Sunday, April 13, 2003
Centralisation - A realistic strategy or not?
Corporate strategies come and go. For most of the 80s and the early 90s, many corporates acquired other, smaller, companies creating far-flung global empires where each operation ran largely autonomously. As I joined Citibank in 1992, it was only just emerging from an enormous problem triggered by a mixture of the Savings and Loans problems in the USA, a cost base that was expanding faster than revenues were growing and an excess of staff performing duplicate functions. If you check back on business magazines from then, Fortune or Forbes perhaps, you'll see that one reason Citi survived was that the CEO, John Reed, implemented a series of significant cost-cutting and revenue driving progammes; at the same time, a Saudi prince bought $1 billion worth of shares at around $8.25. As I left Citibank, those shares were worth perhaps $10 billion, only seven years later. Something worked. Banks are a good example of the "expand at all costs" approach that was popular then, and Citibank was certainly not unique. The consequence though was operations in each European country with considerable duplicate functions but all operated autonomously. I had no part in developing the centralisation strategy - that was all the work of other people who, fortunately for me, I came to work with much later. I worked with a team developing the strategy for how the bank would consolidate and restructure from 1995 to 2000. A fascinating project that taught me much more about every leg of the bank than I could possibly have learned any other way. A generic example of a duplicate function might be trading in foreign exchange. Each country has an IT system, each country books trades, each country receives money and pays money, each country has brokers, sales people and product people. Each country deals with the in-country office of other corporates - the people who need to buy or sell in different countries to pay their bills (salaries and so on). Those are all pretty obvious costs ... but there are slightly more hidden costs of duplication in this case: individual brokers, trading on the bank's account, may take opposing positions in the same currency, a broker long one currency may be oblivious to another broker short the same currency; one broker may have more currency than he needs but not know that another broker needs some to settle a customer transaction who then goes out in the market to buy some more; the central bank will demand some capital to allow the positions to be held and so each country is obliged to tie up important capital. Adding all of these up creates both huge, stable expenses (based on the cost of staff, the cost of infrastructure and the cost of capital) and then enormous swings as cash inflows and outflows are managed. All of the foreign exchange staff are backed up by teams of people who manage cash flows, money market deposits, collections and more exotic derivatives that take advantage of any of these funds flows. From the early 1990s, banks took an axe to these costs: they consolidated their trading books, implemented back to back trades to ensure that risks were centrally managed, managed cash processing through regional centres and put staff into the corporates directly to manage customer money more effectively. The more dramatic programmes included consolidation of legal vehicles, moving all trading to one country, creation of enormous regional centres for all types of cash and currency handling and so on. Most banks completed the bulk of this work in time for the Euro to arrive - a factor which, had they not done it, would have forced such consolidation (after all, 18 countries in Europe trading Euros against each other would not be too smart). Some have continued the work, reducing costs further through consolidating other IT systems or financial products. The stimulus for this back in the early 90s was, clearly, a realisation that revenues weren't going very far but costs were rising. It seems to me that it's possible that there is a similar situation in many governments now. The world economy is precarious at best. Government staff counts are high and probably rising. For somem countries, tax rises may be in the offing unless there is an economic rebound. Back then, the banks had competition to spur them on - if they didn't cut costs and deliver better service, then someone else would and revenues would fall faster. In an increasingly competitive Europe where today the UK outshines all others thanks to some solid fiscal management, other countries are looking a little wobbly - and a progamme such as the banks undertook could bring them back on track, increasing the pressure on their neighbours, including us. The first steps down this path might be a programme to consolidate what a bank would call "back office functions" - payment of cash and receipt of cash as, after all, that's only a matter of reconciling money between accounts; with that underway, an aggressive step would be to consider common internal applications such as finance, payroll, HR, expenses management. A far less aggressive step would be common purchasing of everything from staples and paper to desktop PCs. Some of these steps would create problems at the front end whilst they were underway - customer confusion would be one risk and certainly staff confusion. Much of that could, perhaps, be addressed through using the Internet to mask what was happening underneath. A big programme to not only encourage citizens and businesses to use the 'net but also to get third parties (such as the independent advisors, business councils, accountants and so on) to use the 'net for their customers too would go a long way to buying time for the changes to take place. The 'net can do a lot - and it can certainly give the illusion that all is normal whilst underneath there is a lot going on. Normal may not be enough of course, but it's worth a try if the payoff just a year or two later is much greater effectiveness of services. Such a programme requires significant top down sponsorship. It doesn't happen where fiefdoms are allowed to prosper, where expenses are managed locally, where baseline budgets are hidden from view. With the top down leadership in place, there then need to be positions of power spread across the organisation that don't report into any fiefdom - cross-organisation roles, cross-product roles and cross-process roles. The folks in these roles can look across the whole piece to see the inefficiencies and the opportunities without the history of how the org, product or process has worked in the past. The corporates all had problems like those and had to deconstruct them as they moved ahead. For some it took months, for others it took years. Still others had it done to them as they were, in turn, acquired by more agile competitors. Noone is going to buy-out a government of course, but in an increasingly competitive economic environment, surely the equivalent is relocation of businesses to the strongest economy with the most educated, most available workforce. Weak economies had better watch out, because someone is going to steal your lunch. It's going to be fun to watch as a new set of people grapple with the same old problems.
Posted by Alan at Sunday, April 13, 2003