Why Banks Won’t Sell Legacy Assets Without Changes to PPIP

On March 23, the Treasury Department, the Federal Deposit Insurance Corporation, and the Federal Reserve announced the details of the government’s long-awaited Public-Private Investment Program. PPIP, as it has since come to be known, was proposed as a way to help banks rid themselves of their troubled real estate-related loans and securities by selling them to public-private investment funds or PPIFs. 

The thinking behind PPIP was that the large overhang of hard-to-value legacy assets on banks’ balance sheets was largely responsible for an unprecedented contraction in bank lending and had caused a loss of confidence in banks’ financial conditions, essentially closing the capital markets to them. By matching private investments dollar for dollar and providing favorable nonrecourse financing terms for the PPIFs that would purchase banks’ legacy assets through an auction process, PPIP was seen as a way to unclog selling banks’ balance sheets and narrow the gap between prevailing market prices and the assets’ higher intrinsic values, thereby increasing the value of legacy assets remaining on banks’ books, improving banks’ financial health and getting the flow of credit to the economy moving again. 

Many commentators have since voiced support for PPIP as being directionally correct. However, PPIP seems to have more than its fair share of critics, cynics and skeptics. One group argues that the government’s proposal, which would let private investors contribute equity equal to as little as eight percent of the purchase price for a pool of legacy assets but receive 50 percent of any profits made by the PPIFs, gives private investors an overly generous subsidy that will lead to excessive gains at taxpayers’ expense. A second set of observers believes this subsidy will cause private investors to overbid for banks’ legacy assets, thus exposing taxpayers to the possibility of significant losses. And, finally, some experts are predicting that private investors’ “unjust” enrichment will result in a political backlash – think “AIG bonuses” – that could lead Congress to retroactively impose TARP-like conditions and much less favorable deal terms on private investors. 

All of these mostly “buy side” concerns have some merit, deserve serious consideration, and may warrant modifications to PPIP before the first legacy asset auctions take place. To a large degree, however, they ignore a much more fundamental “sell side” question that will determine whether or not PPIP will be a success: Will banks with legacy assets be willing to sell them at the prices that, through the auction process, PPIFs will be offering to pay? 

Some observers say that this is not a serious concern because, if anything, investors are likely to pay more than fair market value for banks’ legacy assets due to the generous government subsidies they will be receiving. Others predict that PPIP will not suffer from any shortage of product since regulators will “lean on” – think Bank of America being pressured to close on its acquisition of Merrill Lynch – banks with large volumes of legacy assets to cleanse their balance sheets regardless of the sales prices offered, even though government officials have said repeatedly that banks will not be required to accept the highest bids in the upcoming PPIP auctions. Still others predict that banks will be willing to dispose of their legacy assets, regardless of the prices offered, just to put their bad asset problems behind them. 

Unfortunately, none of these arguments withstands any serious scrutiny. 

As an initial matter, the proposed investment structures of the PPIFs make it highly unlikely that private investors will overpay. With up to four-to-one or six-to-one leverage provided by the government’s debt financing, both gains and losses to the investors will be greatly magnified. Also, any tendency to overpay because of the government’s “subsidy” will be more than offset by the fact that, as equity holders with Treasury, the investors will be in a first loss position in the funds. As such, if they pay too much for banks’ legacy assets, they will run a serious risk of losing their entire investments. Accordingly, while investors’ asset bids may be somewhat above currently depressed market prices, their bids will almost certainly be at a discount relative to the legacy loans’ and securities’ perceived “intrinsic” values. 

It is also doubtful that banks’ regulators will force the banks to sell their legacy assets at suboptimal prices. Regulators are likely to “recommend” to some banks that they participate in PPIP by offering up all or a portion of their troubled assets for bid, but it is hard to see how banks could or would be required to accept any bid no matter how low the price, especially if the bank would incur a significant loss on the sale. In short, while the regulators may be able to lead their regulated banks to water, they won’t force them to drink. 

Will banks sell their legacy assets at below book or for less than their intrinsic values just to unclog their balance sheets? Given how troubled assets are accounted for, the accelerated loss recognition and capital holes that would result from most troubled asset sales, the bids they are likely to receive from PPIFs, and the ability of banks to do better in most cases by just standing pat, PPIP’s chances of success are, at best, uncertain. The following somewhat over-simplified troubled loan and securities examples illustrate these points. 

Legacy Loans 

As those familiar with bank accounting practices are already aware, a bank accrues for its projected loan losses by establishing a reserve or an “allowance for loan and lease losses” (ALLL), which is the bank’s best estimate the total amount of loan losses it is likely to incur over the ensuing 12 months. Actual loan losses are “charged off” against the bank’s ALLL as they occur, and the ALLL is replenished, usually on a quarterly or monthly basis, by the bank’s taking of provisions, so that at all times the ALLL should be sufficient to absorb one year’s worth of loan charge offs. 

In determining a bank’s earnings or losses, only the provisions that it takes periodically are an expense deducted from earnings. Loan charge offs, when they occur, are deducted from the ALLL. As such, they have no income statement impact. In practice, however, most banks prefer that their provision expenses at least equal their loan charge offs. Otherwise, if the bank’s charge offs exceed its provisions, it might seem that the bank is understating its loan loss-related expenses and is overstating its income. 

How, then, is a loss on a particular loan accounted for? In this example, let’s say that the bank expects that four percent of its aggregate loan balances will be charged off as losses. Thus, if the bank had a $100 million loan portfolio, it will have already established an ALLL equal to $4 million and will have taken a provisions charge against earnings of $4 million.

HFI vs. HFS Accounting Treatment 

The accounting treatment for an individual loan will depend on whether the loan is “held for investment” (HFI) or “held for sale” (HFS). If the bank satisfies the conditions required to demonstrate its intent to keep the loan in its portfolio as a longer-term assets (HFI), it is able to avoid immediate recognizing any losses even though the loan may have declined in value. Conversely, if the bank plans in the near term to sell the loan (HFS), it must periodically mark the loan to market, with any diminution in the value of the loan being charged to earnings. As one can easily imagine, banks prefer that their loans be classified as HFI rather than HFS. 

In the case of an HFI mortgage loan with an original principal amount of $100,000, the loan will already be treated as if it had incurred a loss of $4,000 since the bank’s $4 million provisions charge is usually spread equally across the bank’s entire loan portfolio. So the loan will be carried on the bank’s books at $96,000. As payments on the loan become past due, the loan will be treated as “nonperforming”, usually after 90 days, at which point the bank may no longer accrue interest on the loan and must reverse previous interest accruals. 

Only after a mortgage loan becomes past due by 180 days or more (or the borrower files for bankruptcy, whichever occurs first) is the bank required to mark the value of the loan on its books to the lower of the loan’s cost on its books or the estimated fair market value (LOCOM) of the underlying residential property at the time. The difference between the book value of the loan ($96,000) and the fair market value of the property is charged off against the bank’s ALLL and will reduce the carrying value of the loan accordingly. As noted above, this charge off has no immediate P&L impact; however, in practice, the bank will most likely want to take an offsetting provision charge against its earnings to maintain the ALLL’s prior balance. 

Assuming that the loan is not modified and it is necessary for the bank to foreclose on the loan and take possession of the mortgaged property as “real estate owned” or REO, the property will thereafter be subject to periodic reappraisals. Subsequent adjustments to the value of the property will be reflected either in the Bank’s ALLL or as adjustments to the bank’s income depending upon the holding period threshold established by the bank. 

PPIP Implications of HFI Accounting 

What are the PPIP implications of the above accounting treatment for HFI loans? Assume again, that private investors are more likely to submit bids for pools of legacy loans that are at some discount to the loans’ perceived fair market values than they are to overpay for the loans because the investors will want to maximize their chances of making a profit when they sell the loans or underlying REO properties to other purchasers. 

For troubled loans that are not yet 180 days or more past due, banks will face a difficult threshold issue of having to decide whether their solicitation of bids from PPIFs for their legacy loan pools will require a change in the status (and accounting treatment) for the loans from HFI to HFS. If so, then the net result of a bank’s submitting a pool of loans less than 180 days past due will potentially be to accelerate its own loss recognition on the loans from an immediate loan write down that will negatively affect current period earnings. (In the case of our hypothetical $100,000 loan, this expense will almost always be greater than the $4,000 loss allocation against the loan that would be released or reallocated.) 

Similarly, if a PPIF’s bid on a bank’s pool of under-180-days-past-due loans is accepted and the loans are then disposed of for less than their carrying value, the bank will be required to take an immediate hit to earnings in the form of a loss on sale. Since most banks are hoping that a combination of capital and earnings from other operations will be able to absorb their loan losses, they will be looking to spread out the financial impact of loan losses over a longer period of time rather than accelerate any loss recognition. Selling troubled HFI loans, which are still held on banks’ books at or near cost, at a much-reduced price will not be an attractive option for them. 

What about over 180 day delinquent legacy loans (and REO) that have already been marked to market? What financial or other incentive might a bank have to sell those loans that have already been either written down or essentially charged off and on which the bank has already taken its earnings lumps? 

Three possible scenarios come to mind where the bank may wish to sell loans and property that have already been marked to their current (lower) values: (1) the bank doesn’t want to deal with the “hassle factor” of continuing to own, hold and manage the charged off loans and REO; (2) the loans and property are worth more to the third party investor because the investor has either better past due loan collection or OREO property management and sales capabilities than the bank; and/or (3) in the case of REO, the bank thinks property values will continue to fall, and it wants to cut its losses while it still can. 

Of these three explanations, the third is the most plausible. But the first two reasons beg the question: Wouldn’t it make more sense in both situations for the bank to continue to hold the loans and REO and simply outsource any loan collection or property management and sales to a third party that specializes in these activities in exchange for some split of the profits? At least the bank would be able to share in any upside potential - funds that might be collected from borrowers who have defaulted on their loans and any profits that might result from REO sales. 

Legacy Securities 

Similarly, how will the accounting treatment for banks’ legacy asset-backed securities influence banks’ decisions on whether to participate in PPIP as sellers? Once again, a little accounting background for the uninitiated may be in order. As in the case of loans, troubled securities fall into different categories, each having its own accounting treatment. (For present purposes, I will not discuss securities that are “held to maturity” (HTM), as I am told by my accounting friends that it is very difficult for an owner of legacy securities to meet all of the requirements for HTM status, and that as a result, banks have very few securities on their balance sheets that are so classified.) The two main categories are “trading” and “available for sale” (AFS). 

Securities held in a bank’s trading account are regularly marked to market, with any gains and losses reflected in the bank’s income statement. As in the case of loans more than 180 days past due, a bank’s losses on its trading securities are already reflected in the bank’s earnings and capital. Consequently, it would not make a great deal of sense for a bank to sell these securities at a discount to fair market value or otherwise unless (1) the bank just wants to tell regulators and the market that it no longer holds any toxic securities on its balance sheet, (2) the bank thinks that the worst is not yet over and that prices for these securities will decline further (due to rising unemployment or additional housing price declines); or (3) the potential private investor has a longer holding period time horizon within which to realize the securities’ higher intrinsic values. 

A near certainty that prices will decline further below their already depressed levels or the bank’s inability to hold the securities for a sufficiently long period of time in order for them to recover in value may be two legitimate reasons why a bank would want to sell its trading legacy securities. However, a bank having legacy securities on its balance sheet per se is no longer the kiss of death that it used to be. If a bank is able to come through its regulator’s stress tests with flying colors notwithstanding its legacy securities, there is no regulatory or market imperative for it to sell. Also, banks have available to them various hedging techniques that they can use to limit their exposure to further price declines without having to close out of their securities positions. 

Available for Sale Securities – Temporary  vs. Other Than Temporary Impairments 

The largest accounting category by far for most legacy asset-backed securities held by banks is AFS, and it is for this group of debt securities that the Financial Accounting Standards Board’s FAS 115-2, Recognition and Presentation of Other-Than-Temporary Impairments [OTTI], issued on April 7, 2009, will have the greatest impact on banks and their interest in participating in PPIP. 

Again, by way of accounting background, when an AFS security declines in value, a bank must make an initial assessment of whether or not the “impairment” is temporary due to market fluctuations or otherwise. If it is temporary, and the bank has the ability and intent to hold the security until the security recovers in value, the decline in price/value is charged to “other comprehensive income (loss)” or OCI. For banks, this treatment is one of the closest things to accounting nirvana. The price drop is not reflected in the bank’s income statement; it only reduces the bank’s equity capital for GAAP purposes. Moreover, when calculating the bank’s regulatory capital, OCI is added back in. The net result is that the price decline in a temporarily impaired security will have no adverse effect on either a bank’s earnings or its regulatory capital. 

Under the prior accounting rules, once it became apparent to the bank that a debt security’s price decline was due to an OTTI, the bank was required to deduct the entire difference between the security’s amortized cost and its lower fair value from both earnings and capital, reducing both. However, going forward, FAS 115-2 provides for a much more favorable accounting treatment for legacy debt securities that have experienced OTTI. If the bank can demonstrate that (1) it has no present intent to sell the a particular security in question and (2) it will not need to sell the security before it recovers in value, it will only need to recognize for income statement and capital purposes that portion of the security’s decline in price that is credit-related. Price declines due to all other factors, such as the trading markets for the security not functioning properly, will only be charged to OCI, and as such will have no earnings or regulatory capital impact on the bank. 

Let’s now take a second, perhaps somewhat extreme, example of an AFS legacy debt security and see whether a bank would have any reason to sell it to a PPIF. Assume that the security has a cost basis of $100 and an intrinsic value (reflecting credit losses on the underlying assets) of $80 but can only be sold at a fire sale price of $20 because of the near-total collapse in the markets for asset-backed securities. Would it make sense for the bank to sell the security for, say, $70? 

In the case of a temporarily impaired debt security, the answer is almost certainly “no”. The sale of the security for $70 would reverse $50 of the prior $80 OCI charge taken, thus increasing the bank’s GAAP capital by $50. However, this gain would have no impact on the bank’s regulatory capital and would be offset by the bank’s accelerated recognition a $30 loss that would be both charged to income and deducted from regulatory capital. 

Under the new accounting rules announced in April, it is also difficult to see why a bank would sell the same security for $70 if it were OTTI absent near certainty of future credit losses or price declines or the bank’s inability to hold the security until it recovered in price. Such a sale would prevent the bank from realizing the full $80 intrinsic value of the security, as measured by the present value of future cash flows on the underlying assets and other factors, and would require the bank to take an immediate additional $10 charge to its earnings and regulatory capital. (The bank will have already taken a $20 charge to earnings when the OTTI determination was made.) This result runs counter to banks’ general preference to spread out their losses on loan and securities assets over longer periods of time.

* * * * 

In short, it may make sense for a small number of banks that are extremely pessimistic about the future of housing prices and the economy to sell now, take their losses, and put their legacy asset problems entirely behind them. But for the vast majority of banks, it will not be financially attractive  for them to sell their legacy assets to PPIFs. 

What If There Is a Capital Hole to Fill? 

There is also one other important but not always acknowledged reason – perhaps the most important reason – why many banks may not wish to participate in PPIP. While the above discussion assumes that banks and their accountants are accurately and reasonably marking their legacy assets to market as required by the applicable accounting rules, it is at least reasonable to speculate that some banks may not be doing so all that “aggressively” because, simply put, they do not have the capital wherewithal to absorb what their actual losses might be. If recognizing the true extent of its losses would create a capital “hole” that would cause a bank to become undercapitalized and subject the bank to more intensive regulatory scrutiny and oversight (or worse), the bank will have an incentive to underestimate the full extent of its loan losses and securities impairments to avoid accelerating the loss recognition that would result from selling its legacy assets.

How to Encourage Greater Bank Participation in PPIP 

So what, if anything, can the government do to make it more likely that banks with legacy assets on their books will want to participate in PPIP, and that PPIP will be successful in achieving its worthy objectives? There are no pat answers or simple solutions, but as a starting point, Treasury, the FDIC and the Fed may want to consider the following modifications to the program.

o Clarification of Accounting Treatment for Loans Submitted for Bids But Not Sold 

As an initial matter, in the case of possible legacy loan sales by banks, the government could work with FASB to obtain clarification that when a bank solicits bids for a pool of its legacy loans from PPIFs, a bank’s preliminary decision to participate in PPIP and its solicitation of bids from prospective PPIF purchasers would not, by themselves, require a change in the accounting status of the loans from HFI to HFS absent an affirmative decision by the bank to accept the highest bid and sell the loans. In this regard, the government needs to help ensure that a bank testing the waters can still treat its legacy loans as HFI if, for any reason, the bank chooses not to sell. Otherwise, the bank will run the “worst of both worlds” risk of forfeiting the more favorable HFI accounting treatment for its legacy loans while at the same time still having those loans on its balance sheet.

o Allowing Profit Participation for Selling Banks

There is already a great deal of discussion (and controversy) as to how any profits made by the PPIFs should be divided between the government/taxpayer and private investors. However, serious consideration should also be given to allowing banks selling their legacy assets to receive a portion of any profits earned by the PPIFs on the legacy loans and securities sold by them in order to further incentivize banks to sell. Such a future revenue share could be a key selling point for banks, especially given that bids made by the PPIFs are likely to be at a discount to loans’ fair values and securities’ intrinsic values, and legacy asset sales will almost certainlyrequire some accelerated loss recognition by the selling banks. Admittedly, how such a profit share would be structured and whose pocket it would come out of are thorny issues. But there is certainly no magic in PPIFs’ profits being split on a 50:50 basis between the government and private investors. Having a profit participation in PPIFs’ future legacy asset sales could tilt the balance in favor of banks selling their legacy assets and getting them off their books instead of retaining the assets and simply outsourcing their management to third parties.

o Additional TARP Funds to Fill Any Resulting Capital Holes 

To address what will perhaps be the most frequent reason why banks won’t willingly participate in PPIP, Treasury will need to commit to selling banks that if a sale of legacy assets to a PPIF causes the bank to experience a capital shortfall, and the bank is unable to raise additional equity to fill the resulting hole by resorting to the private capital markets, TARP funds will be made available to it to offset any losses that are incurred. Once selling banks are cleansed of their legacy assets, and their capital holes are filled with new equity, they should be well on their way to profitability and returning to financial health such that, in a short time, they will be able to repay their sales-related TARP funds

* * * * 

As even Secretary Timothy Geithner, Chairman Sheila Bair and other top government officials acknowledge, it is still too early to know whether PPIP will work or not. No one can say for sure whether there is a significant spread between legacy assets’ intrinsic or fair values and their currently depressed market prices, one of the key premises on which PPIP is based, or whether this difference is, as one observer commented, “just wishful thinking.” And no one can predict with any degree of certainly whether there will be a “sweet spot” where selling banks’ target prices and PPIFs’ bids intersect such that sales of legacy assets will actually take place once the program is started. Nonetheless, the government can and should take whatever steps it can to overcome the accounting and capital-related reasons why banks will otherwise have little reason to sell their legacy assets to PPIFs at this time. Those steps alone will not ensure PPIP’s success, but they will greatly increase the likelihood that selling banks and bidding private investors will be able to get to “yes”, as Roger Fisher, author of the book on principled negotiation “Getting to Yes”, might have said.

About the author

John Buchman is General Counsel and Corporate Secretary of E*TRADE Bank, a $46 billion asset bank subsidiary of E*TRADE Financial Corporation located in Arlington, VA, and is a member of the adjunct faculty at The George Washington University Law School. Prior to joining the Bank in November 2000, John worked as a banking attorney in Washington both in private practice and with the government.