This will be a survey of Citibank consumer bank technology projects from its initiation in the late 1970s through the CORE project in the 1980s and the Systematics project. The immensely successful global cards project developed and deployed by the International Cards Center in Singapore is described in some detail as well as Asia’s lead in installing the Systematics banking suite in 12 countries. All these successful projects culminated in the expensive fiasco of the global Go to Common project – which having cost US$ 1.4 billion remains the most expensive bank technology disaster ever. I have tried to tell this remarkable story in the context of what was happening in the bank management as well as the consumer business because both aspects were intermingled and fed on each other.
I had graduated with an Honours degree in Economics from Cambridge University and was working in the Central Bank of Ceylon Economic Research department in Sri Lanka when they installed the first computer in the country – an IBM /360. Meanwhile I had been awared a Fulbright scholarship to do a PhD at at a US university and I had already applied to MIT and Harvard. Nevertheless I decided not to accept the Fullbright scholarship and instead decided to stay and specialize in computers – purely out of interest and against the advice of the Governor of the bank , William Tennekoon, who had been a colleague of my father in school. He warned me against wasting an expensive education in one of the best universities in the world on a ‘technical’ career when I could have aimed for a place in the IMF or World Bank after I obtained my PhD. However I insisted on joining the Computer Department and I was trained by the bank in the UK, Canada and Japan and I had a very successful ten year career pioneering computerization in Sri Lanka. I automated the National Employees Provident, the National Savings Bank ( 4 million accounts) and the National foreign exchange system .
However when the racial riots erupted in Sri Lanka in 1983, I as a Tamil and a member of the minority ethnic community had no option but to leave and came to Singapore to join the Singapore government National Computer Board, who were then recruiting foreign experts to jumpstart their computerization drive. I was recruited by a very forward looking Singaporean called Robert Iau who was at one the Managing Director of the Singapore CPF
Fortunately I was convinced by a fellow Cambridge alumnus , Mano Appapillai who was the Regional Electronic Banking head for Citibank, to join the bank instead. Nick Greville the country manager in Singapore insisted on recruiting me over the objections of the senior local management, who were against hiring a ‘third word candidate’ – however once I joined I never looked back and never encountered any prejudice after I had proved myself in what was at that time an all Singaporean Chinese technology group.
I was fortunate to play a key role in 4 of the projects described below and my information is therefore first hand My information on the Global Go to Common project came from my former staff still with Citibank, who were horrified at the waste of shareholder funds ( $ 1.4 billion) and valuable technology talent on a useless project, when there were more productive projects with real business benefits. However, this article is primarily a personal memoir, told from an Asia Pacific perspective initially that later expanded to global technology projects.
There has always been a conflict between those who want totally centralized processing and those who prefer a degree of decentralization. The example of the last project described below ( the Global Go to
Common, where over $ 1 billion was spent without achieving the required results and which was cancelled in 2016 ) will show the folly of the former model – nearly 40 years after it was first tried in Citibank with the CBS project .
Citibank, is an institution with a proud 200 year history. It was the second largest bank in the world according to market capitalization by the late 1970s and the most forward looking bank in terms of a global vision for customers, markets and products. Under the legendary leadership of Walter Wriston the bank had what came to be called the Five “I”strategy – Institutional banking, Individual Banking, Investment Banking, Insurance and Information ( the joke was that there was a sixth “I” – standing for Indians , given the remarkable flow of management and technical talent that flowed from the subcontinent to the rest of the global franchise). Walter Wriston also famously said that “information about money is as valuable as money itself“– a perceptive forecast of the extensive role that financial information was to play via the internet 20 years later. His was the first articulation of the mega-bank concept – a financial supermarket on a global scale providing for all the needs of corporates and individuals. Having articulated this visionary concept and having laid the foundations for it, Wriston handed over to his hand picked successor John Reed in 1984.
John Reed, who was the head of consumer banking then set the stage for a global consumer bank with his memorable “Memo from the Beach”, which laid out the roadmap for a truly global consumer bank franchise to serve the emerging global middle class, and handpicked extremely able bankers with a proven track record to implement his vision. This was achieved with superb execution by his lieutenants like Pei Yuen Chia, Rana Talwar. Victor Menezes , Alan Weber and Yashiro San. It was a time of optimism and overarching global visions – the Reagan era banking deregulation had opened the door to building a superbank, global competition, and the growth a huge middle class in Asia and Latin America (Latam) with their aspirations for a better life that required upscale financial services. All these factors created the conditions required to build a global consumer banking franchise. Above all powerful computers were now making their way for the first time into commercial applications from the scientific labs. No industry stood to gain more than banking with its outdated 19th century legacy manual procedures that were both slow and labour intensive and prone to error. The first bank to really grasp this opportunity was Citibank and John Reed set about with determination to implement this new capability which was to transform banking over the next 20 years.
The 70s and CBS
John Reed when he the head of Consumer Banking was a true visionary in technology and a man of immense drive. He realized first and more than any other senior bank executive the potential of technology to transform consumer banking, and being himself a MIT qualified engineer (who contemplated a career in engineering before choosing banking) had immense interest and personal knowledge in the subject. He will be membered for 3 major technology projects – one an unqualified success and the other a partial success or failure, depending on your point of view. The third was his sponsorship of the Global Telecommunications Network (GTN) which I will expand on in the section on the Asia Pacific Processing Center (data center), and which was not only way ahead of its time but a total success.
The unqualified success was his pioneering of the Automated Teller Machine, which is commonplace today, but which, when he launched it, was thought to be impractical and prone to failure and vandalism. He proved the naysayers wrong by acquiring a lab and manufacturing facility in California called Transaction Technology Inc (TTI), which developed and manufactured Citi’s own ATM machines with a revolutionary touch screen. This was an unqualified success and was a tremendous boost to Citi’s North American consumer franchise. It was later deployed internationally as part of Pei Chia’s model branch project that did a great deal to establish Citibank’s upscale brand and image.
The other effort was to build an integrated, all-purpose banking platform, named the Consumer Banking System (CBS) –and the best brains in the bank, TTI and outside were recruited for it. They did all the right things – built a data model with an integrated CIF linked to all product processors which in turn had had all the bells and whistles. This was the first time that such an ambitious project had been attempted by any bank but despite the best efforts of all concerned it was abandoned after a few years after $50 million had been expended – a huge sum at that time. It failed because it was a bridge too far – the technology infrastructure was not mature and they attempted too much in one go – a mistake that was to be repeated on a much larger scale 35 years later with the Global Go to Common project that will be described later.
However the CBS project had two positive legacies – it created a cadre of trained bank technologists who would have a significant impact later, and the Customer data model created for it would later re- appear as the Global Relationship Banking (GRB) customer relationship model that was implemented in South America and Asia but this time on IBM mainframes initially on the Adabas platform and later on the IBM DB2 platform. It was therefore not a total waste of time and effort as it laid the foundation for later successes like the CORE project and Systematics.
The CORE project
The next attempt to create an international consumer bank platform had modest beginnings. Jack Berger, a senior technologist in New York, was looking for a banking system that could be rolled out internationally to serve the growing consumer businesses in Latam, Europe and Asia in mid 1985. He found it in a small software house in Florida called Newtrend – they had a banking package that ran on a mid-range IBM computer. There were 2 unique aspects to this banking system – first it was based on the System 38 machine which had unique futuristic technical features – an integrated database and communications interface, a capability where the operating system handled file placements and other unique features all based on a 48 bit addressing system. All way ahead of its time and also much cheaper to run as IBM did not charge separately for the operating system software as they did for mainframes. With its integrated communications feature and database/file placement functionality it also required much fewer technical staff for maintenance, like the notoriously temperamental prima donnas of the computer world called ‘system programmers’. The second feature was that the Newtrend banking application itself was an integrated software package with DDA, loan ,time deposit and general ledger processors all linked and integrated with a customer information system, all utilizing the unique advanced features of the IBM System/38.
But the drawbacks were also daunting – it was designed for small banks in the South East of the US with none of the features that were required for an international bank, chief of which was a multi-currency capability. But it had the structural integrity on which to build the advanced features required for international banking, so it was chosen.
Then what happened was a text book case for an international rollout. There was a 3 month training in early 1986 in Florida for 20 handpicked technologist from Asia, Latam and Europe (I was one of this select group and in between trips to Disney World and tennis we managed to master the software in 3 months ). This was followed by a modest international kick-off meeting in Tokyo chaired by Alan Weber. The recently purchased Newtrend system was now renamed as CORE – Consumer Oriented Retail System. We then dispersed to our respective regions and collected requirements and passed them on to the technical team in New York to develop a version of CORE that was suitable for Citibank’s international business.
By mid 1987 CORE release 1.0 was ready and implementation started for Japan in Asia and in Europe and Latam as well. There were subsequently 6 monthly new releases of the base with added functionality and within 3-4 years all the international businesses were done with the exception of the large business – such as Hong Kong, Singapore and India in Asia, Germany in Europe and Brazil in Latam. The main reason for this was that IBM deliberately kept the processing power of the AS/400 family (the successor to the System 38 architecture) modest to prevent it from cannibalizing the revenue from their profitable /390 franchise which had much more powerful processors.
The individual businesses were responsible to fund their own projects and hire their own technical staff to handle the conversion programs and country specific legal/reg modifications.. New York provided the base software and generous technical assistance where required. This produced buy in from the businesses and minimized the bureaucracy required the. In addition 2 technical support groups were set up – one in Manilla set up by Jack Berger to provide international software support, and one in Singapore by Tsui Kwok Wo. Tsui was the low key but effective technology manager for Asia Pacific, who was adept at transferring successful products between businesses ( for example the loans system – CLS from Hong Kong to Singapore and the multi-currency time deposit system CitiPlus also from Hong Kong to Singapore) . I was the Reginal System Manager for CORE with responsibilities for deployment in Asia Pacific.
The CORE project succeeded in rolling out an international platform in Latam, Europe and Asia with superior functionality for over 18 businesses worldwide within 4 years with a very modest central outlay – this was a world first for any bank. It succeeded because it did not overreach itself with too ambitious goals, the base platform was robust, the technical direction at all levels was superb, there was no micromanagement from New York, and above all they got the buy-in from the businesses. It allowed for a standard platform globally, while resisting the temptation for total standardization, but which still allowed for the exchange of functionality between businesses. It also fitted in with the prevailing management ethos of ‘Paradisation ‘ – a clumsy term describing the strategy of pushing technology management to individual units rather than centralization.
Jack Berger and the CORE project team deserves more credit than they have been given for this project and it also it showed the way to a successful deployment to multiple businesses globally in the future. Jack Berger and the New York team did a super job and they were ably assisted at the regional and country levels. I enjoyed my time on this project, although it entailed excessive international travel – for 3 years I lived out of a suitcase.
Rana Talwar – the complete banker
Rana Talwar had joined the India business in 1969 as a management trainee with an honours degree in Economics from St Stephens College Delhi. After a stint in the back-office in India, he worked as the Corporate bank Treasurer in the Middle East (Rana was a excellent gambler by instinct and took a lot of money off his staff in poker in later years!). He then moved to a chief-of-staff position in Hong Kong before being appointed as Consumer bank head for Singapore and South East Asia. On these modest foundations, he was to build the most successful banking franchise in Asia (he was promoted to Asia Pacific Head to succeed Alan weber in 1987). He was a complete banker with an instinctive grasp of what needed to be done, a very sharp aptitude for financial numbers an astute judge of the type of people required for each position.
He was above all an inspiring leader – many years later a senior American banker, who was not even working for Rana at that time said of him “I would walk over broken glass for this man “. All his staff felt the same way.
He was moreover an excellent judge of character and chose people of the caliber of Jerry Rao, an intellectual and a visionary, Ashok Dutt, Rajive Johri, Dave Hendrix, (a tough ex-Vietnam war marine) and Ashok Vaswani , all superb at execution and solid bankers , as his business managers. Lou Sanandres, John Botcheller and Bill Chapin ran his group office with the minimum of bureaucracy and maximum effectiveness. All were Alpha type A characters, and avid golfers like Rana, so there was the maximum of camaraderie and zero office politics.
Rana had a few core principles which were regularly drilled into his staff – absolute integrity in all our dealings with customers, regulators and each other, excellence in execution and the necessity of building an upscale branded franchise. One of the few times I have seen him really angry was when our Singapore business launched a product called ‘Citimaids’ aimed at domestic helpers. He shot down the project saying ‘Don’t damage our brand. This is not our target market’.
Rana treated his staff very well – gave them clear objectives and was very generous in rewarding good performance. He was excellent with numbers with an astonishing memory for detail. Needless to say, he was an inspiring leader and those who worked for him did their very best under all circumstances. This was an example that should be emulated by Citibank management today. Rana did not believe in or depend on strategic plans or highly paid consultants. He rather believed in his gut instincts as a banker, choosing the right people and the rest would follow and this was vindicated by results. Rana and Pei Chia were an ideal team at the top of the Consumer bank.
On a more mundane level, Rana was also largely responsible for the tremendous success of the Asia Technology office in Singapore, which was to develop and deploy banking and card systems globally in due course. Though not interested in technology per se ( his eyes would glaze over whenever I ventured into details during a technology presentation), he knew that technology was important, hired the right people, and was over generous in funding it .Above all he empowered his staff, who more than repaid his trust in their capability. I consider it a privilege to have worked for him for over a decade.
Between 1991 and 1995 Rana was a member of the G15 – a group of EVPs who ran the bank with absolute authority reporting directly to John Reed. This was to restore the credibility of the bank with regulators and rebuild shareholder capital which had been depleted by sovereign loan losses – and they succeeded in this. It gave Rana unprecedented freedom of action without interference from New York and he put it to good use. The Systematics project, described later, is one example. Another one was a series of lectures by world leaders like Margaret Thatcher and Colin Powell speaking on leadership. This had a tremendous impact on regulators, politicians and the press and did a great deal to restore Citibank’s credibility. This may never have been approved under normal circumstances.
Rana built the Asia and Middle East upscale consumer banking franchise for Citibank between 1983 and 1996, and the fact that is still survives today (when almost everything else has been sold off) and is extremely profitable and upscale is a testament to the excellence of his efforts and that of his team. This was despite the fact that he was followed by a string of mediocrities as the Asia Pacific Consumer Bank head like Simon Williams, Brian Clayton and Fritz Seegers. Sadly, this franchise is all that survives of what was initially a much larger vision for a global franchise.
The Asia Cards business and the global ECS+ Cards technology project
Now when it came to a decision to launch cards in Asia in 1988 , Rana did not rely on consultants as this was not his style, but after some introspection came up with his own strategy – he would launch in multiple geographies simultaneously and the objectives would be simple – 5 million credit cards in 5 years, each credit card would yield a $50 annual profit leading to a net $250 million increase in the bottom line. A simple objective but one which would require superb implementation. As it turned out this objective was achieved in under 4 years – an astonishing success.
Rana chose as his main players good managers rather than people who had credit card experience – Rajive Johri who set up the hugely successful and profitable Indonesia franchise and later moved to run Hong Kong an, and Raymond Lim who created the most profitable card business in Singapore – a supposedly saturated market. Jerry Rao, the India business manager and Ashok Dutt launched the very successful India cards business (India was the only country to develop its own technology platform and was the last to be converted to the ECS platform in the early 2000s), and Ashok Vaswani who built first a cards and then a banking franchise in Turkey (in the face of intense competition from local banks which had excellent homegrown technology of their own). The only person with ‘cards experience’ was Jeannine Farhi who had worked in the massive US cards business and who came to Asia as the first Regional Cards head and to be the conduit for the transfer of best practices from the US business. Ashok Dutt , Jeannine and Rajive were to play a pivotal role in the cards success story in Asia. Jeannine and Rajive were early supporters of building a Regional Cards Technolgy Centre and in appointing me as its head. I owe them a great deal.
There were many objections to launching cards in Asia. The pundits said -there were no credit bureaus, there was low merchant acceptance, Asians will not revolve their balances, fraud will be endemic etc etc – Rana ignored the naysayers and trusted his gut instinct. He realized that there were many pre-requisites and moved decisively to implement them – first class managers with a proven track record (cards experience was secondary), a robust credit process and above all a world class technology platform. Everything else would follow. I was lucky to be chosen for the latter job as the head of a regional cards technology center and what followed led to the most successful technology project ever for Citibank (with the development and global deployment of the ECS+ platform), or indeed any other bank in Asia. The system is still in use and is superior to the technology platform for payment systems of any competitor bank.
When I struggled in 1989 to draft the Major Expenditure Proposal (MEP in Citibank jargon) to set up the regional cards technology unit and data center Rana told me not to waste my time – “Your job is to build the best infrastructure and a world class technical team. I have enough bean counters to draft the MEP”. Thus I set out armed with literally a blank cheque and hired the best talent in the market and from within Citibank, from my experience in installing the CORE project in Asia. They turned into superstars whose achievements I will detail below – Susan Kwek, the architect of the cards and payment system and chief technical guru who understood both the technology and the business and regularly worked 14 hour days, Dhruv Panchal who implemented the system in the Middle East and Eastern Europe, Tan Sai Tien and Lim Tee Tsian who did Japan, Ong Geok Leon, Derek Mollison, Sundar, Reddy, Yong Choon Bin, Lawrence Tan, Lee Sik Meng and many more. At this point I should include Richard Parry and Vaman, who although a part of the regional credit office, worked seamlessly with my team to develop an advances fraud management system FEWS ( which was developed for under US $1 million, but which in field trials outperformed advanced commercially available systems based on neural networks and that cost multiples of this sum) . We became in a few years a truly international team with staff from Singapore, India, Philippines, Australia, the UK, Turkey, Korea and ,later when we went global from Spain and Poland. Over time they evolved into a “Dream Team” who overcame every technical challenge and did what others claimed was impossible.
I was also empowered to build a best-in-class data center, which I did in 1989 and its opening was a lavish affair attended by Pei Yuen Chia the global Consumer Bank head, who gave a memorable speech.
The base of the cards system was a commercial package called CardPac, but it was soon modified and improved beyond recognition, giving the Citibank businesses a competitive advantage in cost and functionality. The new cards team was called the RCC – the Regional Card Centre – and I was given unprecedented autonomy – my own budget, freedom to hire at will and even set up my own dedicated Data Centre, where I put my previous experience of being the Corporate Bank Data Centre Manger in Singapore to good use. I remain the only Citibank technology manager ever to be in charge of the data center, telecommunications, software development, deployment and business relationship management simultaneously. This complete control of every aspect of the technology platform was a major reason behind our ability to develop and deploy new products quickly. Very soon ‘RCC’ became a ‘brand’ within Citibank in Asia known for its excellence in execution and aggressiveness. The cards system named ECS consisted of the following modules:
Authorizations (OLA later named EAS)
Cards Management system (CMS)
Fees , interest calculations and statement generation
Billing Dispute management (DMS)
Fraud Early Warning (FEWS)
Customer work station
Not all the functionality was available from day 1, but were developed over a period of about 5 years as the business expanded and required more sophisticated functionality.
The first task was South Est Asia – Visa and Mastercard cards businesses were launched in 1989 and 1990 in Singapore, Malaysia, Thailand, Indonesia and Philippines. Thailand, which had a Diners portfolio in addition to Visa and MasterCard was also converted despite the conventional wisdom that it was impossible as CardPac could not accommodate its expanded card numbering system. At the time this was done I remember receiving an email from Pei Yuen Chia, the global consumer bank head, congratulating me and my team and ending with “ …. today Thailand and tomorrow the world”. He was to be proved right eight years later.
Thereafter North Asia – Taiwan and Hong Kong (which were Bank of America portfolios bought by Citibank), who initially planned their own regional processing center saw the results in South East Asia and were brought in. All within time and budget. Then Japan, where previously no financial institution had ever been processed offshore wanted in – the head of the Japan business Yashiro San, who reported directly to John Reed, enquired whether they too could be processed from Singapore – an unheard of event in Japan which prided itself on its own technology. The reason was the poor service and extremely high cost by the cartel which served Citibank. Rana told me to go ahead and encouraged me in his inimitable way “Yashiro–san is a good friend of mine. Don’t screw it up or I will nail you to the wall!”.
The Japanese sent a team that spent a week with us and were satisfied with our technology platform, service levels and quality standards and the I then went up to Tokyo for a week. My immediate conclusion was that the project was not possible mainly because the cartel that processed for Citibank and acted as a central switch for its transactions would not or could not co-operate in providing the technical information and support necessary for a complex de-conversion and transfer – this was in addition to the problems caused by the fact that Japan had its own unique telecommunication communication and interface protocols.
When I explained this to Yashiro-san, he immediately said he would transfer processing to the Sumitomo group – an unheard of change in Japan where cartels or Keiretsu are cozy relationships seldom or never broken. In addition we were fortunate that the cards business manager was Jim Milby, who was to later play a central role in the Global cards project. He was everything one could wish for in a business partner – supportive and meticulous in following up and soothing sensitive Japanese sensibilities. With such support we could not fail – though the Japan project was not the usual RCC ‘Charge of the Light Brigade’ but trench warfare successfully concluded in 20 months for $ 3 million – we not only provided more functionality and equal service kevels but reduced processing costs by 75% and speeded up responsiveness to requests for changes.
The ECS platform was also deployed to the Middle East and Eastern Europe by Dhruv Panchal and his team in another virtuoso performance. In addition to just deployment I encouraged Dhruv and his team to improve the operational performance of the software thereby creating some internal competition for Susan Kwek’s team. Once they had achieved this I merged the 2 teams under Susan – something that Dhruv did not forgive me for a long time. However, it positioned the central team for the global cards project which was to come later.
In 1992 even after the success of the RCC and the ECS platform was established in Asia, there were moves in the US to transfer the Asian cards business processing to the US center at Sioux Falls, South Dakota. The ostensible reason was to save costs – although technology costs in Asia (both data center and application system support) had been reduced – per card cost annually had fallen from US$ 12 to $4 as volumes increased, but the costs in the US with a 30 million card portfolio was $2. They ignored the fact that the cards proposition in Asia was more upscale and had more features, the database was fragmented across 12 countries, was customized for each country legal regulations and language (including double byte kanji support for Japan and Taiwan) and many other issues. This was typical of the know-nothing attitude that the technology heads in the US exhibited after the success of the CORE project, and driven by the desire for total central control. Fortunately, the technology head of the US cards was an exceptional technologist called Larry Fendley, who came to Singapore and spent a week with me. I showed him what had been achieved and he became convinced of the stupidity of trying to move Asia cards to the US platform. Instead he made a generous offer – ‘Come and spend 2 weeks with me in Sioux Falls and I will show you what we have learnt over the last 10 years. Bring your best people with you”
I chose the usual gang of suspects – Susan Kwek, Sai Tien, Geok Lian , Bee Koon and Dhruv Panchal. We spent 2 weeks in the bleak winter in Sioux Falls and Larry Fendley was as good as his word and showed us all the enhancements done by his team. We were particularly impressed by the credit scoring functionality, the use of a segmented database for faster batch processing but above all by the customer service platform – a single screen containing all the relevant customer data rather than the multiple screens that had to be navigated in the native CardpPac version. We absorbed it all, in between steak dinners and horse riding organized by Larry Fendley –we owe him a lot for the unselfish help that he gave me and my team.
After we returned Susan and her team spent 6 months designing the same enhancements for credit, faster batch processing and customer service on CardPac. We also added some innovations on our own – an automated billing dispute management system (BDMS), a parameterized Rewards system (which was so flexible that it was later adapted for use by the banking system as well), a parameterized system to discriminate authorization limits according to the Association merchant category codes, a cook book for co-branding that enabled the base system to be modified to launch a new co-branding card within 3 months an advanced frauds management system – Fraud Early Warning System (FEWS) developed according to specifications from Richard Parry and Vaman, the Regional fraud prevention team (who helped us in other projects as well ). This was the genesis of ECS – the Enhanced Cards System – which was subsequently rolled out across Asia, Middle East and Eastern Europe over 12 months to replace the plain vanilla CardPac system that we had first installed. Incidentally the costs were billed to each business annually and expensed, not capitalized, and this shows how modest was the expenditure for such a major enhancement. ECS gave the Asia cards businesses a clear advantage over the competition and one of the reasons that it became the premier cards franchise in Asia by the 1990s.
There was another reason as well – Rajive Johri, when he was the regional cards head, set up a Regional Cards Council consisting of the cards business managers and the heads of the group support functions – like credit and fraud prevention, marketing, fincon, technology etc. which he headed. He had not only his authority as the regional head but tremendous street cred with the business managers due to his track record in Indonesia and Hong Kong – this enabled him to control and guide a group of aggressive and hard charging Type A individualists. I was naturally in the Council as the RCC head. This council would meet every 4 months and would disseminate best practices, set priorities and define business objectives. I was therefore able to set my technology plan accordingly – the importance and contribution of this group cannot be over emphasized. No bureaucracy – just smart, likeminded and empowered managers setting common objectives and sharing information on successes and failures.
The RCC was now being recognized internationally and in 1997 I received an SOS from Ramon Cueto, the regional cards technology head for Latam on behalf of our Caribbean business – they were being processed out of Tampa and their technology team was unable to meet the corporate Year 2k mandates. Ramon was to prove a great supporter and help to RCC and we owe him a great deal. I flew to Puerto Rico with a small team and agreed to handle the conversion and the Year 2K mandates in 12 months and move the processing to Singapore – unheard of speed for the Carribean business used to lethargic support previously and very ambitious on my part given my other commitments in Asia. But my confidence in my team was not misplaced and they did it on time and within budget – this was the first time the RCC had ventured outside the Asia, the Middle East and Eastern Europe, and was instrumental in the later support from Latam for the ECS system over the US bankcards IBS system. In doing the conversion and transfer of support from Tampa to Singapore, I was greatly assisted by Shay Bridges and Steven Beatson from TI (the international telecommunications group of Citibank), who operated under the radar outside the usual telecommunication’s chain of command to ensure a smooth transition. In so doing they also convinced me o the virtues of TCP/IP over the SNA which was the RCC standard protocol for communications.
But in doing this (and also supporting the launch of cards in Poland and Hungary from Singapore) I made some powerful enemies in the US – chief of whom was Jim Stojak, the head of operations and technology in the US. He saw the RCC and me in particular ( he once described me disparagingly in public ‘ as that Asian cards guy’) as a threat to his plan to use the US Bankcards system globally – he was one of many people to aspire to global dominance in card processing technology and come a cropper! The US bankcards system was deployed in 1998 to 2000 in Europe and Latam with disastrous results which Stojak put down to non-co-operation from the businesses. I do not blame the US card technologists for this debacle. I respect them a great deal as technologists and they helped us when we were starting out. But their system was designed in the early 1980s with the technology available at that time to serve a simple product at very high volumes. It was not flexible enough for international implementation for small to medium portfolios. Larry Fendley the architect of the system knew this would be the result, but Stojak, who was ignorant of technology did not. Shortly after the international expansion project for the US bankcards system predictably failed (now renamed as ICS), Stojak summarily transferred Larry Fendley to a relatively junior position in the Fort Lauderdale technology office (collecting Latam requirements for cards processing !) as a way of forcing him out of the bank and in which he succeeded.
Stojak was responsible for my being removed from the Corporate Leveraged Property program (a group of managers handpicked for rapid advancement) and other assorted humiliations, and I was on the verge of being pushed out from the Bank in June 1999 . However, events were moving behind the scenes that I only learnt about later. Ramon Cueto was advocating the RCC ECS platform for the Latam businesses after the Caribbean experience and convinced the businesses to support this policy and courageously opposed Stojak’s plan for IBS. Finally Fernando Amandi (the stylish and forward looking head of tech and operations in Latam – we owe a great deal for his advice and support during the ECS project) had a confrontation with Jim Stojak and proved to him that the partial conversion of Latam card businesses was a failure and should be aborted. This crisis was brought about by the failure to adapt IBS for the Columbia card business with its complex two stage credit approval process and ‘mora’ interest calculations even after spending more than $20 million and two years. Europe was also in open revolt after their botched IBS conversion and the German card business had just received a 1 rating by audit – the lowest possible. Stojak had no option but to give in and told John Reed and Willumstad that IBS had failed due to non-cooperation from the businesses.
I then received a call from New York from Bob Willemstad, one of the triumvirate of managers who ran the bank, which completely changed the situation for me and the RCC. He simply said – “I know you are having issues – don’t resign as I have plans for your team”
Shortly thereafter I received the mandate to convert Europe (which was by then on the US Bankcards system with very poor results) and Latam partially converted to the US system again with disastrous results. Bob Willumstad himself flew down to Singapore for the kick-off meeting where he ended his speech to my staff by saying – :… every global technology project so far had been a failure. It is your job to succeed and give us the confidence that such projects could be done’.
This came at an opportune time as David Larkoworthy had been appointed by Mary Alice Taylor to head Asia Pacific tech and operations succeeding George Dinardo. He wanted to show quick results and had two objectives – sell RCC and outsource cards tech support to FDR and ‘save millions’ by centralizing all call center work in Darwin ,Australia, starting with the expensive Japan call center located in Tokyo. He had chosen Darwin (with the help of an amiable but incompetent Aussie called Peter Mills who was known for making elaborate presentations during Geoge Dinardo’s time that never led to any action) due to the fact that there were so many Japanese names in the phonebook. He neglected to find out that they were all wealthy retired Japanese businessmen more interested in golf than in working in a call center for minimum wages!
Predictably the project was an abject failure and Larkworthy was forced out by Simon Williams the APAC business head after wasting $30 million on an empty call center. Japan meanwhile had relocated to the cheaper southern island of Kyushu saving 40% of their costs. David Larkworthy lasted barely 2 years and left a trail of exorbitant ‘expenses’ for ‘taxis’ and ‘business entertainment’ in Bangkok and Manilla. These were investigated by Graham Dickson ( the regional control head who showed me his Excel spreadsheet of Larworthy’s and his HR head’s ‘expenses’) but no legal action could be taken as they had been approved by his own HR head, who was also fired. Larkworthy was easily the worst of a string of bad hires at the senior level at this time. But his timely departure saved the RCC.
Bob Willumstad asked me what I needed to succeed after the IBS failure and I told him I need a senior business manager to keep the businesses onside as there was no Cards Council in Europe and Latam to do this necessary function.
He handpicked Ashok Dutt ( the former head of the Consumer Bank in India) for this role and Ashok immediately negotiated the global license for VisionPlus and came up with a very generous (but pragmatic from the bank’s point of view) bonus incentive package for my staff over and above the normal bonus, paid according to achieving defined targets with dates. This was very generous and I and my staff were to benefit in 6 figure numbers, but the target thresholds were very tough and would set a historical record in technology execution costs, quality and timelines at a global level. Essentially 38 countries migrating to a to regional platforms but maintaining a single global architecture in two years and absorbing all conversion and technology build costs for less than $ 50 million. To put it in perspective, HSBC spent several hundred million dollars to build a global Vision Plus platform and ended up with partial implementation in a very small handful of countries. Ashok got Bob Willumstad’s agreement that the cost of this bonus was a miniscule fraction versus the significant benefit to the bank for an accelerated execution in record time and cost which he would plan with me and the team. The plan was pushed through over objections at the highest level by HR, which were over ruled by Bob Willumstad. Ashok also put in place a compelling engagement model between the technology team and the businesses, whereby the businesses felt they owned and controlled the plan for the functionality that the platform would deliver to them, and technology would enable and deliver this plan. In the process he also negotiated with the businesses that they should fully empower the technology team to deliver whatever they signed off on, and insulate them from negative political influences and interference from any part of the business.
Finally he convinced Bob to eliminate any interference and influence whatsoever from the corporate technology office and empower my team to deliver the project under Ashok’s responsibility. He made it clear to Bob that he would not undertake this responsibility without these conditions, and Bob wholeheartedly agreed and made it happen. This model was what enabled the technology and business teams to execute the project at lightning speed. Further Ashok and I decided to convince Germany, who were the most resistant to this project because of bad experience from already two years of a failed attempt to convert them to the US platform, to be the first to convert. He and I spent three weeks with the German business to convince them that the new model would be beneficial to them. Once Germany agreed we knew the “world” would follow. Having done this and set the project on a solid footing, Ashok left for greener pastures to run the Discover Cards in the US .
He was succeeded by Jim Milby (my friend and mentor from Japan days) who was by this time the regional marketing head for Latam and knew everyone in the business. He was ably assisted by his deputy Parkash Mamtaney. These two were indispensable for the success of the project and were indefatigable in helping us to get requirements, getting the businesses to test the software on time and changing their procedures to take advantage of the advanced features of ECS+. They lived out of a suitcase for 3 years traversing the globe. They and a secretary was all there was in the ‘Global Office” in contrast to the huge head office that came later for the Go to Common project. They not only kept the businesses aligned but kept all project records meticulously so we never had issues with Audits or regulators in 2 continents.
The ICC team also received excellent support from Fernando Amandi, Victoria Lopez-Negrete (the Latam regional cards head), Juan Garay and Jorge Lemus from the Mexico business and especially with Ramon Cueto the regional cards technology head, who had originally requested my help to rescue the Puerto Rico cards business from their Y2K dilemma. The Latam team were a pleasure to work with – technically proficient, dedicated to their work and fun to work with.
In Europe we received equally effective support ( more business oriented than technology oriented) from Surinder Sing, the reginal business manager and Hans Staudinger, who despite having an HR background was an effective regional coordinator. Very effective at getting the European businesses to co-operate in giving requirements and testing. I also picked Richard Hosang from the Latam business ( Richard was an experienced technologist who had previously been based in Miami and knew the Latam businesses well) and Ram Acharya to head the Latam relationship team. Richard was very happy to reconnect with his Chinese roots in Singapore. The Latam businesses were notorious for their individuality and their archaic banking regulations – like the credit application process in Columbia and mora interest. Despite this Richard and Ram did a super job (working midnight conference calls and 17 hour trips to the region) in collecting requirements and ensuring that testing went according to our aggressive project schedule.
As an aside, shortly after we converted the Citibank Mexican business in December 2001, Citibank acquired Banamex ( a large Mexican bank) who had a much larger card portfolio. Their cards system was designed to run on Tandem computers which were OK for authorizations and customer service but totally unsuitable for transaction posting, interest calculation and statement generation. It was also an extremely expensive platform and their system did not have a fraction of the functionality of ECS+. The sensible thing would have to also convert the Banamex portfolio to ECS+. Here politics and overarching ambition came in. Banamex management wanted to be the processing hub for Latam (the ‘RCC’ for the region) and insisted that the Mexican Citibank portfolio be converted to their system instead. A senior Citibanker called Al Stevenson, had been posted to Banamex by Victor Menezies ( the EVP for Latam) precisely to stop this type of nonsense, but despite the fact that he looked the picture of an East Coast WASP banker (always impeccably dressed ), he lacked perception, judgment and the character to withstand political pressure of any sort. So he agreed wholeheartedly with this shortsighted decision, over my furious objections and despite my sending Richard Hosang twice to Mexico City in 10 days to explain to the Banamex management the costs and the reduction in business functionality that would result from this decision. But all this fell on deaf years as Al Stevenson was more interested in protecting his lucrative pre-retirement contract by endorsing whatever was decided by Banamex management. It is a fitting comment on this type of ‘head office oversight’ that 4 years later the Banamex management themselves asked to be reconverted to ECS+ – after wasting tens of millions of dollars and losing significant market share due to the inability of their own system to support new products and marketing initiatives. Al Stevenson had meanwhile retired, with his reputation intact, and much the richer from the stock options he received during his stint in Mexico. C’est la vie!
For Europe I chose James Berti from the APPC as my project manager (who I had previously brought over from Australia to handle the data center consolidation), even though he had no previous application management experience. I did this because I had decided to locate data center operations in the European Data Center in Meerbusch in Southern Germany to be compliant with European customer privacy regulation. The ICC handled all system related modifications for the application and James even managed to negotiate the system passwords so that we could set up and modify the CICS and VTAM parameters and the source library from Singapore – an unheard of concession to an application team! The EDC staff did not like to be disturbed at night or on week-ends for this type of work, so our staff had complete freedom to do whatever needed to be done. Despite his lack of previous experience, James and his team, did a very effective job, but admittedly this was also due to Susan Kwek going the extra mile to help him out.
In my home base I was fortunate to report to Stephen Bird, who had joined from GE Capital as the Asia technology and operations head and who took over to repair the damage caused by David Larkworthy during his short but disastrous tenure. He was a common-sense manager who allowed me to handle my international obligations for the cards project without interference, while I also continued my Asia responsibilities for both cards and bank (George Dinardo had left by now and I was charge of the entire technology portfolio). It was obvious to me even then that Steven Bird would go far – he had the unique combination of a penetrating intellect, quick decision making and a low-key style.
I made some immediate decisions (besides renaming the RCC as he ICC standing for the International Cards Center – the global platform would not be based on CardPac but on VisionPlus – an advanced variant on CardPac but in which my team had no experience. The new version of ECS based on VisionPlus was called ECS+. I would use my core team from Singapore headed by Susan Kwek and not rely on external consultants for the design and coding, but they would be supplemented by outside contract staff. I would add some hires from Europe and Latam for the liaison work and hired staff with Spanish language skills as all training materials and documentation for Latam had to be in Spanish. Susan Kwek organized the central team into functional technical responsibility (fees and ineptest, new account acquisition, overdue accounts management and so on) and designed and implemented the required changes on VisionPlus in a remarkably short time. All this while supporting the growing business in Asia, middle east and Eastern Europe. I also took the decision to install an automated testing platform (Test Director and Winrunner) to make our testing process more efficient- this turned out to be a great success.
The budget to convert the 7 businesses in Latam was US $ 12 million and for 5 businesses in Europe was 8 million. This was later supplemented by $3 million for Argentina and Brazil, and US$ 5 million for Japan Diners, which was a late acquisition and a very difficult conversion ( I could write an essay on this ). A total of US$ 28 million – this was and is astonishing modest sum given that over US$ 1.4 billion was spent on the global Go to Common project that attempted the same thing for Bank and cards from an already standardized base.
We also decided to take the bull by the horns and convert the 2 biggest businesses in each region first – Germany in Europe and Mexico in Latam.
There were some initial hiccups as was to be expected in the first deployments in Germany and Mexico, overcome within weeks and the rest of the project was implemented smoothly on schedule with new conversions every 3 months. By 2005 all the businesses were converted including Argentina, Brazil (which was not in the original scope) , Canada, India and Japan Diners. At the endo of the process 38 businesses around the globe were on the ECS+ platform – new releases of of ECS+ were deployed every 4 months catering to new business requirements and enhancements like FEWS were being rolled out globally within 3 weeks, showing that rapid deployment was possible even with regional code bases. Costs were industry best standard despite the support coming from a high wage country like Singapore, and the functionality far in excess of what any other bank had An astonishing achievement not equaled in any other project in Citibank or as far as I am aware in any other bank. Ashok Dutt later told ne that when he had consulted with HSBC on their cards automation he found out hat they had spent over US$ 200 million to convert 4 small businesses in Asia like Sri Lanka and Malaysia. It is not hard to understand why this happened – a superb technical team with over 10 year’s experience on the same system, empowered and working to clear objectives with total business alignment. This was the culmination of 10 years of consistent effort and amply vindicated Rana’s trust and judgement in giving me a ‘blank cheque’ in 1989 and Bob Willumstad’s decision in 1999 to give me the authority to take ECS+ global.
The result was a standard system across the globe with 3 regional codebases but capable of sharing enhancements within and across regions within weeks. Costs worldwide were down to $ 5 per card per annum for both application support and data venter processing. Above all the businesses were happy with the new functionality and the regular new releases 3 time per annum. There were also never any audit issues.
Electronic Commerce Initiative
The ICC also pioneered ecommerce on the internet. In 1998 we implemented the world’s first electronic purchase using the Verisgn Certificate and we rolled out the payment gateway to the Citibank card businesses Germany and Belgium ( I remember the counterpart for our effort in our German business had a PhD in Computer Science). We did this with technical collaboration from IBM and Visa International and the whole project was paid for with a special grant from the Singapore government, who wanted to encourage ecommerce commerce and the development of the relevant technology in Singapore. The technical team consisted of Palaniappan from India, Lu Wang from China and Koh Wee Peng ,a Singaporean management trainee I hired and they did a superb job.. This was a signal achievement but the whole process of downloading an electronic wallet was cumbersome from a customer viewpoint and Citibank eventually moved to an easier process using a technology called SSL. However this pioneering effort built up a great deal of expertise within the bank which we used later in our electronic commerce
The days when my team and I deployed ECS in Asia and ECS+ in Europe/Latam were the best years of my professional life. For us it was like ‘Camelot’ – hard work, lots of fun and a shared objective which we knew we would achieve. We had lavish off-sites in Sentosa Island, Bali and Phuket and even Florida confident that no bean counters would question the expenditure. And we delivered what will be impossible today at even 10 times the cost to incredible timetables.
Two Failed Attempts to Hijack the ICC
Attempt to make it a Common Utility
In 2001 there was a move by some senior Citibankers in Singapore to make the ICC a “utility” for all Singapore banks. What prompted this I am at a loss to explain as the ICC was a competitive strategic advantage for Citibank and the result of 12 years of investment in people and technology platforms. Also the ICC was able to perform so effectively as there was a cards Payment Council setting priorities within Citibank in Asia Pacific, Cemea and the Middle East. Sunil Sreenivasan the CCO set up a Committee to study this and I was the only person in it with any knowledge of cards or payment systems and my view was totally ignored. The others claimed that this was in line with the Singapore government policy but I pointed out that it was the government policy to create the infrastructure and conditions for banks to set up regional platforms and not necessarily to create a Soviet style utility that will require a cumbersome bureaucracy to set priorities. Off the record I was approached by a senior Citibanker who told me that “we can all become rich” by doing this , to which I retorted that this was not the way I wanted to become rich.
When I found that my views were being ignored (indeed expunged from meeting minutes) I had no alternative but to report it to Jim Milby who was my business manager for the global cards project, and he promised to take it up with Bob Willumstad. He must have done so because the whole project was suddenly dropped and the people who claimed to be its sponsors themselves disclaimed it.
Attempt to Create a New ICC
The second attempt was more insidious – Fritz Seegers, one the mediocrities that followed Rana Talwar as the Asia Pacific Consumer Bank head , had grand ideas of making a mark in Asia as a banker ( he was also at one time married to former Indonesia President Sukarno’s daughter). He left Citibank and joined Barclays Bank as the Global Consumer Bank head. Venky Krishnakumar also joined him as a consultant.
Fritz had a vision of duplicating Citibank’s very profitable and upscale consumer business and poured tens of million dollars into Indonesia and India to achieve this. He took as much talent as he could from Citibank and also tried to recreate the International Cards Center technology group. So he lured across Chan Char Ho (who succeeded me as Technology Head), Phiilip Chow (who worked for me as bank technology head), Goh Soon Cheng ( the main technology talent on the Systematics system), Barry Tan, Lim Tee Tsian ( our Japan cards tech expert) Goh Soo Chin , Agnes Teh and many others. All were given big salary increases and sign on bonus and it was featured in the Singapore Press that Barclays was planning a Global Technology Hub in Singapore.
But things did not go according to plan because Fritz was not up to Rana’s stature as either a leader or a banker, and the motely group of ex Citibankers he had gathered around him did not have a shared vision but were mercenaries in it for the money. Soon the consumer businesses in India and Indonesia started bleeding cash amid huge credit losses. The inevitable soon happened and Fritz was fired and Barclays divested themselves of all their international consumer businesses except for South Africa.
What happened to the attempt to create a global technology hub is interesting as it throws light on the ethics of the people involved and also reads like a bad detective story. Citibank auditors found electronic traces of huge file transfers of Citibank technology libraries and alerted the Barclays auditors who did a surprise check and found out that the entire library of Citibank systems for banking and cards( the result of many years work and worth millions of dollars) had been stolen – there is no other word for it. Cha Ho and Phillip Chow , who were responsible were summarily fired and eventually every technologist hired from Citibank, who were to create a “New ICC” lost their job. This was a sad end and shows the morality of people at the highest level, who having spent their entire career in Citibank, were prepared to compromise their principles for money. Neither Rana or I ever did anything like this when we left Citibank and joined other institutions despite the fact that we were approached by former staff who wanted to work with us.
The Systematics project
In 1993, following the example of the success of centralizing cards processing, Rana decided to do the same for banking. This was a much more complex than cards, where a degree of standardization is imposed by the Association rules. But the banking businesses had varied and different processes in each geography, partly due to local conditions and legal/reg and partly due to the idiosyncrasies of the local technology team. He knew that Tsui Kwok Wo, the regional technology head and I, as his number 2 and head of cards technology lacked the ‘gravitas’ required for this immense undertaking so Rana decided to get an international ‘heavy weight’ to head this project.
Rana’s style was never to do anything by half measures and so he dispatched his trusted HR head Lou Sanandres to the US to get headhunters to find the best person available. They found George Dinardo – recently voted the Man of the Year by Computer Weekly and responsible for setting up a computer center for Mellon Bank, which also processed for 30 other smaller banks. An ideal background for the Asia project. George was in the process of moving to his retirement home in North Carolina. Lou Sanandres and Rana had to persuade him to come to Singapore for a 5 year contract by showing him that it would be a world first for banking to be processed centrally in 12 diverse businesses spread across Asia and the middle east and also, in George’s own words “by plying me with stock options” – this was in the days when this particular instrument had real value!
George’s arrival in Asia was like – words fail me – a Tsunami or whirlwind. Bull in a China shop is too mild. Ignoring past precedents, he rode roughshod over the wishes of the business managers and decided – no dictated – what was to be done. All banking business processing was to be centralized in Singapore on the APPC platform on the Systematics suite of applications which would be suitably enhanced for Citibank requirements. Each business was to be allowed only 2 enhancements over and above this. He would decide when and how each business would be converted etc….As was to be expected this did not sit well with the Alpha males who were Citi business managers who were used to taking orders from Rana but not group office bureaucrats – George ignored them all claiming he had a mandate from Rana and John Reed (and God?) to convert Asia into a common platform. In the event only Jerry Rao and Ashok Dutt in India were able to resist George’s onslaught and all others either succumbed to the pressure or acquiesced in the inevitable.
George was a former US Navy destroyer captain and also the most politically incorrect manager I have known – he would not hire senior female technologists to report directly to him (after running through multiple secretaries who left in a flood of tears, Lou Sanandres had to bring in a tough Eurasian lady to handle his work) and liberally sprinkled his conversations with 4 letter words and called his staff sons of bitches. All of us were fired multiple times, which he would forget about next day – but he was very generous with pay, bonuses and stock options and never held a grudge for more than a few hours .He also backed his staff 100% in our frequent tussles with the businesses, auditors and corporate fincon .He was a difficult man to work for (once I nearly came to blows with him, if not for the timely intervention of Lou Sanandres but George did not bear a grudge against me and gave me my highest bonus ever that year) Looking back , he was the right, perhaps the only person, who could have done this job, thus vindicating Rana’s judgement of the type of person needed. In retrospect, I treasure that experience of working for a world class technologist and, under the gruff exterior, a decent man.
George oversaw an enormous investment in technology – both data center and in applications. He would approve only IBM mainframe gear and to service the Citibank account IBM posted a very experienced account manager from New York called Don Hanson. George’s forethought in investing in IBM platforms and technology was vindicated when Hitachi and Amdahl-Fujitsu (they had cheaper alternatives in large water-cooled mainframes, and Amdahl-Fujitsu were in fact chosen by the corporate bank) dropped out of the mainframe race when IBM introduced powerful air cooled CMOS processors. The corporate bank, whose investment was now at a dead-end, later transferred their mainframe workload on to our computers.
The task ahead was daunting – convert 12 businesses into a new standard platform based on Systematics but significantly enhanced for Citibank, and transfer data center processing to Singapore simultaneously. All this had to be done in under 5 years to meet the looming Y2K deadline that mandated that all systems had to be Y2K by mid 1998. But George and his team were up to the task – George asked me to write the plan and prepare the MEP (US$ 60 million for applications development and software licenses, a massive expansion in Data Centre capacity and telecommunications – an astonishingly low figure given the $ 1.4 billion spent later on the Global Go to Common project with less scope but admittedly greater geographical coverage). I was chosen to handle the Data Centre expansion (we moved to new state of the art facilities which are still in use today and is one of Citi’s largest data centers worldwide). I was also responsible for the conversion of Singapore (the lead country) and come with the pilot version of Systematics converted to Citibank requirements, and then convert Malaysia and Japan.
3 seasoned technology professions were also brought in to handle the other countries – Jagdeep Chahal from IBM Australia to handle Thailand and Australia, Dan Hartman from Bank of America to do Philippines, Taiwan and Hong Kong and Sergio Araneda , a lateral Citi transfer from our business in Korea to handle the middles east and Eastern Europe. Yew Seng Tan was George’s technology architect, responsible for keeping us compliant with CTO guidelines.
There were 2 application teams – one under me to handle the 3 businesses entrusted to me, and one common team under tri-partite leadership (Dan oversaw product processor development – his version was later adopted by my team when we combined the two different releases afire Y2K) to handle the other businesses. This team was based in Singapore and Manilla and included some excellent technologists like Agnes Santiago. This may have seemed a clumsy arrangement but it worked. Most of the key staff were seasoned Citibank veterans, supplemented by 5 key hires from the US who were experts in the Systematics codebase (of whom the most outstanding technically were Dave Hale who went to become the chief architect for mainframe banking in the US and Tom Phillips, an expert on the Systematics ALS loans module. They were supplemented by the usual army of contractors from India and the Philippines.
As an aside Australia, under the extremely arrogant manager Brian Clayton, tried to go it alone with support from Alltel the vendor of Systematics and messed it up so badly that the business suffered a near death experience. They later had to be rescued by Jagdeep Chahal and the regional team. But this is another story.
Compared to the ICC approach for cards automation, which was like a rapier, this project was executed like a blunderbuss. Full steam ahead and damn the consequences – it says a great deal for the technical prowess of the team that there were no major blunders, that the project was completed roughly on schedule although we exceeded our budget by US$ 12 million. Just as a comparison, Citi budgeted US$ 500 million for Y2K compliance globally and Asia did not have to dip into this honeypot for the Systematics codebase for banking or ECS+ for cards.
When the Sytematics project in Asia was well underway, and clearly going to succeed, Europe started to show interest. This alarmed the New York Office and to show their relevance to the process, they came up with their own banking solution – a package called Sanchez (after the 2 brothers who developed it) and claimed it was superior as it allowed for on-line real time posting of transactions. This argument can be traced to the New York technologist called Mike Veale who, while he had no hands-on experience in any real system, was a master at elaborate ‘architectural’ presentations that seem to impress higher management so much. This argument was spurious as real time posting was unnecessary for consumer banking needs, where the illusion of real time processing can be created at touchpoints like the ATM, teller, call center or internet using shadow transactions files. The transactions during the day can then be posted safely at night (when there can also be reversals of erroneous postings) as interest was only calculated on a daily basis, unlike some corporate systems involving large sums where intraday interest was also calculated.
The results were predictable – huge teams were set up in London and New York for requirement gathering, project management and with the vendor handling business specific changes. There was never any real progress made and with the Y2K deadline looming, the project was quietly cancelled without any fanfare.in late 1998. The pilot business to be converted was Hungary with only 2000 customers and even this was not achieved despite the expenditure of US$ 17 million – many times more that this business would ever have made even if it had been converted. In the event only the few real techies involved were axed mainly from the London office – Mike Veale and the New York bureaucrats went safely back to their real job of making elaborate presentations to impress management with their technical prowess.
The Systematics project once completed left the Bank in Asia and CEMEA with an enhanced platform with a high degree of standardization and full Y2k Compliance by mid 1998 There were some variations in the codebases because 2 teams had developed the base code. In the event this was easily corrected after Y2K by a project called the Enhanced Banking System (EBS) and the 2 teams were merged under me and the chosen platform was the one developed by the common team under Dan Hartman. This positioned the bank for future developments which were sadly misused as will be explained in the last section of this article.
The Asia Pacific Data center (APPC)
As I mentioned earlier, when I set up the RCC, I was given the signal freedom to set up my own captive data center, which I named the Asia Pacific Computer Center (Lou Sanandres once joked that APPC stood for Ajit’s Personal Computer Center). This gave me the freedom to proceed at a faster pace and design a system that was optimized for performance as well as functionality. I was the only technology manager in Citibank to have the responsibility for applications development, the data center and telecommunications simultaneously. In handling this responsibility, I put to good use my former experience as the Data Center for the Corporate bank in Singapore. I decided to to order the latest equipment (largely IBM) and more importantly recruit the best possible team. The most important members of my team were Leong Soon Peng, the technical team head, and Esther Goh the data center operations head. They were the best that that Singapore produced – technically proficient, hardworking, dependable and utterly loyal. I could not have done it without them. They were ably supported by John Harding (a maverick American who was my technical consultant and Jack of All Trades), Ricardo Nuqui ( from the Philippines), Koh Chong Chia (a communications technology wizard),Kim Hwa (our expert in CICS), James Berti (from our Australian business), Kim Hwa, Martin Loh and others. Together we built a superior and robust foundation that was later tasked to handle banking support across Asia as well with the Sytematics project . When we did the Latam cards project, the APPC also extended support to Latin America , as it was deemed that there was no suitable data center in the region.
This grew into a massive, world class facility – the largest in Asia outside Japan, with superb technical infrastructure and 6 Sigma service levels. From the outset we maintained a perfect 5 rating from audit. We also pioneered many firsts – the first to use mirrored mainframes connected by an IBM technology called Sysyplex and CICSplex, the first to use a communications protocol called MQ series, the first to use a dedicated non-switched fiber optics connection for data center back-up and so on. But we were not blinded by the new – we remained with an older IBM communications protocol (Called System Network Architecture – SNA ) over the new protocol called TCP/IP as the former was more robust and manageable despite immense pressure from our communications unit. It was only when TCP/IP had matured and the tools were available to manage it that we switched.
It may be a surprise that in all my essay on regional and global processing I never mention telecommunications. This was for the very good reason that John Reed our Chairman, ably assisted by his chief technology officer Colin Crook, had decided to build one of the second corporate global telecommunications networks (after IBM). This they did a decade before any other bank and it was entirely funded by New York – this also benefited by the global expansion of fiber optic submarine cables which increased by multiples the bandwidth available. When we needed to process for the Middle East and Eastern Europe, and later Europe and Latam, all we had to do was “plug ‘n play”. Connectivity was never an issue. The GTN was a fundamental pre-requisite for international processing and we could not have done it without the GTN . John Reed’s vision in investing in this was amply vindicated.
When I eventually gave up the data center in 1999 to be merged with the Corporate bank center, it became the foundation for the combined facility and today runs the entire bank’s data processing. The Corporate bank facility was eventually closed and processing moved to APPC. The APPC was one of the less obvious reasons for the success of the International Cards and Systematics projects.
I had no part in the development of the India technology platforms as they were never a part of the regional projects and independently developed their own systems. There were a variety of reasons for this – the Indian business had its own unique requirements, there were severe constraints on technology imports till the 1990s, etc. But I suspect that the real reason was that, given the wealth of their technology talent, they felt that could do it better – and Jerry Rao the business manager fully supported this view. And I believe they were right at least for the India business.
Under a very experienced technologist from IBM called Ram Bhagwat, the India team independently developed an entire suite of applications for both cards and bank – a remarkable achievement. Under different circumstances they and not Singapore may have developed the global systems. That this was not so was due to a number of constraints they faced – the inability to import the latest hardware and software till the 1990s ( for a number of years they had to use an obsolete computer called PRIME) , lack of exposure to international requirements such as multi-currency and much lower funding levels. But nevertheless their achievements deserve to be recognized. The India bank business is still on its own home developed application platform and it was only the cards business that was converted to ECS+ in 2005.
As it was India technology still made a considerable impact on the international business by its ideas and products (like “unfixed time deposits) but mainly though sending talented staff, like Arvind Rao who became the guru of loans processing for the international businesses. However the two best examples of the talent and ideas coming out of India were the data analytics system and the mutual funds product processor.
S Ramakrishnan (SR from the elite IIT stable) came to Singapore to develop and deploy the the analytics system called the data warehouse for bank and cards. This was one of the first banking implementation of pre-determined data cubes that were the basis for analytics. SR used off the shelf components that were commercially available. This data warehouse was used throughout Asia and was a necessary compliment to the ECS+ and Systematics platforms that lacked a data analytic capability. It was later taken up by Europe and Latam. At the same time the US cards business under Roberta Arena contracted with Lockheed Martin to build a data analytics capability for their portfolio. The result was a cumbersome system that used proprietary technology specially developed for this application. It was touted as a global solution but eventually was never adopted outside the US. The cost of the US system was $60 million versus $32 million for the Asia project.
Similarly, Rhagavan (Rags) came to the Singapore regional office with a mutual fund processor, developed for the India business, where I helped to transfer it from its Oracle platform to the IBM DB2 platform. It was a complete solution for managing funds and was later extended to cover share finance as well. This too was taken up by Europe.
Citibank India was also fortunate in having two successive business managers who were interested and knowledgeable about technology and who encouraged innovations – Jerry Rao and Ashoke Dutt and both went on to achieve a great deal in technology after leaving Citibank. Jerry Rao, set up a software and call center facility called Mphasis, which he later sold to a US firm called EDS. Ashok Dutt went on to run Discover cards in the US and now runs a data analytics company in Chicago.
The Year 2K boondoggle
From the 1980s all computer technologist knew that they faced a problem when the century changed on December 31 1999 to January 1 2000. This was because hardware microprocessors and software systems used 2 digits for the year in the date field. So computer applications with a 2 digit date field would have seen January 1, 2000 as 01/01/00 – indistinguishable from January 1,1900. This would have caused chaos in the date and above all interest calculation routines. This saving of 2 digits done partly to save on storage (that was expensive in those days) but mainly because no one thought so far ahead and expected their systems to be still in use 20 years on. COBOL, the main language in which banking applications was written, had this problem. The job of locating and changing the thousands of instances of a 2 digit year field into 4 would be tedious and time consuming but not rocket science.
However, this problem was seized on by a few technology scaremongers in the popular press who forecast a Y2K disaster of huge proportions when all banking systems would not work let alone the electricity supply, water systems etc, all of which depended on microprocessors to some extent. Time magazine and Newsweek and others carried stories predicting a disaster of major proportions. In the US there were even ‘survivalists’ who stored supplies of food in mountain cabins preparing for the day of doom. This panicked the public and more relevantly the senior management, who were told by their legal teams that there would be catastrophic litigation liabilities – the technologists acquiesced in this, probably happy to be in the limelight for once. The results, especially in the US was a huge overreaction driven by staff lawyers, not technologists.
In the Citibank case a central budget of US$ 500 million was set up to rectify this and a corporate office was set up to monitor compliance by mid 1998, after which there would be a freeze on new developments. Every technology office had to have a Y2K compliance plan, monitored first monthly, then weekly and in 1999 even daily (!). All businesses used this pot of money to rectify their systems and replace hardware whether justified or not. Only the Asia business we did not take a dollar for either cards or bank – we were able to do the Y2K rectification within our normal budget, as it was a point of pride for George Dinardo and his team to show the whole thing was over hyped. However as the head of the computer center APPC, I could not resist the temptation to replace my water cooled mainframes (even before they were fully depreciated) , as they required enormous amounts of power and air-conditioning capacity, with the new air-cooled IBM CMOS processors, which the APPC was the first to install in Asia ( the saving in electric power and air-conditioning actually justified the switch but it would not have been approved under normal circumstances).The order I placed, one of the largest in Asia, merited me a one-on-one breakfast with a direct report to Lou Gerstner, the Chairman of IBM during the annual IBM business conference in Tokyo!
In the event Y2k went off without a whimper – no incidents in Citibank or US banks, which had spent a fortune and equally so in Banks in Europe and Asia, which had spent a fraction of this. The only beneficiaries were the technology hardware companies and the Indian software houses, all of which made fortunes and contributed to the tech stock price bubble, which predictably crashed after Y2K. This should be a salutary lesson not to overreact to popular hysteria fueled by the press
Global Go to Common – the Mother of All Technology Projects
After completing the global cards project I decided to leave the bank as I was exhausted physically and mentally by the work and excessive international travel. In any case I would have had to leave as I would have definitely opposed what was to come – a single system to cover the whole world including North America, as impractical and unnecessary – and would consequently have been forced out. I joined a local Bank called DBS as the Managing Director for consumer bank technology and operations, where I realized how superior were the Citibank technology platforms.
My feeling was that we had done what was necessary on the back-end and should now concentrate on the internet and mobile applications and cloud computing.
By 2008 the Citibank International Consumer platform was in good shape – the card businesses worldwide (except for the US was on the ECS+ platform (arguably the best cards system in the world), Asia, the Middle East and Europe was on the EBS/Systematics platform and Latam was on derivatives of the old CORE platform. The only significant variations were in the front-end customer service platform, wealth management and in customer analytics. Even here Asia had developed advanced solutions that could have been tweaked for international release.
There was a project called Eclipse to convert the banking customer service platform to the Cards Workstation architecture, which was superior technically so that there would be a common customer interface – this modest program was given the grand title of Eclipse. Then DK Sharma, the Asia technology Head, launched the even grander named Rainbow project (DK was a master salesman), which was essentially the continuation of the old process of continually upgrading and standardizing the components of the consumer platform in successive releases between different countries across Asia.
This was where the trouble started – DK Sharma and Mark Terkos the Global Technology Head decided to expand roll out this program globally, including the very different North American business. This was scope creep of huge proportions as the Asia platform consisted of the following components:
(It is interesting to note that all the above applications, except the last 2, originated in the Singapore Asia regional technology office, developed often in the teeth of opposition from the corporate technology office the US.)
This is a complex multi-layered suite of applications consisting of hundreds of millions of lines of code. What was promised was a single version of all modules covering all the countries in the world including the US – what hubris (or probably ignorance and lack of hands-on technology experience) caused this promise to be made I will never know.
They also promised enormous savings in operations processing, technology maintenance expenses, instantaneous roll out of new projects globally (never mind the business rationale), and unspecified new business revenue – $ 4 billion over 10 years! The cost of the Global Go to Common project was equally gigantic – US$ 1.4 billion consisting of multiple projects and MEPs for all regions. Ignored was the fact that the Citi back-end technology platform was already well standardized and it only required successive releases to continue the convergence. Also overlooked was the obvious fact that the North American business was very different to the upscale Asian franchise (in fact a study showed that there was only 17% common functionality between Asia and the US), and the fact that the focus of advances in banking technology had moved to the web, the cloud and mobile applications from the back-end. All of which were common knowledge but overlooked by DK Sharma and Mark Terkos probably mesmerized by the opportunity to present the first billion-dollar technology project to the Board and the kudos and financial rewards that would follow.
DK Sharma’s Bollywood good looks and persuasive presentation skills were not matched by his execution ability, as was well known by all his colleagues (during the 8 years he worked for me he never completed a single project from beginning to end). DK and Mark Terkos persuaded Jonathan Larsen (The Consumer Bank head), who took it to Medina Medora who approved it and finally the Board approved the project. DK Sharma moved to Florida as the International Technology Head to implement the project. The results were inevitable and predictable.
After global kick-off meetings accompanied by a reiteration that Citi was on the path to technology Nirvana the project ground down to interminably long and complex requirements gathering sessions dominated by US participants (there was a requirement meeting in Singapore with over 200 US participants – when we did ECS+ for cards we only consulted a few experts from each region) and the smaller businesses was shut out. The old simple accounting method of directly charging each business for the changes made for it was superseded by taking all expenses centrally and allocating a “head office tax” on each business – this naturally removed all constraints and the businesses asked for the moon since they know they would in any case be allocated a pre-determined sum. All relationship between functional value and cost was lost.
The number of staff employed and the head office bureaucracy ballooned with a cast of thousands in Singapore and India – the only real beneficiaries of this monster project in the end were the Indian software houses supplying contract technical staff like Tata Consultancy Services and WIPRO. They made a fortune.
In the past I had a simple technology organization – a relationship management team of business savvy technical staff, who oversaw requirements gathering, acted as domain experts and standardizing requirements between countries and regions and were in charge of testing. The technical team were organized according to functional expertise – authorizations, fees and interest etc. who handled all the technical work. Simple but effective, cutting down on the number of internal interfaces and communications. Now the teams were split up into numerous stand-alone units -requirements gatherers, business analysts, functional designers, coders , testing teams, deployment teams etc. stationed in the US, Singapore and India The internal communications and interfaces ballooned (accompanied by unnecessary CYA emails as each manager tried to justify that he did his bit in a process where there was no end-to-end visibility on responsibility ) , endless staff meetings across time zones to co-ordinate the teams and many layers of management and complex reporting structures made progress painfully slow.
The results were not difficult to predict – after 5 years of continuously increasing complexity accompanied by escalating delays and costs and only partial implementation the project was finally shut down by Stephen Bird (who was by then the head of the Global Consumer Bank) in 2015. The $ 1.4 billion, authorized in the original series of MEPs had by now been spent and a further $ 700 million was authorized but fortunately Steven now diverted these funds to mobile and cloud applications. He was prompted by complaints from the profitable Asia franchise that they saw no value in the project and their urgent requirements had been frozen for years because of this project. This was a sensible decision taken quickly by an intelligent and experienced executive who saw through the technical smoke and mirrors that justified this huge waste of shareholder capital. In his gentle style, he blamed no one or pointed any fingers – (although there was plenty of blame to go around) but merely said that the bank would in future concentrate their technology resources on Fintech, internet banking, cloud computing and mobile applications– sensible but this should have been known 5 years earlier especially to the so-called experts like DK Sharma and Mark Terkos.
The project was buried without fanfare and no postmortem was held on why the bank spent a fortune and 5 years on this project. Neither was there any effort to identify if any of the promised $4 billion in savings and revenue in 10 years had been realized – the simple truth is that there were negligible returns. Whether this was the best course of action from a shareholder perspective is best left to the judgment of the current management. The main actors – Jonathan Larsen, Mark Terkos an DK Sharma – have left, no doubt with highly lucrative golden parachutes. The axe instead fell on the senior technical staff, who were given an impossible task and tried their best.
Citibank which had led the banking world in technology for 40 years – the first ATMs, the first international banking system (CORE), the first global data center serving 3 continents (APPC in Singapore), the first global proprietary telecommunications network for a bank (the GTN) and the first global cards and payment system for 38 businesses (the ECS+ system developed by the ICC in Singapore), now had another first – the first billion dollar banking technology disaster ($1.4 billion to be precise). It is not hard to find the reason for this fiasco. The former firsts were achieved under the management of seasoned professional bankers, who understood the value of technology and the competitive advantage it brings, and were willing to provide patient long term funding for technology. They also selected technologists with a proven track record and not just on a few superficially impressive presentations. They were – to use an old fashioned Biblical term – Wise Men.
They were succeeded by a management who were not real bankers but former consultants, deal makers and marketing types who were looking for a quick win, and latched on to the Go To Common project as an easy way to cut costs in technology and operations by cutting headcount, and outsourcing future development in technology – $ 4 billion of savings over 10 years to be precise was promised to the Board to win approval. To achieve these dubious objectives they were convinced (or conned?) by two technology gurus, who had no real track record in project implementations, but were expert at selling grandiose ideas, however spurious (I have seen some of these presentations and they can only be described as works of art (or fiction?) in PowerPoint! ) – and they predictably failed. There was no legacy from the massive expenditure in this failed mega-project unlike in the in the earlier CBS failure. No savings in operations costs, not one new application or new banking product. Nothing . After 5 years of effort and the expenditure of $1.4 billion we reverted back to the status quo ante. A pure waste of time, money and valuable talent. Imagine what could have been done with $1.4 billion in 2009 – we could have, for instance, bought ownership control of PayPal and been a leading player in global payment systems today. The irony is that the people responsible – both bank management and the senior technologists- were either promoted or retired with a generous severance package – the golden parachute. The real price was paid by the working level technologists, who faced a brutal downsizing (with minimal benefits) after the fiasco, and diminished career prospects for those who survived. And of course the long suffering shareholder of Citigroup. Life is not always fair – or as the cynical French say – “C’est la Vie”.
I will now spell out why the whole project was unnecessary and why it failed as it is a salutary lesson in technology governance:
The sensible thing would have been to keep regional code bases of a common architecture, with 3 or 4 releases a year and an exchange of business relevant enhancements within and among regions in the releases This is common sense not rocket science and is, in fact, what has been now adopted for ECS+ for cards, which is the only global application now left. It is a fitting comment on the Go to Common fiasco that 6 months after the project was cancelled by Steven Bird, that the US decoupled its banking system from the ‘global’ version saying it was impractical and prohibitively expensive to stay in synch!
It is important to learn the lessons of this costly failure, a repetition on a much larger scale of the failure of CBS 35 years ago, as in the words of the great historian Lord Macaulay – “those who forget the lessons of history are condemned to repeat them”.
I should mention that while the above saga was going on, Citibank’s global business and ambitions in consumer banking imploded – Brazil, Argentina, Chile, Peru, Columbia and the Caribbean in Latam was sold off, as was of all Western Europe (including Germany, the crown jewel of the international business, on which so much effort and treasure had been spent) and Japan – the latter with the Diners portfolio consisting of 500,000 of the most upscale customers in the country. When Citibank started the Japan business and John Reed recruited Yashiro-San to head it, the latter had only one question – “does your bank have the stamina to see this through in the long run?” After John Reed left the answer became clear – we had no vision but to reach quarterly earnings targets even on a shrinking base. There is nothing now left of Walter Wriston and John Reed’s original vision of a global and profitable upscale Consumer franchise differentiated from the competition by world class customer service, credit procedures and technology. There was a window of opportunity to have achieved this in the late 90s if Rana had been retained as the International Consumer Bank Head based in London .
John Reed must share part of the responsibility for bringing in Bill Campbell (a tobacco salesman to head the Consumer Bank) and Larry Phillips from GE as global HR head. These two decimated the senior management of the bank (Pei and Rana either left or were forced out) and brought in a consumer package goods marketing executives who knew nothing of the complexities of banking. They ran the franchise built over the past 20 years into the ground. In fact, in 2001 Bob Willumstad told me that after they took control they could not find a single competent banker in the international business – what a damning indictment on a bank that was once the incubator of talent not only for itself but other institutions as well. In Asia Rana was followed by a succession of mediocrities like Brian Clayton, Simon Williams and Fritz Seegers. However, the Asia franchise was robust enough to withstand poor management but the days of innovation, superior customer service and growth were over.
The overriding factor in this drama was also the merger with the Travelers group in 1998. The management of Travelers, who took the driving seat, were more interested in deal making and short term profits (Bob Willumstad was an honorable exception) rather than in Walter Wriston’s vision of building a global financial brand. From the point of view of technology for instance they downplayed John Reed’s drive to achieve 6 Sigma quality and the other similar initiatives. This strategy worked briefly, and I remember at a global meeting for senior managers in 2001 Sandy Weill proudly announcing that we had overtaken Exxon to become the ‘most profitable corporation on the planet’ with $15 billion in profits. The financial press lionized Sandy Weill but veteran Citibankers knew that trouble was brewing. So called financial ‘heavyweights’ like Robert Rubin, Bill Clinton’s Secretary of the Treasury, who was on the Board, either did not know or care what was going on. This strategy worked briefly but as was inevitable led to the subprime mortgage and CDL crisis which led to the decimation of shareholder capital. The bank barely survived with a shrunken capital vase. $ 650 billion was lost in this debacle dwarfing the $ 1.4 billion wasted on the Go to Common project to set things in perspective. The share price fell by over 90% from over $50 to at one time under $2, and has still not recovered ( the fact that it is now trading in the $50 range is because there was a 10 to 1 stock conversion). Today Citibank is driven by the tyranny of meeting Wall Street analysts short term quarterly expectation rather than achieving an overarching vision. We are now left with only the profitable and upscale Asian business built under Rana in the 1980s and 1990s – profitable but not the global bank originally envisaged and operating in saturated markets with limited growth prospects.
As for Citibank technology, staff morale is down (I keep in touch with my people at Chinese dinners and golf games) and the best have left or moved on to other areas of the bank – people like Susan Kwek (currently Citi Singapore O & T head), Dhruv Panchal, Goh Soon Cheng, Esther Goh, Goh Soon Peng , Derek Mollison, Yong Choon Bin, Yap Bee Koon, Koh Chong Chia and Ong Gok Lian. For those who remain this is just a job like any other– a far cry from the days when it was a joy to work incredibly hard to achieve a shared objective. In short Camelot. There are still a very few of the old stalwarts left like Simon Chiang, CP (not DK!) Sharma, Arvindh Rao, Sundar and Reddy. The bank should treasure them.
I would like to dedicate this memoir to the technology professionals I worked with in Citibank for over 2 decades. They were the best and I consider it a privilege to have been associated with them.
Singapore March 2018