Goldman Forecast: Citi, Merrill, JP Morgan May Write Down $34 Billion

Goldman’s William Tanona predicts further sizable losses at major brokerage firms, with Citi and Merrill taking particularly large hits. Tanona also expects Citigroup to cut its dividend. From Bloomberg:

Citigroup Inc., JPMorgan Chase & Co. and Merrill Lynch & Co. may write down an additional $34 billion in securities linked to the collapse of the subprime mortgage market, according to Goldman Sachs Group Inc.

Citigroup, the biggest U.S. bank, may reduce the value of its holdings by $18.7 billion in the fourth quarter and cut its dividend 40 percent, Goldman analyst William Tanona said in a Dec. 26 report on the New York-based companies. JPMorgan Chase & Co., the third-largest U.S. bank, may write off $3.4 billion, double Goldman’s previous estimate. Merrill Lynch & Co. may reduce its holdings by $11.5 billion, he wrote.

Losses and writedowns at the world’s biggest banks and securities firms total $97 billion this year, according to data compiled by Bloomberg…

“It will be a couple of quarters before the current credit crisis is fully digested by the markets,” wrote Tanona, who has a “sell” rating on Citigroup’s stock and a “neutral” rating on JPMorgan and Merrill. “Given the magnitude of the writedowns we assume and Citi’s remaining exposure, we believe the firm has a serious need to preserve or raise additional capital.”

Tanona downgraded Citigroup’s shares on Nov. 19 and was the last of six analysts who follow the company to advise clients to sell the stock. Nine analysts rate Citigroup a “buy” while eight recommend holding the stock, according to Bloomberg data…

Writedowns at the biggest banks are still likely to be “significantly larger than investors are anticipating,” Tanona wrote…

Tanona previously had estimated that Merrill, which replaced CEO Stan O’Neal with John Thain, would have to write down $6 billion of securities.

“Many of the December year-end firms are likely to be more aggressive with their marks,” Tanona wrote. “Particularly those with high levels of exposure such as Citi and Merrill Lynch, both of whom have new CEOs at their helms.”

Sanford C. Bernstein & Co. analyst Brad Hintz estimated in a note dated today that Merrill will have a CDO-related writedown of $10 billion in the fourth quarter.

Print Friendly, PDF & Email


  1. doc

    There are lots of people looking at new ways to talk about liquidity, like:

    1. But according to Jean-Francois Robin of the French bank Natixis, “it’s not that there is a lack of liquidity (in the global financial system), it’s that it is not circulating.”

    2 “Overnight rates are very low, indicating lots of liquidity,” said Niels Christensen, economist at Nordea in Copenhagen. “Unfortunately, they (the ECB) can’t affect the three-month rates as easily as the overnight rates,” he said. “It’s not only a question of liquidity in the banking system… but capital requirements are stretched at the moment,” he added

  2. Yves Smith


    You make a good point, that there is a question of nomenclature. What is liquidity? In a specific market, it’s intuitive, it means you can buy and sell readily, but how do you define it systemically?

    This relates to a pet peeve, that academics and central bankers don’t pay much heed to money supply (certainly not the broader forms of money). The reasoning is that with the proliferation of many forms of near money, it has become an increasingly difficult problem

    To me, that would suggest that understanding the role and impact of near money was very important, but that isn’t a widely held view.

  3. Mencius Moldbug


    That round thing you find yourself fumbling around with is actually the handle of a door. If the room were not dark and full of smoke, you’d see a sign on the door. The sign says “Austrian Economics.”

    If you open this door and step in, you will find a bright hallway full of fresh air. In the hallway there are many, many answers to these questions that trouble you. If you don’t like the answers, you can always step back out again!

    I would not say that the Misesian canon is absolutely perfect and infallible in all regards. But a smart, curious fellow like you should see this as an opportunity.

    You are absolutely right: the maturity of claims to money is the most important thing about them. Even if you don’t care to read the Austrians (you can just ignore the politics), you will find this line of reasoning quite fruitful. Misesian economics is much like Darwinian evolution: it’s not complicated at all, and it’s easy to just reinvent it yourself.

    In an accounting system which considers maturity transformation fair play (from classic deposit banking to SIVs), you will see balance sheets in which long-term claims balance short-term obligations.

    This “maturity transformation” is profitable because long-term interest rates have a natural tendency to be higher than short-term ones.

    This is true because long-term periods of production offer more ways to profit than short ones. For example, since it is almost impossible to imagine any production process that produces a return on investment in one day, the natural interest rate of one-day money is zero.

    But just because maturity transformation is profitable doesn’t mean it’s healthy, either locally or systemically. Think about this for a minute. Suppose, just for simplicity, that you own a gold mine which will produce one ton of gold every year for the next 30 years. Suppose you then issue 30 notes, each of which is a promise to pay a ton of gold, on demand, to the bearer. Or to pay it in 30 days, or 60, or 90.

    In what sense is this operation solvent? And in what sense is it different from an SIV?

    The word “liquidity” in its classic meaning, which you explore above, relates to the difficulty of selling an asset. For example, a van Gogh is illiquid in the classic sense. It may be worth a lot, but it is hard to mark to market, and it takes a considerable amount of energy to set up a transaction in which it is exchanged for dollars. You have to call Sotheby’s and all.

    How did this term get to be used in the completely different sense of a credit crunch? Granted, long-term payment streams such as mortgages are inescapably nonfungible, like the van Gogh – every unhappy mortgage is unhappy in its own way – but this seems quite unconnected to the problem of a shortage of short-term money.

    The basic problem is that maturity transformation creates an enormous demand for long-term claims to money which is thoroughly spurious, because its source is people who demand notes that entitle them to gold in 30 days, not 30 years.

    If this financially engineered demand disappears for any set of long-term claims, we see implied interest rates that are much, much higher. Supply and demand.

    One reason for such demand to disappear is that the validity of said long-term claims is unclear. Another is just that some people think they might be unclear. There is a game-theoretic process by which the slightest disturbance can shut down either the whole maturity transformation engine, or that engine as it relates to any set of long-term securities. This process is called a “bank run.”

    The important fact is that a bank run is self-sustaining. When the present-money market price of a bank’s assets is lower than that of its obligations, no one has any reason to buy back in. Since the bank has to sell these assets in order to pay back the depositors, the feedback loop is clear.

    In extremis, the maturity transformation engine shuts down for all securities across the whole financial system. We find out what the real demand is to exchange present gold (or euros, or whatever) for 30-year gold or whatever. In today’s world, this calculation is unlikely to produce a 5% long-term interest rate.

    (Our engine cannot shut down for truly risk-free securities, such as Treasury bonds. So a risk-free number for the “natural interest rate” will not appear. However, if you estimated the actual probability of default for some of today’s toxic waste, and compared it to the market price as seen in, say, the E-Trade transaction, I bet you’d see some crazy double-digit implied long-term rates.)

    If you know that the engine is going to restart, you will make a tasty, tasty profit by buying subprime loans before everyone else does. If you think you know this and you’re wrong, you will get burned. Given the combination of political and game-theoretic uncertainty that determines the answer, anyone who claims to know when the crunch will end is probably lying.

    The result at present is that these instruments do not trade, because those who hold them have institutional reasons to believe that the engine will restart soon. Those who do not hold them are understandably skeptical. So bid and ask do not cross, and the market is frozen.

    Is this like the difficulty of auctioning a van Gogh? Sort of. And thus the term “liquidity.”

    Google “Diamond-Dybvig” for the game-theoretic model of this multiple equilibrium problem. These people are not Austrians – they write in math and work at the Fed. Perhaps this will increase your confidence in them.

    The Austrian point is that, in the absence of loan guarantees issued by an authority with infallible power to back up those guarantees (eg, a central bank which can print fiat currency), there is no multiple equilibrium. There is only one equilibrium: the bank run.

    If there is any chance at all that your bank is insolvent, the ideal strategy is to pull it all out now. If you don’t follow this strategy, others will. Since your bank will have to sell long-term claims to satisfy its short-term creditors, its sales will depress the entire market for long-term claims, or at least for the category of claims held by the class of banks that is under pressure. So all holders of these securities will become insolvent.

    Ergo, absent official guarantees, rational investors will shun maturity mismatches and demand temporally matched balance sheets.

    Ergo, the layer of political band-aids that has built up, in successive financial crises, to supply short-term demand for long-term instruments (“liquidity”) when the Diamond-Dybvig switch flips over, is the entire cause of the problem. There is nothing inside the ball of duct tape. Central banking is the ultimate self-licking icecream cone.

    Another way to describe the problem is to assert that there is no way for a Bagehotian lender of last resort to insure against illiquidity without also insuring against insolvency, unless the LLR or its delegates (eg, NRSROs) can evaluate the quality of all loans. Like all systems of administrative pricing, this scales pretty much the way a lead balloon flies.

    A financial market which uses a system of accounting in which balance sheets must be balanced not only by scalar valuation, but also by temporal maturity, would not be subject to Diamond-Dybvig panics and would not need any mechanism of centralized loan quality control.

    Transitioning from the present Bagehotian money market to a maturity-matched financial system in which the State need not guarantee private loans, formally or informally, would be no mean feat! However, the first stage of recovery is always the realization that you have a problem. I hope I’m not wrong that you seem to be approaching this realization.

  4. Mencius Moldbug

    Correction: of course, by “all holders of these securities will become insolvent,” I meant “all holders of these securities which have issued short-term liabilities against them will become insolvent.”

Comments are closed.