A hedge fund correspondent pointed us to an article at Institutional Risk Analytics on the Bear-JP Morgan deal. While we don’t subscribe to its view that Bear was “raped” (please, the firm was going to file for bankruptcy a week ago Monday), it contains an intriguing analysis of JP Morgan.
Differing with popular opinion, IRA argues that JPM is far from a financially strong institution. It has the highest gearing of any of the three large US banks (and remember, that includes the CDO-laden, walking wounded Citigroup) and by their measures, also has the highest level of economic risk per their metrics. JPM’s chickens have not yet come home to roost because its book is heavily weighed toward corporate business, and those problems are coming to the fore later. (The cognoscenti may take issue with their use of RAROC as another measure, but I’m not troubled when making cross company comparisons if you have access only to published financials).
Although IRA does not say so explicitly, the reasoning appears to be that the Fed pushed Bear into JPM’s arms as a way to shore up JPM. If asking a firm to take on a $13 trillion derivatives book, of which only $2 trillion is exchange traded, is a favor, I’d hate to see what punishment looks like.
I believe JPM will regret this deal (assuming it comes off) and not simply for the impact it is having on Jamie Dimon’s reputation. Bear is stuffed with the some of riskiest assets in the credit game: mortgage debt credit defaults swaps, JPM is thus increasing its exposures at time when it would be more prudent to reduce risk, effectively doubling down. As Smart Money describes it:
The Martingale, gambling lingo for what experienced traders call “doubling down,” is perhaps the quickest means to a bloody end. I got my first gray hair the day I understood why the Martingale system, despite all its attractions, simply doesn’t work.
The gambling system, which dates back to a London gaming house in the late 1700s, is completely irrational, yet incredibly seductive. The thinking: If you keep doubling your losing bets, eventually a winning trade will make up for the losses. Like making a deal with the devil, the Martingale system will always comes back to haunt you — and often more quickly than you might expect……
Sure, you might win once in a while with the Martingale — perhaps enough to keep you interested in the strategy. But eventually, you’ll lose it all, in a very quick and undignified fashion. Take it from somebody who has tried it — the Martingale will kill you. It has to kill you. Why? The strategy is inherently flawed. It’s designed to have you increase your bet at exactly the wrong time.
To understand the grim outlook for JPM, start the analysis with derivatives. Because of its huge market share in all manner of OTC derivatives, JPM represents a “super sample” of overall OTC market risk. In terms of total size vs the bank’s balance sheet, JPM’s derivatives book is more than 7 standard deviations above the large bank peer group.
Because of this huge OTC derivatives book, the $1.6 trillion asset bank can tolerate just a 15bp realized loss across its aggregate derivatives position before losing the equivalent of its regulatory Risk Based Capital (RBC). And much like the GSEs, JPM’s positions are too big to hedge – despite what Mr. Dimon may say to the contrary about laying off his bank’s risk. And note that we have not even mentioned subprime assets yet.
Look at the balance sheet of JPM’s three main subsidiary banks and the mounting stress from loans losses is apparent. At the end of 2007, JPM aggregated 97bp of gross loan charge offs, 1.25 SDs above peer, and produced a Loss Given Default of 85%, likewise well above peer. The Exposure at Default calculated by the IRA Bank Monitor using data from the FDIC was 202%, more than 2 SDs above peer.
At the end of 2007, JPM’s Tier One Risk Based Capital held by its subsidiary banks was just $88.1 billion, a tiny foundation for the bank’s vast trading operations. The Economic Capital (“EC”) simulation in The IRA Bank Monitor generates an EC benchmark of $422 billion for JPM or a ratio of EC to Tier One RBC of 4.79:1, suggesting that JPM needs almost five times current capital levels to fully support its economic risks.
This EC produces a “Stressed” result for JPM under IRA’s Counterparty Risk Rating and a RAROC of just 0.22%.
While JPM currently boasts the highest Tier One leverage ratio of the top three US banks by assets, in EC terms it appears to be the clear outlier in the marketplace with the highest levels of economic risk vs. capital of any large bank in the US — with one exception: Commerce Bancorp (NYSE:CBH). The relatively large MBS holdings of CBH push its ratio of EC to Tier One RBC over 8:1 in the IRA Bank Monitor simulation.
Why do we take such a dim view of JPM and the US banking sector generally? First, because the US real estate market is not yet even close to the bottom. Second, the commercial real estate and corporate credit sectors are being dragged down by the same deflationary forces that are causing the US economy to slow dramatically. When you consider that US real estate markets and bank loan losses are unlikely to bottom before this time next year, you begin to understand our bearish outlook.
JPM has been lucky so far because its risk book is heavily weighted toward commercial rather than consumer risk, unlike our beleaguered friends at Citigroup (NYSE:C). But like last week’s debacle involving BSC, the fast deteriorating situation at C could provide a catalyst that takes JPM down a couple of notches in the next few months.
We hear in the risk channel that the internal situation at C is going from bad to worse as veteran Citi bankers are in near-mutiny against the new, two-headed management team imposed by regulators. Meanwhile, former CEO Chuck Prince, who is a consultant to C, is leading the discussions with regulators on behalf of the bank and is, in effect, acting as shadow chief executive of C. One insider predicts that the C annual meeting in several weeks time will be “very messy” and notes that acting Chairman Robert Rubin is nowhere to be seen.
Keep in mind that C, JPM and many other large banks are still trying to get their arms around the full dimension of the risks facing their institutions, this even as bank loan default rates remain well-below long-term averages. All of the subsidiary banks of C, for example, reported 127bp of charge offs in 2007, a full 2 SDs above peer but well below 1991 loan loss levels.
Click here to see a loan default series for Citibank NA going back to 1989. Notice how low bank loan charge offs were in the 2006-2007 period compared with previous recessions. Of note, the ratio of EC to Tier One RBC for C at the end of 2007 was 3.46:1, significantly better than JPM, but still suggesting that C really needs more than 3x current capital to address its economic risks.
BTW, the maximum probable loan loss (“MPL”) calculated by The IRA Bank Monitor for C over the next 12 months is 400bp or well above 1991 loan default levels. We expect to see the MPL for all of the large money center banks move higher as 2008 progresses.