The Morgan Stanley Puzzle

Grab your jackhammers. Let’s do some drilling.

The Federal Reserve’s behavior towards Morgan Stanley is puzzling. How much of the puzzle is Morgan Stanley’s own doing, and how much is the Fed’s?

The Morgan Stanley Puzzle, Part 1

As noted by the Wall Street Journal and in Morgan Stanley’s third quarter 2011 earnings press release, the bank increased its risk-weighted assets in its regulatory capital calculation by $44 billion. The third quarter press release attributed the move to a change in the Fed’s permitted treatment governing collateral for over-the-counter derivatives. The press release stated that the Fed advised the bank in October 2011 of the change, after a review.

Before the Fed was purported to change its mind, Morgan Stanley was already dramatically reducing its measure of derivatives credit exposure in its regulatory capital calculations in 2011 and possibly at the end of 2010. Those who study the bank’s report to the Fed on Form FR Y-9C would have noticed the drop. And it pretty much entirely occurs in the 20% bucket. (You can easily access that report by clicking on the link in each bank name in the Fed’s Top Fifty Bank Holding Companies ranking. The derivatives number is line 54 in section HC-R — Regulatory Capital.)

This is the other, hidden financial bubble, playing out in the 20% bucket of the regulatory capital calculation. It hugely involves credit ratings. The 20% bucket is the risk-weighting category relating to exposures to counterparties rated A-/A3 or better. That’s the rating category where the biggest banks’ long-term senior unsecured ratings have been parked throughout the post-meltdown period. Credit ratings drive much of the need for banks to post collateral to support their derivatives contracts. The banks disclose in their annual reports (their SEC form 10K accessible in their investor relations websites) both that government support “lifts” their credit ratings and also that the rating agencies are reviewing this assumption of government support. The credit rating agencies have begun to drop certain of the biggest banks’ ratings through that A-/A3 threshold.

There’s big, and then there’s big. Comparing the biggest banks by asset size gives rise to a different ranking than using derivatives exposures. The single derivatives number I use in my book is the “credit equivalent amount” on line 54 (see above), the estimate the Fed requires them to provide in the regulatory capital ratio calculation. It’s a measure of how much their derivatives counterparties owe them. The 20% means that only 20% of the exposures in that bucket go into the regulatory capital calculation (into the denominator).

Ranked by assets at year-end 2011, JPMorgan Chase is the biggest at $2.3 trillion in assets, followed by Bank of America at $2.1 trillion. Citigroup and Wells Fargo exceed $1 trillion. Goldman Sachs had only $924 billion in assets. Morgan Stanley is seventh, after MetLife, with $750 billion in assets.

When Goldman Sachs and Morgan Stanley first submitted a Form FR Y-9C to the Fed in the first quarter of 2009 (as newly-minted bank holding companies after their rescue post-Lehman), Goldman Sachs ranked first in derivatives exposures. Their credit equivalent amount at March 31, 2009 was $612 billion, followed by Morgan Stanley at $564 billion. JPMorgan Chase was third at $455 billion.

The whole group’s derivatives “credit equivalent amount” numbers dropped through subsequent quarters, due to risk management, interest rate impacts and perhaps other reasons. By mid-2010 the group’s derivatives exposures pretty much leveled off. They stopped declining. Goldman’s was around $400 billion, JPMorgan Chase’s at around $350 billion, Bank of America’s ranging between $320 and $340 billion and Citigroup’s just over $200 billion.

But Morgan Stanley’s continued to drop. By the end of June 2010 it was $350 billion. At the end of 2010 it had fallen to $239 billion, over $300 billion less than than first quarter of 2009. Goldman’s had declined a third. Morgan Stanley’s had declined more than a half.

Morgan Stanley’s derivatives credit equivalent amount dropped precipitously to $86 billion at the end of March 2011, then to $78 billion at the end of June of last year. After an email to the bank and then an email to the Fed’s regulatory staff — and some nudging, the Fed replied in August that indeed there was a disagreement with Morgan Stanley over these numbers. When the third quarter 2011 FR Y-9C came out, the exposure was back to $289 billion, a $211 billion difference. Primarily in the 20% bucket. 20% times $211 billion is close to the additional $44.4 billion in risk-weighted assets mentioned in the third quarter earnings release. Yet from the first quarter of 2009 to that third quarter of 2011 Morgan Stanley’s total notional amounts of derivatives increased $16.6 trillion. (That is the sum of all four derivative product categories in lines 12 and 13 in section HC-R of that report. Derivatives held for trading comprised 99.8% of all derivatives.)

If one can get one’s regulator to agree with this interpretation, it is a way to insulate one’s balance sheet from the need to post more regulatory capital as one’s counterparties are downgraded. As exposures move from the 20% bucket to the 50% bucket (for the weakest derivative counterparty credit ratings, those below A-/A3), the reporting bank would have to post capital on an additional 30% of the exposure.

Studying the bank’s third quarter earnings release required by the Securities and Exchange Commission — the 10-Q — showed little explanation and only a slight change in the derivatives portfolio. However, reading through the bank’s press releases and the Basel II regulatory capital requirements led this analyst to theorize that the bank was using exchange trading to reduce its derivatives exposures as measured under Basel II rules and therefore capital backing them. Towards the end of 2010 Morgan Stanley began to issue the occasional press release that it was now clearing certain groups of over-the-counter derivatives on exchanges. As Dodd-Frank intended to reduce the perceived risks in OTC derivatives by moving them onto exchanges, these press releases probably didn’t raise any concerns.

Any regulatory capital system could never be perfect. The Basel II requirements are in some ways a little lax for global markets, but in other ways rather stringent. In Annex 4, Section II, point 6 in Basel II, regulators allow banks to avoid posting any capital against OTC derivatives contracts with exchanges. ”An exposure value of zero for counterparty credit risk can be attributed to derivative contracts … that are outstanding with a central counterparty (i.g. a clearing house).” Any number times zero equals zero. So no derivative contracts with exchanges need be included in the denominator of the regulatory capital ratio under Basel II. One can debate whether exchange-cleared derivatives are really risk-free after MF Global’s demise. (See my earlier post.)

The Basel II point goes on to nail two key requirements to remove exchange-traded derivatives from the denominator. First, the exchange needs to be the legal counterparty — the buyer for each sell contract and the seller for each buy contract. Second, margining must take place daily. Margining is the assessment of credit exposure followed by posting of collateral (preferably cash) to protect the bank if credit exposure to its counterparty increases. The latter is a difficult standard to meet. All the banks disclose that their derivatives are often illiquid and have lengthy maturities. It is hard to believe that those can be reasonably accurately measured on a daily basis on such huge portfolios. Not having details, one can only speculate. Even JPMorgan Chase CEO Jamie Dimon testified to Congress on June 13th that he can’t manage based solely on models.

The clincher was the behavior of Morgan Stanley’s peers. In such competitive markets, why weren’t the other biggest banks doing the same thing? Had Morgan Stanley believed it had an agreement with the Fed about how these contracts would be treated? Taxpayers and investors will never know what happened behind closed doors between the Fed and this bank. But there’s more to this puzzle.

The Morgan Stanley Puzzle, Part 2

Grab your hard hats. We’re going underground for the next step.

Was the Fed in fact allowing just Morgan Stanley to do this given the seemingly unfair treatment it was getting from the Fed over moving derivatives to its FDIC-regulated bank? (See my earlier post.) Others in the press have pointed to the Fed’s allowing Bank of America to move its Merrill derivatives into the FDIC-guaranteed bank, but not allowing Morgan Stanley to do the same. The Office of the Comptroller of the Currency is the multi-state banking regulator. The OCC publishes a quarterly report on derivatives. Great information for those who want to know more. First part reads like a newspaper, followed by charts and tables full of insightful data.

At the beginning of 2009, when the formerly investment banks Goldman and Morgan Stanley had become bank holding companies, JPMorgan Chase had all its derivatives in its FDIC-guaranteed bank according to the OCC derivatives report. Citigroup had nearly all there (93%). Bank of America, now owning Merrill, had 50% (its own, but not Merrill’s) in the bank. Goldman had 84% of its derivatives in Goldman Sachs Bank USA. Goldman set up its FDIC-guaranteed bank in 2004, the year of the infamous SEC decision allowing the Consolidated Supervised Entity regulatory regime to supervise investment bank holding companies. Morgan Stanley set up its FDIC-guaranteed bank, Morgan Stanley Bank, in 2001,  just after the repeal of the Glass-Steagall separation between investment banking’s trading intensity and commercial banking’s federal deposit guarantee. Morgan Stanley’s is the older “bank,” yet in March 2009 they still had no derivatives booked in the FDIC-guaranteed banking subsidiary.

It gets worse, at least for Morgan Stanley. After the meltdown Morgan Stanley Bank had rather small amounts of derivatives — $24.6 billion of purchased credit default swap protection and $16 billion in interest rate swaps held for trading (both notional amounts) at December 31, 2009 (in their FDIC call report). At the end of 2011 the percentage was still only 3% (OCC derivatives report). Goldman Sachs Bank USA, in contrast, already had booked $718 billion in credit default swap purchased protection (notional amount) and had written $640 billion in credit default swap insurance protection (what AIG provided that got them into trouble) by the end of 2008, just after Lehman’s failure. Goldman’s gross notional over-the-counter interest rate swap derivatives held for trading in their FDIC-guaranteed bank at that date was $38.6 trillion according to their FDIC call report. By the end of 2011 Goldman’s percentage of derivatives in its federally-guaranteed bank had risen to 92%. Morgan Stanley’s was still only 3%.

What is driving the Fed’s uneven treatment of these two investment banks? This favors Goldman as a counterparty in the derivatives markets. It allows Goldman’s bank counterparties to reduce the amount of regulatory capital they would otherwise have to post if their derivatives contracts were subject to Goldman’s holding company rating rather than its FDIC-guaranteed bank rating, which is one or two notches higher. As a Goldman counterparty, this helps Morgan Stanley, too. But the perhaps unintended consequence could well be to drive some of Morgan Stanley’s future derivatives trading opportunities into the arms –or onto the books — of its peers. Retained profits are the best source of regulatory capital. If a bank doesn’t trade, it can’t earn profits.