Moral Hazard at the FDIC

There has been a lot of discussion about Moral Hazard regarding the large bank bailouts (which the recent financial reform legislation apparently now has “institutionalized”).  And there has been a lot of commentary about the lack of help for America’s community institutions; in 2010 alone, the FDIC has closed 108 such banks (at least as of this writing).  It just appears to us that Washington protects the Wall Street firms to the detriment of Main Street.  We have written several blogs about this “unholy” alliance (http://www.thestreet.com/story/10771337/1/the-washington-wall-street-alliance).   Now, it appears that the FDIC, whose insurance fund was in the red by $21 billion at the end of the first quarter, has decided to become a player in the private sector by partnering with, you guessed it, Wall Street!.  This is a major strategic change in direction at the FDIC in how they deal with the disposal of troubled assets at failed institutions.  Specifically, the form of the strategic direction change is to allow a partnership with the private sector in order for the FDIC to participate in the upside potential of loan work outs and sales.  In technical lexicon, this is called an “Entity Sale” as opposed to a “Direct Sale” of the assets.

Partnering is not new to the FDIC, and it appears that the current strategic shift had its roots in the Resolution Trust Co. (RTC) model.  The RTC was the entity that dealt with the troubled assets from the S&L crisis era.  At that time, the FDIC, through the RTC, took back stock and warrants as part of their resolution in the largest failure cases of Continental Illinois and First City.  This created the public perception and significant outrage that “nationalization” was occurring; this seems eerily similar to today’s attitudes.

In one of their first forays into the entity sale strategy, the FDIC sold the assets of First National Bank of Nevada for between 30% and 50% of face value to Private National Mortgage Acceptance Company, LLC (PennyMac), a company founded by the former president of Countrywide.  The FDIC provided interest free loans to PennyMac, but kept an 80% equity stake, hoping to recover some of the losses should the real estate market turn upward.

Since the PennyMac deal, the FDIC has found it increasingly difficult to find private sector participants, so the last few deals have seen a reduction in FDIC’s equity stake to 60%.  For example, Colony Capital and the Cogsville  Group paid $445 million for a 40% stake in the assets of Community Bank of Nevada, First Bank of Beverly Hills, and New Frontier Bank.  The FDIC gave the group an interest free loan for slightly more than half of the purchase price, thus lowering the upfront cost to the group to $218 million in cash.

A July 21^st news report at Bloomberg (FDIC Selling Corus Bank Loans is Bet on Failed Condos, Sterngold and Keehner) summarizes the policy shift.  “The new FDIC strategy for managing assets seized from failed banks has turned the agency into a long-term investor, making a multibillion-dollar bet on the recovery of some of the most distressed condominium markets in the country, from Miami to Las Vegas.  Instead of selling the assets to maximize cash in hand, the agency is offering its private-sector partners zero-percent financing, management fees and new loans to complete construction projects it can hold until markets recover.”  In the Corus Bank Deal, the partners, led by Starwood Capital Group, took a 40% stake in the new LLC entity while the FDIC retained 60%.  The partnership received an interest free loan for half of the purchase price.  It also received a promise by the FDIC to provide up to $1 billion in additional financing for working capital to complete the condo projects.  The private group also receives a management fee of 1% on the outstanding face value of the assets.  In order to put liquidity back into the deposit insurance fund, the FDIC sold $1.38 billion of notes backed by the partnership’s assets into the fixed income market with an FDIC guarantee.

We have several issues with this new FDIC approach:

• First, a government agency has now become a private sector competitor with the ability to give zero interest rate loans to its partners, effectively giving these chosen ones a competitive market advantage.  These chosen few private sector partners can offer lower prices on their projects than the rest of the private sector.  Not only is this unfair, but it also puts more downward price pressure on the already strained real estate market.
• Second, a government agency is now taking on the kind of risk normally associated with private sector capitalism.  Capitalists, of course, are looking for large returns.  At the same time, they are putting their capital at risk.  Risking capital just doesn’t seem appropriate for a government agency which is charged with protecting the public’s principal (i.e., FDIC insurance), not in risking it for outsized returns.
• As a corollary to the above issue regarding risk, the FDIC is charged by law with resolving the problem assets in the “least cost” fashion.  Depending on what happens in the real estate markets in the future, the new approach may or may not be the “least cost” approach.

This last issue led us to examine the Corus deal more closely.  The Bloomberg article cited above provided details of the Corus Bank assets sold to the FDIC/Starwood public-private LLC.  We analyzed the possible outcomes and concluded that, only under very favorable market outcomes does the FDIC appear to come out better than if it just held on to the loans and sold them as the RTC did in the S&L crisis.

In the model discussed below, when we had to make choices about which assumptions to use, we always chose those that were favorable toward the FDIC. Here is the detail along with the assumptions:

• The FDIC sold $4.5 billion of Corus Bank condo and other commercial real estate loan assets to the partnership for $2.77 billion, or about 61.5% of value.  The $4.5 billion of face value had already been written down by Corus to appraised value to somewhere south of $4.5 billion, but probably not to $2.77 billion.  We have no access to the written down values on the Corus books when the FDIC seized the assets.
• Bloomberg indicates that because of the zero percent financing, the bid by the Starwood group was 20% higher, indicating that the “fair market” value was closer to $2.22 billion, or 49% of face value.  In the analysis below, we show the results for the “deal value” of $2.77 billion, as well as for the “fair market” value of $2.22 billion.
• The FDIC advanced a $1.385 billion zero percent loan (50% Loan to Value ratio).  Then, in order to re-liquefy its balance sheet, on March 8^th, it sold a note into the fixed income market for a similar sum ($1.377 billion).  That note was backed by the Corus assets and carried an FDIC guarantee.  Because this note was privately placed, there is no information available on the collateral (transparency?).  However, we have found the private placement announcements:
  • $150 million 0% coupon due October, 2011 at a discount of approximately 50 basis points (.50%);
  • $850 million 0% coupon due October, 2012 at a discount of approximately 110 basis points (1.10%);
  • $377 million 0% coupon due October 2013 at a discount of approximately 165 basis points (1.65%).

We assume that the interest rate on the note will average 2.5% if the assets are disposed of in 5 years rising to 3.5% if the term is 10 years.  This seems reasonable given the current yield curve.

• The additional $1 billion in working capital financing to complete the condo projects is not analyzed in this model.
• Normally, one would expect the collateral to have some cash throw off from monthly payments.  However, the assets all appear to be severely distressed, either in an unfinished stage (thus the need for additional working capital), or, if finished, with little or even negative cash flow.  As a result, we did not include any cash income to the partnership from monthly payments by the borrowers.
• A Starwood private partner entity receives a 1% management fee based on the face value of the $4.5 billion of assets.  If nothing is sold, that amounts to $45 million per year.  The fee is paid, not to the public-private partnership, but to a separate “management” company owned by the private partners.  This is the typical hedge fund model.
• The model assumes that the sales take place evenly over the time period assigned (e.g., 10 years).  In addition, the FDIC’s interest costs and the management fee costs are evenly amortized in the same way.  This assumption, in particular, works to the benefit of the FDIC.
• The private sector Starwood partners’ investment is 40% of the required equity, i.e. $2.77 billion of assets less $1.385 billion of loan times 40%, or $554 million.

The tables below show the percentage of recoupment relative to the deal value ($2.77 billion) (i.e., FDIC Net/Deal Value) or the fair market value ($2.22 billion)(i.e., FDIC Net/Fair Market Value) that inures to the benefit of the FDIC over 5 and 10 year periods at various revenue realization levels relative to the deal value.  The appendix to this piece shows a complete model iteration using a 7 year time frame.

Sale Time Frame: 5 Years;   Equity: 60% FDIC, 40% Private

Sale Time Frame: 10 Years;   Equity: 60% FDIC, 40% Private

Observations:

• The zero percent loan and the 1% management fee are perverse incentives from an FDIC point of view.  That is, the shorter the time frame to sell the assets at a particular price, the better off the FDIC.  But the zero percent loan and management fee make it to the advantage of the private partners to hold for longer periods of time.  The tables show that the private sector partners don’t make any money selling at fire sale liquidation prices, so it is in their best interests to wait for the real estate markets to firm up. (In the model, the private partners break-even at sale prices of about 90% of deal value.)  So far, the FDIC has done 14 such deals and is likely to do many more.  The overhang of the resulting assets, especially if banks continue to fail at their current rate, will continue to put downward pressure on real estate prices and makes even a 10 year time frame appear optimistic.
• For the FDIC to receive the “deal price” ($2.77 billion) on a net basis, the assets would have to sell for nearly 150% of deal value ($4.16 billion) in 5 years and 160% ($4.43 billion) in 10 years.  On the other hand, if the FDIC kept the assets in an RTC type entity, it would only need to sell the assets at 100% of deal value ($2.77 billion)(plus costs).  This leads us to question whether or not this model is “least cost” to the taxpayers.
• Said in another way, if the market does not “come back” in 5 years, but remains where it is, and the private partners decide to exit and sell for 100% of today’s deal value, the FDIC will get only 73% of deal value.  If this happens in 10 year, the FDIC gets only 65%.
• As expected, if the ownership percentages are altered to 80% FDIC, 20% private partners, as in the PennyMac deal, the FDIC benefits at every sale price level to the detriment of the private partners.

Conclusions:

1. Only if the real estate market recovers significantly (rising at least 50%) in the next five to ten years will the FDIC be “better off”.  It could be significantly “worse off” if the real estate recovery is modest or non-existent over that time frame.
2. In this new “strategy”, we once again observe an “unholy” Washington-Wall Street alliance.  Large Wall Street institutions are given sweetheart deals and, as a result of such subsidy, have a significant competitive advantage over regular Main Street businesses.   If the FDIC is now giving out zero rate loans to the Wall Street entities so that they can “wait” for the markets to turn up, why don’t they just give that same zero percent financing to the institutions that they are about to seize, allowing them to count it as capital and giving them that same precious “wait” time for the real estate markets to recover? (Note: If they do this with the small banks, they don’t even need the cash – just accounting entries!)
3. Finally, given the fact that the FDIC appears unable to structure a deal without a perverse incentive, we ask, do we really want the government in the risk/return business where poor decision making leads to loss of capital?

Robert Barone, Ph.D.

Joshua Barone

August 4, 2010

Appendix:  The Model

Sale Time Frame: 7 Years

Deal Value: $2.77 billion

Average Interest Cost to FDIC for Private Placement: 3.00%

Private Partner Ownership Percentage: 40%

Cash Invested by Private Partners: $554 million

In Bills $ except Percentages