder. And, consider the alternative; if the bank fails and the FDIC is the note seller at 50 in the first example, that’s a $3,500,000 loss to the taxpayer, versus a 4% return on a PPIP $6,800,000 loan. If you think FDIC loan sales are the best way to maximize value for the taxpayer, this excellent post from REIT Wrecks will open your eyes.

PPIP is not easy to love, but I’ve not seen a better alternative. If you’re not familiar with the PPIP program see a description here.

Investor Returns on FDIC Discounted Note Purchases

John Reeder over at Real Property Alpha and I have a discussion going on the implications of the FDIC selling performing CRE loans at 50 cents on the dollar (see John’s posts here, here and here, and mine here). One of John’s commenters asks about cap rates on these transactions, which I’m taking as an opportunity to show how these deals can be such a home run for the note buyer.

I’ve expanded my previous example to show net operating income, cap rate, and collateral value. There’s not enough information in the FDIC sales info to use an actual example, so I’ll show a couple of hypotheticals. The first example is a marginally performing loan where the property generates just enough income to cover the interest payment:

I’ve picked an 8% cap rate – the real life non-distressed transactions I’m seeing have cap rates in the 7%s for multifamily on up to the 9%s for retail. At this cap rate the property has negative equity, so at maturity the borrower will presumably be unable to pay off the loan and the note holder will foreclose. Here are the numbers for a note buyer who bought at a 50% discount:

Even though there’s a $2,500,000 loss on the $10,000,000 loan, since the note buyer only paid $5,000,000 for the note they are up $2,500,000 on their position.

Let’s say the original loan is performing well and the NOI is more than the interest payments:

In this example, the borrower has equity and will be able to pay the loan off by selling the collateral. That works out even better for the discounted note buyer, because even though they only paid $5,000,000 to buy the note they get repaid the whole $10,000,000 (100% return).

This kind of apparent no lose proposition is what gets people very excited about buying notes at a discount. There are risks, of course; property income might deteriorate, cap rates might continue to increase, or the borrower might file bankruptcy and get the loan restructured on terms less favorable to the note holder. That’s why note investors are looking for a relatively high yield and a big discount going into the transaction; a 50% discount gives a lot of room for things to go wrong and still get an acceptable return. However, selling at a 50% discount is a huge hit for a bank to take (I’ve posted about the impact on the bank’s balance sheet here), so as far as I know the FDIC is the only active seller that is discounting to this extent.

Why Are Performing CRE Loans Selling for 50 Cents on the Dollar?

Zero Hedge and Real Property Alpha have picked up on the results of recent FDIC auctions of CRE loans (Zero Hedge posts here and here, Real Property Alpha posts here and here). A graph from Real Property Alpha shows performing CRE loans are being sold at roughly 50% discounts:

You might think the sales price on the performing loans indicates the collateral backing the loan is worth only half the loan amount, but that’s not the case. This is about a change in investor yield requirements, not CRE fundamentals.

Let’s say you have a well secured, performing $10,000,000 CRE loan paying a 6% interest rate:

Now, let’s say you are taken over by the FDIC, and the FDIC wants to sell the loan. You might think that since the loan is well secured you could sell it for the full principal amount, but you would be wrong; note buyers want a 12% yield on their investment (actually, they want more – I get two or three calls a day from people wanting to buy notes, and return requirements are 12% to 25%). To get a 12% yield on a loan paying 6% interest, you need to buy it at a 50% discount:

You might question why the FDIC would sell – 6% is not a bad yield when 5 year Treasuries are at 2%. If the answer is the same as when I worked there in the late 1980’s, it’s because their job is to liquidate assets at the best price they can get for them. But, most banks would be content to collect 6%, and that explains why you don’t see many banks selling performing CRE notes.

How Big a Hit Can Lenders Take on Note Sales?

I’ve previously posted on how driving away borrowers can leave a bank in a better position to handle losses on the remaining portfolio (link here). Here is the simplified balance sheet side of the math:

In this example, $250,000,000 in loans are paid off and used to reduce liabilities. The loss reserve and equity are unchanged, but have increased in size relative to the remaining portfolio, so the bank is in a better position to absorb losses in that portfolio.

This suggests that a bank could sell loans at a discount without damaging its ability to deal with future losses. Here is the same transaction above, but the bank sells the loans at an 11% discount:

Note that there is an actual loss of $27,500,000 which needs to come from somewhere in order to pay off the liabilities. In this example, it comes from cash and a reduction in the cash held in the loss reserve (but still maintaining a reserve level of 2% of remaining loans). Even though they took an 11% hit, the banks ability to weather additional losses remains unchanged. However, note the bank’s cash position has declined substantially.

What happens when the discount is 22%?

The bank is in a worse position, and has wiped out it’s cash position.

The real world is obviously much more complicated, but the rule of thumb is a bank can take a 10% hit on a note sale without much pain because the capital and loss reserves are already on the balance sheet to handle the loss. In general, as the market deteriorates banks have been building reserve levels, and specific loss reserves are being taken against some assets. To the extent these reserves exist, bigger discounts can be taken.

According to Zero Hedge, the FDIC commercial loan auctions are clearing at a 50% discount. For a bank to take that kind of hit on a note sale of any size, they would need to have built up very large reserves, or have substantial excess capital, or both. There aren’t many (any?) banks with substantial CRE exposure in that position, hence there are not a lot of note sales going on.

This Time is Very Different: Attack of the Zombie Properties

The last time we had a severe CRE downturn was 1990 – 1995. For those of us who were around, the current situation feels similar – plummeting employment, deteriorating income fundamentals, spiking cap rates, and loss of liquidity in the market. However, there are some huge differences this time which have important implications.

First, some history. Here is a chart of cap rates taken from a paper by Philip Conner and Youguo Liang (Income and Cap Rate Effects on Property Appreciation, worth checking out):

Current value cap rates bottomed at around 6.7% in 1990, were around 8.25% in 1992, and peaked at around 9.5% in 1995. Based on the sales and appraisals I’m seeing and talk with colleagues, current cap rates seem to be in the 8% to 8.5% range, so today is somewhere around 1992 levels.

Now, let’s consider interest rates. A typical variable rate CRE deal in 1990 used an 11th District Cost of Funds index (COFI) plus 2.25%. An equivalent CRE deal in 2007 would have been priced at 30 day LIBOR + 2%. Here is how the interest rate would have changed on those two deals over the last 2 years:

Interest rates this time are much lower. In 1992, the cap rates were right around the interest rate, which meant a property with no equity also probably couldn’t make it’s payment. Today is much different; cap rates are 5.5% to 6% higher than the interest rate. This means a property could be severely under water and still make it’s payment. Here’s an example:

In an ordinary world, a property overleveraged to this extent would be foreclosed on and sold, but because interest rates are so low it can continue to make its payments.

What are the implications?

• CRE loans are collateral based, so under FAS 114 the bank probably needs to recognize the loss even though the loan payments are current. If the loan term is long enough, it’s possible the bank can make an argument the value will recover, and avoid recognizing the loss. But regulators and accountants these days tend to be pessimistic in their outlook, so the bank is probably stuck with recognizing the loss.
• If a bank attempts to foreclose on a basis other than a payment default (for example, loan maturity or a non-monetary covenant violation), the borrower will probably file bankruptcy. It is very difficult to obtain relief from stay and foreclose on a borrower willing to make their contractual interest payments (more on that here). So, the bank is probably stuck with the deal until interest rates go up and there is a payment default, unless they sell the note.
• If the bank sells the note for the collateral value, the return to the note purchaser is equal to the cap rate (in the example above, 8.25%). Note buyers are looking for returns in the 20% range, so these deals won’t appeal to them either.

I believe the result is we will have a lot of zombie loans on bank books, and a lot of zombie properties that are grossly overleveraged, but which can’t be cleared to market values because the borrowers can make the payments at today’s incredibly low rates.

Why Aren’t Banks Selling More Distressed CRE Debt?

People who are trying to buy distressed CRE debt tell me banks aren’t willing to sell at prices which will clear the market. Why?

In a New York Times piece, Casey Mulligan argues banks anticipated a government program to subsidize sales, and have held back waiting for it. An excerpt:

[…The] secondary market for legacy mortgages has stagnated largely because of the (ultimately correct) anticipation of a huge government subsidy. Banks were not “unable” to sell their legacy mortgages; they were prudently unwilling to sell because they expected the government to eventually step in and help push the prices of those assets higher.

We all witnessed last week the big capital gains to banks that came with the unveiling of the Geithner plan. A bank would have been foolish to sell off its legacy mortgages during the fall or winter, before such a plan was unveiled and executed, because a fall or winter non-bank buyer of legacy mortgages would likely be ineligible for the ultimate subsidy.

Thus, the secondary market for legacy mortgages has failed so far because of the lack of a plan rather than a lack of clarity. To get the market operating again, the Geithner plan does not need to alleviate the market weakness improperly identified by its authors, but needs only to stay on the path to execution.

The subsidy Mulligan is referring to is the PPIF program. David Kotok of Cumberland Advisors lays out the clearest explanation I’ve seen on how the program boosts prices and reduces buyer risk here (it takes eight minutes to read, but it’s well worth it if you’re interested in this topic).

I believe Kotok’s example overstates the value of the subsidy because the “win” side of the bet is too high.  A more realistic example is provided by the example from “a hawkeyed reader who embellishes the math”, about three quarters of the way down the post. Even with this example, the price support provided by PIFF is a big boost.

So will PIFF free up the market? I think it will definitely help, but there are still three very large issues. The first is that many banks are still hoping for the best on their loans and will hold back. This position will be harder to sustain if CRE continues to deteriorate, but it may take some time. The second issue is that some banks will be unwilling to take the hit required even at subsidized price levels because it will put them out of business. Waiting and hoping for a turnaround may be the only survival strategy for some banks. The final issue is that, even with the PIFF subsidy, buyers will hold back because they believe CRE still has a long way to fall. This view is succinctly summarized by the “expert in a rating agency” quoted in the Kotok post:

 

CMBS prices are terrible, but underlying asset prices are soon to follow, so prices reflect collateral, not liquidity discount.

I think this will be a real problem. I’ve posted on how CRE prices tend to spiral down here.

Workouts 101: Hold'em or Fold'em?

Income property workout people have been idling for many years now, but it looks like those days are over. This seems an opportune time for a series of posts outlining what I learned the last time around (1989-1994).

I will be laying these rules out in a series of binary choices - one or the other, true or false, yes or no. Obviously, there are a lot of moving parts to every income property workout, and it is tempting to try to weigh all the factors which could influence the outcome (formally, this is a Bayesian approach). The binary approach seems simplistic, but there is compelling evidence it can lead to decisions almost as good as more complex rule systems, and it's much simpler (important when you're up to your waist in alligators). If you want to delve into this topic further, a couple good books are Simple Heuristics That Make Us Smart by Gerd Gigerenzer and The Either/Or Investor by Clark Winter.

So, Hold'em or Fold'em? Although this decision applies at the individual loan level too, first the decision needs to be made at the portfolio level, i.e., do I liquidate the portfolio as quickly as possible (through a portfolio note sale, for example), or try to maximize value working out individual loans? There is no right answer - subsequent events which are unknowable at the time of this decision will determine whether or not the right decision was made, and even later the answer may not be clear. Here are some implications of an immediate liquidation:

• Liquidation will almost certainly result in a higher immediate loss than holding the portfolio. The buyer is going to make a determination of the value to be realized from working out the portfolio, and this is almost certainly going to be less than the value a lender might reasonably justify to a regulator. For example, a lender might mark a loan to the value of the underlying collateral, but a buyer of the distressed note will start with the value of the collateral and then deduct a further haircut for the risk and time required to realize that value.
  
• Since an immediate loss is involved, you probably wouldn't liquidate the portfolio unless you believed the market was going to continue to fall for a material time period. If the market turns around shortly after the sale, the buyer will have a windfall and the seller will have an unnecessary loss and will look stupid in the bargain.
  
• It is much easier to make the decision to dump a portfolio if you were not around when it was originated. Prior involvement creates all kinds of biases which tend to keep people in situations once they've committed to them (see endowment effect, post-purchase rationalization, status quo bias, sunk cost effects, loss aversion, optimism bias, and valence effects). If you're the guy who is brought in to clean up the mess, it's much easier to attribute the problem (and the loss you take from the liquidation) to the old regime and move on.
  
• Frequently if your organization is publicly traded the market has already built in the full loss (and maybe more) into your stock price, and a liquidation will actually improve the value of your stock (see, for example, SL Green's divestiture of their interest in Gramercy). This has some interesting implications if you are working out debt with a publicly traded borrower. You might assume they are interested in maximizing value, while they might see more value in dumping the collateral and disassociating themselves from the problem.

To summarize, although liquidation might result in a large immediate loss, there are some compelling reasons to consider it. This is especially true if it looks like the downturn is going to be protracted (as it does this time around). So why aren't many lenders liquidating their positions now? I think at this stage it's because there have not been senior management replacements at many institutions, and because many institutions cannot afford the hit.