Citi: John Reed reshuffles the pack


Published on:

Chairman John Reed's successor could be any one of half a dozen managers running the new streamlined Citicorp - Reed is giving nothing away about his favoured choice. But this group is fast changing the culture of the "largest small bank in the world" as it retools its approach to branded products and global coverage. Peter Lee reports.

After five years of crisis management and recovery from its near-fatal brush with commercial real-estate lending, Citicorp is implementing its plans for the next century. It intends to be a very different bank from the leviathan that lurched from one crisis to the next in the 1980s. That bank was decentralized, thinly capitalized, poorly managed and diverse.

This one will be the opposite. It will be dominated by just two strong divisions: consumer banking worldwide, and emerging market wholesale banking. A third and lesser division, commercial banking in the developed world, will concentrate on a handful of profitable products, dealing with an exclusive customer list. One of its main tasks will be to track these customers as they expand across borders, especially into emerging markets.

There will be no sudden lurch into insurance, information technology or global investment banking.

If the bank can stick to its plans, its opportunities are considerable. "People tend to think of Citicorp as some huge behemoth," says Robert Albertson, director of US bank research at Goldman Sachs. "It's not. In some ways, it's the largest small bank in the world. Except in US credit cards, in no product or market is its market share significant."

Citibank has room to grow. Supporting that growth is an unrivalled network of banks in 100 countries, including 75 emerging markets. No other bank comes close. "That network is the only true competitive advantage we have," admits one senior executive.

A change of structure is linked to a change in culture. The bank is turning away from what it calls the old silos: independently run groups in different countries or product areas, chasing exclusively their own revenues, spending furiously on their own people and systems, monitoring their own risks.

A bank whose managers once competed as strongly against each other as against other banks will be forced to cooperate more. Pay is changing to encourage this. The bank's senior managers will be more obsessed with the nitty-gritty details of consolidating back offices and improving productivity than with pursuing grandiose new designs in their own fiefdoms.

A new generation of senior managers is emerging at Citicorp to lead the bank into the next century. Most of them are in their early 50s, with a score of years at the bank. Many won their spurs in far-flung corners of its big emerging-country network. One will eventually take over from John Reed, possibly the most fascinating and controversial bank chairman in the world. Within a few years, America's leading bank could be headed by a chairman called Menezes, Aziz, de Souza or Talwar.

Reed is 57. He might run Citicorp for another eight years. It's too early to predict who will succeed him. And he's not giving anything away. Paradoxically, those who once seemed closest to the crown have tended to get blown away.

Shock wave

When Citicorp announced its latest management changes in January, a shock wave passed through the bank, its customers and shareholders. While a new generation was pushed forward, others had to make way. Most surprising was the fate of Pei-yuan Chia, head of Citicorp's global consumer businesses and widely regarded as the most important executive after Reed. Chia, born in Hong Kong, raised in China and Taiwan, joined Citicorp 22 years ago from General Foods. After running the bank's largest and most profitable business for four years, he appeared something of a bank stalwart. "It will be very strange dealing with Citicorp without having Pei Chia to talk to," says one American fund manager. "He was a friend."

But Chia is also 57. He was too old ever to take over from Reed. And Reed didn't want to make him an official number two by filling the post of president. That has lain vacant since the sudden departure of Richard Braddock in 1992. Braddock had also run the consumer business, until Chia replaced him in 1992. Braddock became president, and apparent number two to Reed, with responsibility for cost-cutting as part of the bank's five-point recovery plan. He left in October 1992 in circumstances never fully explained.

Braddock's departure foreshadowed that of Christopher Steffen, who joined the bank late in 1992, following his high-profile restructuring efforts at Kodak. He too was charged with controlling costs and introducing new management disciplines, such as benchmarking. In January 1995 he was promoted to vice chairman and came to be talked of as a possible future CEO. His resignation came suddenly in December 1995.

"These changes have been on the cards for well over a year," says one Citicorp insider. "But John [Reed] often holds his cards close to his chest. He appears to be abrupt." Abrupt and autocratic.

Chia will at least stay on for a few months. That is interpreted as a gesture of goodwill designed to calm nerves within the organization and silence any talk that there has been a palace revolution, or that the veterans of the bank's emerging markets business, who have supplied so much of its profits during the tough years, are flexing their muscles.

Chia's retirement clears the way for the bank's other vice-chairmen: Onno Ruding (56), William Rhodes (60) and Paul Collins (60). But their ages make it unlikely that any of these, the bank's most senior remaining officers, would take over from Reed. If any nursed such an ambition, it was probably killed in January when Reed announced that the three vice-chairmen "will continue to provide a senior policy overview".

The clue to the most likely candidates lies in a select ginger group once dubbed the G-15. Almost four years ago, at the height of the bank's real-estate loan crisis, Reed created this committee of the bank's top-15 business managers, who all reported directly to him. He required them all to fly into New York once a month for meetings that lasted an entire day and sometimes two. It set a punishing schedule.

"It was a killer," says one member of the group. "Reed would call us all together - from Asia, Europe or wherever - once a month and basically beat us up. We've had nearly 50 of those meetings." It was a management structure designed to help the bank through the worst crisis of its modern history. And it worked.

From 1993 onwards Reed knew that the bank had survived the worst. But he persisted with the G-15 while the bank repaired its balance sheet, rebuilt its tier-one capital and restored its rating. Of the original members of the G-15, 12 remain in key operating jobs. Last year after some joint consultation they told Reed that the G-15 structure, which had worked well during the crisis and recovery, wasn't suited to the present strategy. Reed responded by nearly halving the number of line executives who report directly to him.

The survivors are: from the consumer side, William Campbell, who runs the Citicorp branch network; Roberta Arena, who runs credit cards in Europe and North America; Shaukat Aziz, at present working with Arena to organize credit cards as one global business; Alvaro de Souza running private banking. On the wholesale side are Robert McCormack, who runs Citi's global relationship bank (GRB) for selected corporate customers; Paul Collins and Dennis Martin, who run emerging markets. Also reporting directly to Reed is Victor Menezes, newly appointed chief financial officer.

If these men and one woman feel they are starting a contest to succeed Reed, they are wise enough not to admit it. Reed could run the bank for another eight years. "Will I be the next chairman? No, I don't think so," says one executive who reports to him. "For all I know, the next chairman of Citicorp is running a sub-division of the credit card business right now."

Victor Menezes is a clear candidate one day to head the bank which he joined as a management trainee in his native India. He has wide experience of all sides of the bank, having worked in Latin America and Asia as well as in developed markets. He worked most recently in consumer banking and also spent some time in strategic planning in the 1980s. '"It's his name that you hear more and more," says one Citibanker. Menezes is said to be liked by Reed. A clue to Menezes's standing is that, while he presently fills the slot of chief financial officer, he seems to have taken on responsibilities that range well beyond managing the bank's balance sheet, funding and liquidity.

Says a colleague: "He's not going to be a numbers-obsessed CFO. His value is that, unlike Steffen, he has worked all over the bank and knows which buttons to press to make things happen."

Menezes describes his own job more in terms of a bank's number two than a CFO: "Trying to drive the fundamental performance of the bank; making sure we're implementing the strategy in a controlled way; making sure we have a balance between growth and control and taking responsibility for costs and productivity." He explains this proactive approach: "If I just sit back and wait for numbers to come in, I'll be in trouble, because numbers are a lagging indicator." The deployment of such an experienced banker away from management to a corporate position underlines the move to a more process-driven structure.

Branded goods

William Campbell, aged 52, is a rare outsider and something of an unknown quantity. He is a former CEO of Philip Morris USA, a company of which Reed is a director. Philip Morris is one of the bank's role-models in its quest to give its Citibanking branch network a powerful brand identity, so that it doesn't have to compete on price. Campbell has been a consultant at the bank for over a year. His official appointment in January was over-shadowed by a fuss over his testimony to congress denying that Philip Morris manipulated the nicotine content of its cigarettes to make them more addictive.

Partly because of that, partly because of Steffen's fate, there is some scepticism about his prospects. "We'll see," says one Citibanker. "He comes with great experience in marketing branded consumer goods, but he is not a banker. It will be years before he understands fully how this place works."

Joining Campbell is Roberta Arena, 48, who started at Citicorp 23 years ago and now runs the credit card business in the US. She had the difficult task of taking over from the highly regarded James Bailey, long-time head of that division. Analysts say her strength is marketing and that she has a strong staff running the card business. Shaukat Aziz, born in Pakistan, aged 46, joined Citicorp in 1969 and has worked in corporate planning and for the wholesale side of the bank covering the Middle East, Africa and Asia Pacific. Bankers who worked with him in Asia regard him as highly competent and likeable. "It will probably be a short-term assignment for him," says one bank insider. "He doesn't work for Roberta and she doesn't work for him." The card business will be reorganized as one global business later this year.

Alvaro de Souza, 48, made a success of the bank's new cross-border finance group and, before that, was country corporate officer in Brazil. His appointment to run the private bank, shows one of the drawbacks of moving senior managers around the organization. It robs the cross-border finance group of continuity of leadership after only a year of operation, insiders say. The group will be run by Dipak Rastogi, fresh from pulling the bank's derivatives units into one global business.

Bankers say de Souza has a reasonable chance of success in this part of the consumer bank. As a former country credit officer in Brazil he will have worked closely with the private bank staff there. The company CEOs and CFOs he would have courted for commercial banking business would also be clients of the private bank.

Another less well known figure to watch for is Rana Talwar, 48, an Indian who joined Citicorp 27 years ago, spent his early career in operations and corporate banking in India, and now runs the bank's consumer branch business in Europe and America, one of the bank's biggest divisions.

Dennis Martin, 53, joined Citicorp 28 years ago in his native Argentina. He now runs the emerging market business and bridles at any suggestion that its emerging market network exposes Citicorp to excessive risk. He points out in tones of mock horror that in a single day in early March the Dow Jones index fluctuated from a 171 point fall to a 100 point gain. That's real volatility. Unusually for a Latin native, he spent a large chunk of his career, seven years, in Asia.

Asked why emerging markets is a good training ground for Citicorp senior managers, he replies: "If you work in emerging markets as a country head, you run a whole business. The first person who comes to see you in the morning may want to talk about a credit matter, the next may come with a regulatory issue, the next with a staffing problem, the next will haul you off to meet one of his customers."

Of Martin's emergence, one colleague wryly notes: "We seem to have a lot of Argentinians at the top of this bank." Reed spend much of his boyhood and youth in Argentina and some close to him say that he is culturally more Argentine than American.

Robert McCormack runs global relationship banking (GRB) which aims to provide banking services to 1,400 large corporations with big international businesses. McCormack served with the US Peace Corps from 1968 to 1972. He spent much of his early career running the bank's operations in central America. He was later put in charge of working out the bank's real-estate portfolio. That meant he usually led off the monthly meetings of the G-15 by reporting how big a hole the bank was in.

At a meeting of senior bank managers one asked whether he wasn't scared of taking charge of this huge problem that threatened the bank's solvency. He replied, so the story goes: "Scared? No. I was scared when my branch in Nicaragua was bombed for the sixth time." He regarded real estate as a problem that the bank could work its way out of, if the top people put their minds to it. McCormack earned much kudos for his handling of the real-estate problem and would almost certainly be a candidate to run the bank if Reed decided to retire early.

A new culture

Citibank, like all global banks, is becoming more obsessed with systems and operations than with lending money. At management meetings the talk is of "strategic cost management", "enterprise-wide technology platforms", "mass-customization", "standardizing products". But you don't often hear are the words "entrepreneurial", "creative", "independent".

Citicorp managers are happy to be shifting this way. They want the bank's earnings to more stable and predictable, because stable earnings attract higher share-price multiples and they are all paid partly in stock and stock options. It's also the best way they can see to throw off the bank's reputation for being accident-prone which it gained with a string of credit problems and failed investments in the 1980s.

McCormack says: "In the past this bank was full of artists, solo-product guys who would say 'give us a lot of the shareholders' capital and we'll make a ton of money by going off on our own and doing something clever'. Well, the commercial real estate problem came from a bunch of artists trying to be clever." There's no room for artists in today's Citicorp.

The bank has decided that, instead of relying on such artists, it will build its performance on three sturdy businesses. The single most important source of profits will be the consumer banking business which Reed himself established at Citicorp over 20 years ago and which last year supplied over 60% of its earnings. It is the largest and fastest growing part of the bank. The bank's role models in building this business are not other banks but consumer product leaders, such as McDonald's, Coca-Cola and Philip Morris. The plan is to give the bank a brand identity as powerful as Levi's jeans or Marlboro cigarettes.

Citibank is striving to ensure that its branches are exactly the same around the world so that a customer arriving with his ATM card will find the same kinds of machines, services and corporate style in Belgium, New York or Colombia. But, as only 5% to 6% of its customers travel regularly from country to country, why bother? Partly because it makes it easier for the bank when it is opening new branches to have a standard design. Previously, when the bank sent executives overseas it left them to install whatever technology they preferred.

"When the customer touches the bank," says Menezes, "he has a set of experiences which may be pleasant or unpleasant. Obviously we want them to be pleasant, but also consistent. Remember, people don't go to McDonald's for the hamburgers, they go for the convenience, because they know the bathrooms will be clean and that they will be able to sit down with their kids and have a meal for not too much money."


Citibank still has to decide which of several technologies to embrace. Most bankers agree fewer customers will need a branch. It is not clear whether the future of consumer banking lies with home computers or telephone banking, and it is even less clear whether talking to a representative over the phone about mutual funds can become a branded consumer experience. To find out, Citibank, like others, will have to invest in several areas of technology at once.

The bank's wholesale business, which includes foreign exchange, global relationship banking and emerging markets, provides 40% of earnings. Some time this year, Citicorp will probably open in its 100th country. Some 75 of these are emerging markets, with generally high rates of economic growth and wide margins on banking business compared with OECD countries, albeit with greater political risk. Its nearest rival to the title of most international bank is HSBC which is present in just 45 emerging countries. For the last four years Citicorp has produced an ROE of over 30% from its emerging markets business (see box above).

Citicorp's commercial and investment banking business in the developed markets was once the mainstay of the bank. But this business now looks like a poor relation of the consumer and emerging market divisions. Its earnings have been volatile, partly because they encompass the bank's trading businesses. Its ROE was 12% last year, 10% in 1994 and 20% in 1993. This part of the bank has been reorganized into the GRB run by McCormack.

Here more than anywhere else Citicorp is defining itself as much by what it will not do as by what it will. It will not enter into mergers or acquisitions with other large commercial banks, nor buy an investment bank. It does not wish to be a leading trader and underwriter of equities. It is not keen on lending large amounts of money and will not seek earnings growth through balance-sheet growth. The GRB actually reduced assets from $96 billion in 1994 to $90 billion last year.

McCormack says: "The banking business in OECD countries is over-capitalized, very competitive and the pricing is horrible."

In recent years many international commercial and investment banks have struggled with the question of whether to organize themselves as collections of global businesses, or to be organized as complete banks in different time zones and countries. Many have plumped to replace country heads with product heads. The top executives don't run Europe, Asia and America, they run bonds, equities, foreign exchange all round the world.

Relationship structure

Citicorp has come up with a new idea. It has decided that it will place geographic reporting with customer reporting. The idea of the GRB is to organize the bank into customer groups. Their aim is to serve 1,400 customers, many already large and growing across borders. They will provide customers with certain wholesale banking services which multinationals value in many countries: transaction services such as funds transfer, custody and cash management; foreign exchange and securities trading; corporate finance; and some lending. The bank is clearly trying to concentrate on businesses it is already good at and which are profitable. An organization chart puts relationship managers and industry heads at the top, supported by key businesses which are also run by global product heads. These are all in turn supported by operations and technology.

For any bank to proclaim that it is customer-driven is usually an empty cliché. But Citicorp seems to be making it an organizing principle. Its model is the bank's own world corporation group, a team of relationship managers and product specialists which for more than 20 years has developed the art of winning large amounts of revenue from 217 of the world's biggest firms. It has always been an elite group within the bank, earning a disproportionately large amount of revenue. Now the bank wants to extend this model to cover most of its activity in wholesale banking in the developed world. The ultimate aim is for many of the bank's key people to define themselves as working at Citicorp in the Sony, GrandMet or GM group, instead of saying that they work for Citicorp Japan or for Citicorp in foreign exchange.

In the world corporation group, the most important people were the parent account managers and the subsidiary account managers. A brilliant parent account manager could report to his boss monthly, leading off with contact details for each person within Citicorp that worked on his customer's account, and adding details of revenue generated and prospects. As with everything else, Citicorp maps this out. It analyzes the balance between effort and revenue from the moment it starts working on a target account to the take-off point when it becomes a key relationship bank and beyond.

It is not altruistic. Citicorp analyzes its customers in terms of their wallet - the amount they spend on banking services - of its own share of their wallet, and what may be the changing characteristics of the different wallets the customer has in its home market, in international markets and in emerging markets.

Citicorp would seem to be at a disadvantage in competing in the developed world for business from large companies. It has no equities business, it is not a leading bond underwriter, it is not a leading mergers and acquisitions adviser. It is not even keen to lend money. Take the large credit arranged last year by the Kingdom of Spain carrying a drawn margin of 4bp. When it was drawn, some of the Japanese participants found it costing them 25bp just to lend money. Citicorp shareholders don't want the bank chasing too much of that business.

So what does Citicorp offer the corporate CFO who expects his lead bank to lend him money and will reward him for doing so with bond and even equity mandates? McCormack replies: "Those bond and equity deals may have a large share of his mind, but they aren't necessarily a large share of his wallet. Ask him how much he spends each year on transactions and cash management and how much his subsidiaries pay to raise money in Timbuktu. It will be very much more than the fees for bond and equity deals." He adds: "Even if you're a big firm in bonds and equities, even in a good year you don't make an ROE much above 15% and your stock doesn't get a high multiple."

McCormack argues - hopes perhaps - that what the biggest companies in the world want from Citicorp are not things that consume a lot of capital, like loans, but apparently boring things such as cash management and transaction services. Even more important, they want these banking services all round the world. Citicorp has offices in more countries than any other bank. "They actually take it for granted we have the products," says McCormack. "What they really want is the network, the fact that we can deliver a service which is basically the same for them to use in Russia or Malaysia as it is in Chicago. They also of course want it efficient and low-cost."

Ruthless pull-out

McCormack stresses that these are not like middle-market customers. By dealing mainly with the 1,400 biggest customers in the world, Citibank is concentrating on less risky customers. The bank has been ruthless in weeding out businesses it no longer wants to be in. The bank used to have many large mall operators around the US as customers but they wanted the bank only to lend them money. When Citicorp managers talked about doing securitization deals, they found the operators preferred to deal with investment banks. Citicorp has just pulled out of the sector.

Reed has said that the GRB faces a tough year. It has to do more with less. It has a complicated reorganization to absorb. There is bound to be some infighting, either between customer people and product people, or between customer people and the old geographical heads. As well as customer groups, the GRB is organized into product groups which will also be run as single global businesses, such as foreign exchange under Julian Simmonds and transaction services under James Bailey, former head of the bank's US credit-card business. These product groups exist to serve the 1,400 customers. But to be big enough to do that, they have to deal with counterparts outside the 1,400. The transaction services group has its own sales force to deal with such single-product customers, as does foreign exchange.

An obvious source of tension will be the price at which relationship managers want the bank to deal with their customers, versus the product groups' desire to be seen to run profitable businesses. It will be tough for Citicorp to report earnings broken down by customers, very easy to report numbers for three or four product groups. Ruding, who oversees customer relationships for the GRB, hopes the product groups won't feel they have been subordinated to the customer people. "Because the product heads now run global businesses, their power has been widely expanded," says Ruding. "Someone like Julian Simmonds, running forex, is stronger than before. The product heads are all heavy-weights."

Ruding admits there may be some concern among the bank's former country heads. These used to enjoy a high degree of autonomy because Citicorp has decentralized decision making.

Now Citicorp has merged its businesses in North America, Japan and Europe. The bank hopes to centralize certain operations to cut expenses and make its businesses more efficient. It has already done this in Europe. Five years ago Citicorp ran its operations there as 18 different banks in 18 countries. When it began to centralize processing, accounting, credit management and foreign exchange market-making, mainly to the UK, it found itself able to reduce staff and still deal with higher volumes of foreign exchange and cash management business.

Now, the bank has the chance to pull off the same trick on a grander scale. It needs to get all its businesses running on the same systems. In the 1970s and 1980s, country heads installed technology that was customized to automate particular processes. But the bank won't spend wildly to do this. "We have to self-fund this," says McCormack. "We have no reserves for this. Our attitude is we have to find a way to fund the investments we need to improve."

Meanwhile, Citicorp must avoid any more big accidents. "You have to do two things says Rhodes, "first spot what's coming down the line, second do something about it quickly before it hits you." Citicorp is heavily exposed to financial institutions. Three years ago, it became worried about the Japanese financial system and cut its exposure to second-tier banks, ahead of many of its peers.

Sceptical foot soldiers

It will take several years for Citicorp's latest plans to take hold. The bank's senior executives and its chairman have always been great ones for coming up with plans and strategies, management matrices, performance grids and the rest. They have not always been so good at executing their plans. Bankers on the ground are used to change, but they have probably had their fill of change and new strategies. True, senior management devised the five-point plan to rescue the bank in 1991 and 1992, and delivered on it. Then they delivered on a second three-year plan to recapitalize the bank and win back its AA ratings. Still, the troops will be sceptical.

One senior executive recalls a recent meeting with managers in Europe. They told him they were unsure who was now responsible for taking certain credit decisions. His rather unsympathetic reply: "You're all in this room. Figure it out." There are also hints of politicking and in-fighting. Bankers inside the GRB are keen to point out that though the ROE of their business (12% to 14%) looks poor next to emerging markets (with a consistent ROE of 30%), nearly a third of the earnings credited to the emerging markets business come from OECD-headquartered customers that are covered by the GRB.

There is some enthusiasm for the new strategy. A good aspect of the reorganization is that people have a clearer idea of how they are being judged and paid. Truett Tate, one of six customer executives to whom industry heads report, says: "In the past anybody could build appraisal forms for their business. I could focus on net new revenue, another manager might judge his reports on how well they built a new product area. At the end of the year, managers would sit down and evaluate their best people. I might ask how did so-and-so do at bringing in net new revenue and the reply would come: 'I didn't measure that. I measured something else.' The result was that we had almost no chance of getting everyone pulling in the same direction."

For the first time in his 25 years at the bank, Tate finds he is not being measured purely on numbers. The six customer executives are judged partly on their combined financial performance, but they no longer receive monthly reports with an individual P&L. Over time that should change behaviour.

"We are being judged on collective performance, not individual performance," says Tate. "So when we explain to a manager why he should do something for a product he's not responsible for, we can say it's because a large portion of his variable compensation doesn't depend just on his own P&L; it relates to the performance of the bank as a whole. This is the first time the bank has done this at a level below senior management."

He adds: "There's a whole group of people at the bank in their 20s and 30s who don't identify with the parochial, narrowly, selfishly defined culture of the past 20 years."

Climbing the emerging country grid

Some investors still worry about Citicorp's dependence on earnings from emerging markets. Doubters might draw some comfort from how the bank's risk management systems coped with the January 1995 Mexican crisis. When the peso devaluation hit, Citicorp's stock was hammered, as investors assumed it would suffer because such a large amount of its business is in emerging markets. But it didn't lose a cent in Mexico.

"From the Colossio assassination onwards, we closed our price risk and our interest-rate gap risk and increased our liquidity," recalls Dennis Martin who, with Paul Collins, runs emerging markets. In fact, the bank had one of its best-ever years in Mexico because, with international flows of capital disrupted, local companies pay higher margins for simple local-currency loans. Similarly, the bank had a great year in Venezuela, another country out of favour with international investors.

Martin can obtain a measure of earnings at risk for each country, updated daily. Recently he has been talking regularly to his country head in one emerging market suffering from a recession and from conflict among leading politicians. He has decided that the bank's operation in that country must increase its liquidity - which will hurt margins in the short term. He will review the decision in 60 days.

The emerging market group's reward for supplying strong profits to Citicorp and avoiding losses through the hard times is that it will continue to benefit from investment. Even while managing its way through the tail end of the real estate crisis, the bank invested in emerging markets at a rate of $100 million a year during the last two-and-a-half years - all this from capital the business generated itself. This year it will invest slightly more, opening up in new countries, expanding in certain existing ones and standardizing certain products, so that, as far as possible they look the same to customers everywhere. The theory is that it will make life easier for large companies if the process for obtaining a letter of credit from Citicorp is the same in Hong Kong, Argentina or South Africa.

Typically for Citicorp's process-obsessed managers they have come up with a grid to map this. The grid splits countries into five phases of development: early, growing, rapid development, maturing, mature. Onto this grid it adds the types of services appropriate to each stage. Citicorp typically follows its huge multinational customers into more obscure markets and they want basic cash-management and foreign-exchange services. These produce local-currency deposits which in turn fund short-term local-currency lending. During the second phase of development a country might require project finance and might develop some local bond or treasury bill market. Rapid development brings demand for more sophisticated services such as derivatives, and securities custody and, eventually, for asset-based securitization and American depositary receipt services as international investors seek to buy into it.

Using this grid, as well as projections for GDP growth and the pace of deregulation, Citicorp has chosen nine pilot countries as offering the greatest opportunities for profits and is mapping out the investments required to capture some of the wholesale banking revenues which will become available in those countries in the next five years. It will begin funding those investments this year. Martin says: "In a country like Taiwan, the central bank has told us what individuals and consumers will be able to do with their money for the next 10 years. If you can imagine the future, you can own that future." In other countries it's a matter of interpreting signals. "If a government says it wants to fund itself with something more than loans, it's a good bet there will eventually be a domestic bond market. But often bankers simply don't see that."

In some countries, Citicorp is building local brokerages, it may consider buying some small to medium-size banks or, in more mature markets, asset finance companies. But it has no plans for very large acquisitions. Unlike the rest of Citicorp, where country heads are becoming less important and businesses are being run as single global entities with their own P&L, in emerging markets Citicorp still has to build up sizeable banks in each country. But it can reduce its costs by centralizing some back-office functions. It has just centralized its Indian cash-collection system in Delhi; previously this was done in several cities. The bank has cut its unit costs by 30% or more. Most of the computers which process business for it in Latin America are in Springfield, Maryland. In Asia those computers are all in Singapore.