WILLIAM BRADY (pictured, in the middle) and Howard Sheperd had each spent more than 30 years at National City Bank before becoming its leaders in 1948. But even after all that time, they were not really sure how the sprawling financial conglomerate that would become Citigroup made money. George Moore, who would later rise to chairman, was appointed head of a “New Look Committee” to unravel the mystery. His conclusion: “We have never really known just where we made our net profits, but have generally proceeded on the assumption that we should encourage the growth of all these businesses to the maximum, on the theory that the more they grow the more money we could make.”
The assumption under the current boss, Michael Corbat, is precisely the opposite. But working out how Citigroup makes its money—and therefore which parts of the business are most dispensable—is just as vexing. In total, Mr Corbat reckons 60 businesses have been sold since the crisis. Among them are brokerage arms in America and Japan, a student-loan operation and some credit-card units. The most visible contraction has been in Citi’s consumer business, which is shrinking from 50 countries to 24, and in America, from 14 cities to seven. This week, Citi announced the sale of OneMain Financial, a subsidiary that makes high-interest consumer loans, and a stake in Akbank, a Turkish lender.

In spite of all this restructuring, Citi’s performance remains dismal. Part of the problem is the endless restructuring itself: provisions related to it were $148m in the first quarter of 2013, $75m in the second, $133m in the third, $234m in the fourth, $211m in the first quarter of 2014, $397m in the second, $382m in the third and $655m in the fourth, according to S&P Capital IQ, a financial-data firm. Its return on equity last year was 3.4%. Regulators have frozen its dividend at one cent per share (a yield of less than 0.1%) since the crisis. Even as the share prices of other American banks approach or pass their pre-crisis peaks, Citi’s is down by 90% (see chart). Its market capitalisation is well below its book value, suggesting it would be more valuable if broken up.
On March 11th the Federal Reserve will reveal the results of the second of the two annual “stress tests” it conducts for big banks, which attempt to simulate downturns to make sure that banks have enough capital to withstand them. Citi failed this test in 2012 and in 2014, and as a result has not been allowed to raise its dividend. The first failure prompted the depature of its then chief executive, Vikram Pandit; a third one would probably put an end to Mr Corbat’s tenure.
Citi’s sprawl is hard to fathom. It is present in 101 countries, and handles $3 trillion of transactions daily. It finances $600 billion in trade every year. More than half its deposits are foreign—far more than any other American bank, according to Moebs Services, a research firm. Foreign operations provide 60% of its revenue; two-thirds come from emerging markets. This part of the bank initially catered to multinationals, and to the American government and military. But as ever more, and smaller, firms expanded abroad, its customer base grew.
Citi’s international presence has always been both a strength and a liability. Take its Chinese operations, which helped keep the bank afloat in the 1930s, only to be forcibly closed in 1949. By the same token, Banamex, its Mexican unit, has been one of the most profitable parts of the bank, but lost $360m due to fraud in 2013. Frank Vanderlip, the CEO who first expanded abroad, nearly killed the bank with a hefty bet on Russia just before the revolution. He was succeeded by James Stillman, who staked Citi’s fortunes on global sugar markets with similarly disastrous results. Next up was “Sunshine Charley” Mitchell, who expanded both abroad and at home in the run-up to the Depression, leaving the bank in need of its first government bail-out.
If anything, the risks of big international operations appear to be growing. The fines and investigations the bank faced for its dealings in American mortgages before the crisis seem to be fading away. But it faces continuing litigation over allegations that it manipulated currency markets and interest rates and helped its clients launder money. Banamex is the subject of a fresh investigation; foreign transactions that may have helped companies to avoid American taxes have also caught the authorities’ attention.
Increasingly, big banks are held responsible not only for their own misdeeds, but also for the conduct of their clients. They have to be on the look-out not only for specific crimes like fraud, but for unsavoury people and activities, from tax evasion to terrorism, more generally.
On top of this, big global banks face ever higher capital requirements, with overlapping regulations set in multiple jurisdictions. The discussion of Citi’s obligations in this respect in its latest regulatory filing extends over 17 pages. The most important elements, says Steven Chubak, an analyst with Nomura Securities, are the 9% in common equity demanded by international banking standards (more than double the pre-crisis level of 4%) and various surcharges imposed by the Fed, which push the total to 11%. Further impositions may be on the way, including a “countercyclical capital buffer”. Rules on holdings of liquid assets add yet more complication and expense. Each time the overall capital requirement increases by a percentage point, Citi’s return on equity declines by a similar amount, Mr Chubak estimates.
These rules are giving smaller banks, or purely domestically focused ones, a competitive advantage. American banks that regulators do not consider systemic, for instance, need an equity buffer of just 7%. Even Wells Fargo, America’s biggest bank by market capitalisation, must hold only 10%, Mr Chubak estimates, thanks to its domestic focus. It is no coincidence that Wells recently exceeded the record valuation for an American financial firm set by Citi in 2001, of $283 billion.
Mr Corbat’s response to all this, and Mr Pandit’s before him, has been to shrink and simplify. Citi’s workforce has dropped by over a third since 2007, from 375,000 to 241,000. It will continue to fall. Consumer businesses are being jettisoned in various countries in the Middle East and Latin America, where concerns about money-laundering and financing terrorism are most acute.
But just how far to go is hard to decide. Some activities still look like anomalies: it is the only big bank that uses proprietary designs for its cash machines. Others, such as OneMain, are easily shorn, but very profitable: its return on equity is 19%.
Reducing the retail business is especially fraught. Citi initially tried to shrink its operations in Texas to two cities, Dallas and Houston. But customers were put off by its diminished presence, so it was forced to pull out of the state altogether. Analysts question whether it has an adequate network in the suburbs of Chicago and Boston to preserve its business in those cities.
Contraction abroad carries similar risks. Operations in obscure or hazard-prone countries may provide little or no return on their own, but may also be where Citi provides the greatest value to multinationals, in that companies have few reputable alternatives when doing business in such places. Hiving off Citi’s corporate business in different countries would not make any sense: American clients come to it for seamless access to foreign markets, and vice versa.
It is possible that Citi is already shrinking too much, consuming the seed corn that will produce profits later on, speculates Charles Peabody, an analyst with Portales Partners. Operations like retail banking, brokerage and asset management take time to build but provide steady pay-outs once established. It is this part of Citi that is being shed.
What remains is a more opaque and volatile corporate bank. Citi, Mr Peabody says, has recently become far more eager to undertake various risky sorts of transactions, such as “bought deals”, in which a bank buys a huge block of shares in the hope that it can sell them in smaller parcels at a profit. The most recent data from American regulators, from September, show that Citi held derivatives contracts with a notional value of $70 trillion—far more than any other commercial bank.
Citi says it is simply catering to its customers, who want to hedge against movements in commodities and currencies. But it is also on an increasingly urgent hunt for decent profits. How to produce them, however, is a question that has bedevilled Citigroup for more than a century.