Description

MBNA Trust Preferred Series E & D, junior subordinated debt of Bank of America holdco, currently trades at a 20% yield and should trade at 8% for a 150% gain.  Nearly half that gain is a simple arbitrage toward BoA's pari passu Trust Preferreds trading at 14.5% that are the same in all but name, and the other half is based on two things.  First on an understanding that trust preferreds can only be permanently impaired in bankruptcy, and second believing that BoA's favorable credit position and backing by the government prevents bankruptcy.

Previous bank trust preferred write-ups on VIC have been arbitrage write-ups, but this will focus mostly on the mechanics of trust preferreds and a credit analysis of BoA, and only quickly touch the arbitrage.

There are two major components in the 12% journey from 20% to 8%:

First half of gain (from 20% to 14.5%)   The arbitrage of different trust preferreds under BoA's Holdco umbrella has been well covered by majic's excellent Merrill report, with Rasputin providing well-spoken details.   I will refer readers back to that report (or even the Wachovia arb), but again emphasize that these are basically the same securities, except the BoA Trust Preferreds (U Series) have a 5.88% coupon vs. the 8.1% coupon for the MBNA, but the BAC costs 1% more!  MBNA debt was long ago assumed by BoA.

Repeat: Only the name on the trading security is substantially different.  Pari Passu, and both held at Holdco.   

Second Half of Gain (from 14.5% to 8%): Understand thatTrust Preferreds are only permanently impaired in bankruptcy, then realize that the potential for bankruptcy is remote.

Trust Preferred Background (TruPS) Background:   

According to a Fed Policy paper, "Trust preferred securities are...cumulative preferred securities issued....in the form of a trust...whose sole asset is a deeply subordinated note issued by the Bank Holding Company (BHC). The subordinated note is senior only to a BHC's common and preferred stock. In event of failure to pay the cumulative dividend, an event of default and acceleration occurs, giving investors the right to take hold of the subordinated note issued by the BHC. At the same time, the BHC's obligation to pay principal and interest on the underlying junior subordinated note accelerates and the note becomes immediately due and payable."

The only difference between general subordinated debt and Trups subordinated debt is that Trups interest payments can be deferred (with interest) for up to 5 years, which I discuss later. 

Peoples have argued that the gov't can "jump in front of these" or "stop payment on these." I disagree because:

First: Preferred or common stock isn't entitled to receive a dividend unless formally declared. The gov't doesn't have to formally wipe out equity shareholders to wipe them out in practice by never paying out anything. 

Key Point: Trust preferred have a contractual right to quarterly interest payments which only bankruptcy can change. Larry Summers was on Meet the Press today saying that the Obama administration will not be touching the sanctity of a legal contract, no matter how egregious, and that was a cornerstone of their philosophy.  

Trust Preferreds receive interest payments from junior unsubordinated bonds which have a set interest payment (8.10% rate, payable quarterly, for the MBNA E), are cumulative, and have a set maturity date (2033).  Paying the principal + accumulated interest (if deferred) is a contractual obligation which pushes the company into default if not performed.

Second:  While the company can dilute equity holders expected dividend unit per share to zero, with debt holders it is binary - either you get paid the coupon or not and is determined by banktuptcy or not.   Even if the government jumped into debt senior to the Trups, the all payments would have to be made (eventually) unless bankruptcy is filed and the contract is broken.  An exchange offer would necessarily be voluntary.

The Fed's Feb 24 supervisory letter saying banks should "cease" paying dividends on trust preferred securities if they have seems wrong.   The word "temporarily defer" is appropriate, but "cease" would be impossible outside bankruptcy, so I disagree and think it was wrongly worded.  I cover what happens under deferment last.

Retail has driven part of the irrationality

Trust Preferreds were sold almost exclusively to retail by banks, REITs, and auto makers who had complex financing structures making an institutional sale more difficult, but a sale to retail looking for "good yield" easy.  Retail bought Preferred stocks, exchange traded debt (GM issued lots of this), and Trust Preferred and some issues are near zeros as the companies go bankrupt and junior unsecured debt is worth little for these companies (e.g. Ford Motor, BankAtlantic)

 In addition, there really is a lack of information on these instruments on the internet.  Except for reading the prospectus (which retail never does, and stockbrokers rarely do) little info will be found to give retail the right answer. Thus, general liquidations to get out of dangerous "yield plays" in general has wilder arbitrages than the BoA series.

Finally, the obvious name similarity between Preferred Stock and Trust Preferred Stock (Preferred Stock issued by a Trust that holds sub debt) undoubtedly has caused mass confusion. And frankly, for a company going to chapter 11 with lots of debt and few assets, the recovery between the two won't be very different.

Risk of Nationalization/Bankruptcy:

From a high level, I find it hard to believe that the government wants allow both Citi and BoA to enter bankruptcy or nationalization.  By all accounts, Citi is worse than BoA.  Financial bear Meredith Whitney has said BoA is "solvent" and isn't at risk of insolvency, but that Citi is toast.

Looking at the details, the Federal Reserve has written (and Bernanke said in speeches) that banks should defer trust preferred interest payments if/when the banks either aren't earning enough to cover the payments, or they don't meet capital requirements (say a 4% Tier 1 ratio, which defines adequately capitalized).    I don't think either option are remotely likely for 2009 (or beyond) based on this math:

BAC current Tier 1 Capital is 10.7%, vs. 8.57% in Q1 2007.  Since ML's portfolio is probably well protected by the $120 billion in large loss protection issued by the Fed, I don't include ML's potential losses on assets and just use BoA's year end numbers.

Risk-Weighted Assets: $1,311B

Tier One Capital:           $150B  (includes recent $30B Preferred raise after ML deal)

2009 Earnings:               +$53B  (1.7% cost of funds, 5% average return on earning assets)

Credit card loss:            -$8.2B  (est. of 10%, far higher than previous cycles and highest est. I've seen)

Housing:                       -$64B   15% additional loss estimate, on top of current reserves

Commercial:                  -$34B   10% loss on commercial, mainly

Consumer/Other:           -$16B              

Total Tier One Cap.:      $80B    

Pro Forma Tier 1 %      6.1%  (Bear case)

I'll note that this $122B in estimated potential write downs are fairly draconian given the ML backstop, the likely complete write down of Countrywide (which is well reserved by most accounts and the MBS market is priced right, or even cheaply, by some accounts), and the $8.5B BAC write down in Q4 2008.  The current estimate of BAC's Q1 2009 write down expectation is under $7B, so much room remains.  BAC was only levered 10x to 1 at the end of last year.

There are also other levers to pull to drastically raise capital ratios:

• Covert all or part of the $67B in preferred stock to common. Would take Tier 1 to 15% as of today. Going from 15% to 4% would require a $84 Billion additional write-down beyond the above example, or a total of $206 billion.
• $1.4B/yr in non-governmental preferred stock dividends to cut (~$4B for all preferred). 

Other points of comfort.:

1)      Unlike other banks, BoA is very US focused. The US helpfully started the downturn way before other countries and should at least be fully realizing most losses by now.

2)      Their asset valuation estimates aren't necessarily worse than the industry. For YE 2008, their used to price asssets called for a drop in US home prices of25-30%, vs. 24% for JP Morgan, and 23% for Citi.

3)      Insiders buying.  For what it's worth, BAC management wouldn't be buying across the board if something glaring was wrong (although they can totally miss the macro environment)

4)      They do have some good assets, mixed in with bad assets.  This isn't a zero.  That said, they didn't get here by being winners in investment banking or trading as they have been terrible in that business, but thankfully Ken Lewis, the CEO, seems to realize that.

5)      The macro environment for lending is great.  Low cost of funds (1.5% - 2%) and decent asset earnings (5%+) should allow at least some banks to earn their way out of this.  Buffett said this was a wonderful time to be a banker, and even incompetents should be able to make money

Given the government backing of BoA and the decent credit situation for bondholders, I'd say this debt deserves to eventually trade near 8%, which is 4% above treasuries.

Deferral Case/Downside Scenario

For up to 5 years, BAC can defer the dividends (with interest), after which they must pay the cumulated dividends.  The NPV per units remains the same, but surely investors will sell the instrument off.  I assume a selloff to a 12.5% yield, near $17.5 dollars/share.

However, deferring isn't the best choice unless things get really bad. Why:

1)      Why defer the dividend, and have to pay 8.13% on the dividends, when can issue TLGP debt at under 3%, or when it's just about the highest cost of capital BoA has, or when your assets are yielding 6%?

2)      Capital ratios aren't helped.  Are still accruing the dividend as an expense.

3)      The trust preferred requires that BoA can't pay dividends or principal on any equity if the trust preferred is in deferral.  This includes the gov't preferred.  Now, the gov't could just pick up trust preferred or senior debt, wipe out their preferred stock, and solve that problem. But, there is only $20.7B worth of trust preferred outstanding, so deferring trust preferred only saves $1.2B/year, which seems trivial in the scheme of things unless BoA goes all the way to nationalization.  And even if the government wipes out all equity and puts the trust preferreds in deferment and then guarantees the rest of the company, you're still sitting on a 20% note that will get paid once the government rehabilitates BoA.

Redemption Scenario:

I see an early redemption as improbable. A long standing rule is that all Tier 1 and 2 capital instruments (incl. trust preferred) shouldn't be redeemed by companies unless the Federal Reserve is consulted.  While possible, I find it hard to believe a bank could get away with "reducing tier 1 capital" while in the crisis stage.  Maybe after the storm blows over, but, as Rasputin put it, the risk of transmogrification might be too great to do an economically rational thing (e.g. buy back debt yielding 20%, and issue debt at 3% that is gov't backed).

Catalyst

1) BoA earns itself out of the hole in 2009

2) Someone somewhere writes these arbitrages up on a widely read publication.  It'll happen somewhere sooner or later