Tuesday, August 21, 2007

Shinsei Bank

Business India, Januray 19-February 1, 2004
Karma in Japan

Indian executives of Shinsei Bank and Indian software companies have played a significant role in a great turnaround story in Japan

Shivanand Kanavi

Dhananjaya Dvivedi, corporate executive officer (banking infrastructure group) of Shinsei, is a driven man. He was recruited from Citibank, where he had headed many an IT innovation globally. “When we took over LTCB the challenge was dealing with the unknown. The bank was not visible to anybody. Internal workings, business, people’s capability, products that were sold and were part of the warehouse, everything was an unknown. Meanwhile, Yashiro-san (‘san’ in Japanese is a respectful address like ‘ji’ in Hindi) had a very sharp definition of what he wanted. He wanted the company to be profitable, and he did not want a short-term solution that involved social trauma to customers, depositors, or employees. Moreover, we had a timeline of only one year. We also had revenue targets defined in the very first year. These became our engineering design criteria and we had to work backwards, says Dvivedi.”

Today Shinsei’s IT infrastructure, built by Indian software companies like Nucleus, i-flex, and Polaris under Dvivedi’s direction, has become the standardsetter in Japan. The bank has turned around and is now the most profitable bank in Japan. Wall Street is eagerly awaiting its $2 billion IPO expected in the next few weeks.

However, to understand the change wrought by Yashiro and his team one needs to understand the bank’s previous Karma.

Birth of LTCB
To understand the change wrought by Yashiro and his team in Shinsei Bank one needs to understand its history. Long Term Credit Bank is what we in India would call a development finance institution. As the Americans handed over power to the Japanese in 1952, the new Japanese administration decided to discuss the nature of the financial system needed to fund Japan’s revival. “Will it be stocks, bonds, or banks?” asked the then finance minister Ikeda. The answer he gave later, with America’s tacit support, was: “Banks, not capital markets.” The Tokyo Stock Exchange was reopened, but it only played the role of tying up industrial families, the keiretsu, through crossholdings.

LTCB was started to fund the industry by issuing debentures, which were like bearer bonds. It was directed to lend to core sectors and later to automotive, etc. In fact a young car manufacturer with global ambitions called Toyota came up owing to support from LTCB. However, soon lending was not based on cashflow projections but on the fact that only companies with a web of social contacts would survive, while those without allies would naturally die. The interest rates did not reflect risk, since the government set the price of money. Money was not an end but a means to an end — and the end was the revival of Japan.
Over three decades LTCB funded a large number of companies like Kawasaki Steel, Bridgestone, Toshiba, Tokyo Electric Power, and so on. It was highly successful in its mission. But that created a problem. The successful manufacturing companies in the 1970s and ’80s with strong cashflows no longer needed loans from LTCB; if need be they could directly access global finance. This led to some experiments in learning merchant banking skills from foreign markets, but these activities were marginal. Then came the real estate and stock market bubble of the ’80s. LTCB, like others, grabbed this opportunity for high profits immediately. However, the collateral for these loans was suspect. But the LTCB culture did not allow for serious financial appraisal.


Death
And then the stock and real estate bubble burst in 1991–92. Since then the economy has stagnated. The financial system is burdened with bad loans that have been estimated to be as high as $1 trillion! The government, however, has continued to muddle along without taking hard decisions, riding on the still very high domestic savings of over $14 trillion.

After meandering along for six long years, with over $40 billion as bad loans and its capital base totally eroded, and carrying a capital deficit of $3 billion, LTCB declared bankruptcy on 23 October 1998 and the government nationalised it. R.I.P.

Rebirth as Shinsei
Then came a group of US investors led by Tim Collins of Ripplewood Ventures.
Collins, along with Chris Flowers, a former Goldman Sachs dealmaker, put together a consortium of investors including Citigroup, AIG, GE Capital, Mellon, Paine Webber, and David Rockefeller personally. It was as high-powered a slice of Wall Street as you could get. Moreover, Collins took on board powerful individuals like Vernon Jordan (close to Bill Clinton), Paul Volcker (former chief of the US Federal Reserve), and John Reed (former Citibank chief and currently chairman of NYSE) as advisors, while co-opting Mitsubishi and Nippon Steel bigwigs on the governing board.

This group eventually won the bid. They bought the bank for a total capital of about $1.2 billion. Since the quality of the loan portfolio was opaque to the new investors, the Japanese government allowed a put option for any ‘bad loan’ discovered later.

It was path breaking and not smooth, to say the least. It was the first ever foreign entry into mainstream Japanese banking and was perceived by many in Japan as another invasion of the gaijin (aliens) and that too by a detestable ‘vulture fund’ from Wall Street. For the protagonists, however, the deal became the proverbial pudding to test Japanese commitment for globalisation and deregulation and led to some highpowered lobbying in Washington and Tokyo.

Tim Collins persuaded Masamoto Yoshiro, former head of Exxon in Japan and Citibank Japan, to be the chairman and CEO (see interview with Yashiro) to execute the turnaround. He recruited a team of investment bankers to start that line of business and a host of senior executives like Dvivedi, Sajeeve Thomas, and Janak Raj from Citibank to execute his plans for a total remake of the bank.

“Japanese do not like abstractions like software, they excel in manufacturing the concrete”-- Masamoto Yashiro

A great believer in harnessing the skills of Indian software companies, Masamoto Yashiro chairman & CEO, Shinsei Bank, explained the issues in Japanese banking, in a candid interview with Shivanand Kanavi


Q You were an outsider to banking. So how do you look at banking in Japan?

A I spent many years with the oil industry, 30 years with Exxon. Both oil and banks were regulated in this country. Japan does not have oil resources, so they let the oil companies operate much more freely than banking, especially in the 1980s. In the case of banking, till recently banks were not free to offer new products. So like the CIO, CFO, etc, they had one ‘Chief Ministry of Finance Contact’. These officials used to go every day to the MoF and sit around in the corridor and see who was coming and going. It was that bad. There was no initiative.

Q When Tim Collins invited you to join him in Ripplewood you had already retired. So what made you take his offer?

A He was very insistent. He thought there were many business opportunities in Japan. I thought I must help in restructuring Japanese companies. But initially it was outside of financial services. But very soon the banks that were nationalized were available for investment even for foreign interests, and that came about very quickly. So I had to divert my energies to the banks. That is how I ended up here. But I continue to be a limited partner of Ripplewood.

Q When you took up the assignment at Shinsei did you think it would be a model for ot her Japanese banks?

A Yes and no. LTCB had only sold debentures and given corporate loans, so I thought we should change the business model completely. But to do all these things you needed to change the infrastructure. Infrastructure as it existed was antiquated and was 20 years old. We had mainframe computers, we were using Cobol! Everything was old. The story of Jay’s (Dhananjaya Dvivedi’s) contribution is very well known. Anything we want can now be supplied by Jay’s group. The product is à la carte.

Q Is this unusual for a Japanese bank? Are others acknowledging your success?

A Yes, it is unusual. Others do not want to acknowledge it openly, but IT people know that we are doing the right thing.

Q Japan is the second-largest software market. Who is servicing it?

A It may be in terms of money. If you have to pay five times more than elsewhere, you can even be number one very quickly.

Q What is your advice to Indian software companies?

A Strangely, Indian banks seem to go to Australia for software! Japanese companies focused on manufacturing and brought their costs down, but they did not touch non-engineering parts of overhead. They keep spending more, year after year. They are worried about visible products, which they can improve. Abstraction is the last thing we attach value to. India has many philosophers; we don’t have any.

Q Would Shinsei bank be interested in emerging markets like India?

A We are looking at Korea and Taiwan, but India is too far for us. But since we have many Indians working in the bank, maybe we should take a look at it.


“We had no management information systems here when we came in. Reports were produced once in six months, but top management could not monitor on a monthly basis which products were doing well, which customer groups were behaving in interesting ways, and so on. We also wanted to launch the retail bank with a whole new way of banking access. For example, ATMs here closed at 3 pm. It was not the law, more an industry practice, but everyone followed it. Also people were being charged for using ATMs. But a retail bank with very few branches like us had to invite people to use ATMs and the Internet more and more to do normal banking. So all these equired a new infrastructure, which at the same time had a low total cost of ownership. Jay’s team, with the help of Indian software vendors, has delivered that. Jay is like a cook — if we want caviar added to a product he can do it, if someone wants garlic removed from something else, he can do it,” says Yashiro with a twinkle in his eyes.

“I have known Yashiro-san from my Citibank days,” says Jerry Rao, CEO of MphasiS-BFL, which did a lot of preliminary groundwork for the IT roadmap at Shinsei. “In fact we both retired from Citi at the same time in 1998. I invited him to be the advisor to MphasiS, which I had just started. Yashiro-san and Jay had to cut costs, implement a retail banking solution in a 100 per cent wholesale bank and migrate from an old system to a new one, while retaining the old people. Strangely, Japan, which is very advanced in automobiles, electronics, and so on, is very backward in banking and insurance. Yashiro promised 100 per cent improvement. Executing five different meshed projects simultaneously isn’t trivial, even though the technology was known in global banking,” adds Rao.

“We had to choose a design that would allow flexibility for our business plan. Customer self-service was the first point of our system, not labour saving. When the customer comes in he should have full power to service himself. Flexcube gave us the first generation packaged software to run retail banking. It is an extremely well engineered product if you work within its limitations, but if you want to add something in there like a credit card, or any other complex product, then it has a problem coexisting in this package. The new-generation product is from Polaris. We use three or four products from Nucleus, collection systems, general ledger, etc, but the more important thing about them is tenacity. If something does not work, they make it work. Complex engineering pieces have been developed by Nucleus,” adds Dvivedi.

Hitting the headlines
What would have cost $500–600 million for any other bank in Japan using traditional mainframes and customized software has cost Shinsei just $60 million. Naturally the engagement is continuing and expanding. This has led to headlines in Japanese newspapers like: “If you want to learn about software, learn from India”— Nikkei Sangyo Shimbun (17 September 2001); “Not enough IT engineers, foreigners save in software development” — Nihon Keizai Shimbun (22 September 2002); “Two secrets of ‘Cruel’ Shinsei Bank: open computer systems, Indian companies” — Nikkei Sangyo Shimbun (26 October 2001); and so on.

The robustness of the system, besides its low cost, was also highlighted in early 2002 when two major banks – UFJ and Mizuho (the largest bank in the world) – had crashes in their mainframe-based IT systems. Now grudgingly other banks are trying to copy what was done at Shinsei. Thus, Shinsei is becoming an agent of creative disruption within Japanese banking circles in more ways than one. Bill Gates takes a keen interest in Shinsei and has visited it. After all, it is the largest installation of Windows based banking software.

But the change did not go smoothly. There was a definite culture clash. “Even now only about half the old employees appreciate the change. The other half might still want the placid old days of LTCB. But change is a one-way street,” says deputy GM (IT) Michiyuki Okano.

Culture clash
What were the roots of culture clash? “Firstly, Japanese believe in a consensus approach (namewashi). Any change has to be discussed and only after a consensus has been achieved things move. But here change was happening through orders from the top, and at a furious pace. Of course change could only come this way, otherwise we would be debating endlessly,” adds Okano.

“The other was the Japanese obsession with perfection. Europeans and Americans may accept some imperfections, but Japanese do not. Now this can take an inordinately long time,” says Okano insightfully. “In the beginning we might have been a little cocky too. We had installed Flexcube at 60 sites and Shinsei was the 61st. But soon we learnt a lot. The drive towards zero defects was unprecedented. We improved greatly through this experience,” says i-flex CEO Rajesh Hukku.

Nucleus had the tough job of supporting the migration from legacy mainframe systems to new Intel servers and a PC-based system. “We built a reconciliation engine to show the equivalence of both the systems. It was crucial to migration. However, many times we had people holding two printouts from the two systems to a light and pointing out a font difference or a change in the position of a column!” says Vishnu Dusad.

“Indian companies have good engineering skills, but to make an impression in Japan they do need to realise also that time in Japan is of utmost importance. A Friday morning deadline means that the client expects it at 9 am Friday Tokyo time, not Friday afternoon, evening, or Monday or Tuesday. In fact a familiar ‘no problem’ response always scares me unless a detailed road map is given to show how the problem will be solved,” says GM (IT) Peter Franken, another former Citibanker.

Already Dvivedi and Yashiro are thinking of leveraging the IT skills developed in-house in conjunction with Indian software engineers. There is a plan to start a Shinsei Bank support centre in Pune. Nucleus already has 100 engineers at a centre in Pune. “We are creating a pan-Asian banking model. The Pune centre will be a big step for us,” says Dvivedi.

The results are there for all to see. Shinsei ATMs are working 24 hours, and the retail bank has attracted over half a million customers starting from zero less than three years ago. Customers can access their accounts at railway stations and even through 52,000 third-party ATMs at no cost. Deposits by new customers have crossed $14 billion.

Of course technology is not the only thing that has brought this in. A brand new retail banking culture brought into Japan by the Shinsei team headed by another Indian, Sajeeve Thomas, is responsible for that. Every branch of Shinsei looks like a modern showroom of an upmarket store rather than the congested paper-laden office it used to be. Earlier each transaction used to take forever. It was assumed that the customer had infinite time. To count the money half a dozen or more officers used to be involved before handing it over to the customer! Today Internet kiosks at every branch take care of that.

In fact the impressive headquarters of Shinsei had a 100- foot-high glass entrance with not a soul there. “I could not believe it in the beginning. How can the lobby of a bank be empty and forbidding?” asks Thomas. He quickly changed it by starting a retail branch there with a Yahoo! Internet café and a Starbucks coffee shop thrown in. “We did a survey among our own employees on what should a retail bank look like. The names that popped up were Starbucks, Seven Eleven, Uniqlo (a Japanese clothing store), and Sony. We called in designers to remake all our retail branches to give a customer-friendly experience. The elimination of enormous amount of paperwork and centralised data processing created a lot of space in each branch. Shinsei had a brand recognition of 4 per cent when it started. Today it is 87 per cent,” says Thomas.

The self-effacing Thomas has brought in more than an exterior change. The culture has changed —the stiff old manager hidden behind paper in a forbidding room inside the branch has been replaced by a smiling one, right at the reception to direct any customer to the appropriate kiosk.

But what about the basic corporate lending business that Shinsei inherited? “Well that too has changed,” says Janak Raj another ex-Citibanker who heads risk management group.
“When I was in Citi Japan in the early 1990s we wanted to sell a building and there was no problem getting an offer of $500 million for it, by just walking around and visiting a few nearby offices! That was the nature of the bubble economy those days. Obviously these assets shrunk to 10 per cent of their value after the bubble burst,” adds Raj.

However, Shinsei broke the mould of zombie lending (bankrupt banks lending to bankrupt firms to keep them afloat), a term coined by Anil Kashyap of the University of Chicago’s Business School. Immediately it was declared a “cruel bank” and a representative of Wall Street’s greed. But Yashiro stood firm, in not lending to bankrupt firms just to keep the “relationship” going, while being sympathetic to viable ones. However, now the recovery has become smoother and many old borrowers have paid up, thereby reducing the bad loan portfolio — by $27 billion in three years. Out of those only $10 billion were reduced by exercising the put option and returning them to the government as agreed in the contract of sale, while $17 billion were actually recovered. “Now other Japanese banks are also collecting a lot of money. Otherwise
no-one can survive,” says Okano. Considerable progress has also been made by Janak Raj’s team in securitizing loans by issuing new bonds, adding another stream of revenue to the balance sheet.

As a result Shinsei has become the most profitable bank in Japan today— it declared over $440 million in net profit in March 2003. GM (management accounting) Sanjeev Gupta, another former Citibanker, and his team are happily preparing the balance- sheet for the much-awaited IPO. Investment bankers expect the offer, slated for mid-February, to fetch close to $10 billion for the bank. The investors plan to raise $2 billion through this issue. Considering that they bought the bank for $1.2 billion, the ‘vulture fund’ attribute of Ripplewood might change to ‘venture fund’.

Will Shinsei change the Japanese financial sector, as Tim Collins and David Rockefeller suggested? The jury is still out. Change anywhere, and more so in Japan, is painfully slow. When it does come, as in the Meji restoration at the turn of the century, it can be amazing. A mixture of angst about change and pride in being Japanese drives Yashiro at 74. But he can claim modest success – by the grudging imitations that are happening in other banks – and retire a satisfied man. Obviously he has absorbed global best practices and rebelled against inefficiencies and irrationalities in the Japanese system. But he has handled the change with firmness and sensitivity. Not an employee was fired in all this process and many more customers have been added. He put a stop to zombie lending and took his fiduciary responsibilities towards his depositors seriously, combining them with sympathy (not loyalty) towards his corporate borrowers. Shinsei Bank rose out of the ashes of Long Term Credit Bank of Japan, which went under in 1998 with $40 billion in bad debts

Yashiro loyalists, like Jay Dvivedi, Sajeeve Thomas, and Janak Raj, might move on once Yashiro retires. In fact Dvivedi might eventually even set up his own software consulting firm. The Indian software companies are using Shinsei as the edge of the wedge into Japanese market. Already Nucleus is finding new customers like Honda, and more banks. I-flex has installed its products in six more banks in Japan. And so on.

As for the 2,000-odd Indian software engineers in Tokyo, it is not really ‘Jaapaan, love in Tokyo’ à la Shammi Kapoor. Very few learn Japanese or study the Japanese culture. Most yearn for home food, for which there are over 200 Indian restaurants in Tokyo alone!

Shinsei has been good karma.