Tue, Apr 23, 2002 page19

• Mizuho's mess, or how not to operate a bank

  • By Patrick Smith / BLOOMBERG
  • • 
    • 
    • 

  "One bank takes you to a world of new solutions," read advertisements for the April debut of Mizuho Bank, the world's largest in terms of assets and one of the amalgamations that is supposed to represent the future power of post-crisis banking in Japan. A world of near- chaos has been more like it.

  Maybe the mess at Mizuho ought to have been expected of a bank made of three dinosaurs in uncomfortable communion. Dai-Ichi Kangyo, Fuji and the Industrial Bank of Japan surprised everyone in August 1999 when they announced that they would merge into "a national champion," as a colleague who was there put it, "and a champion with a somewhat anti-foreign tilt to its ambitions."

  A long time has passed since then, and the stock has traded well, especially during this spring's (manipulated) run-up in bank shares. But when Mizuho's doors formally opened for business on April 1, you would have thought the merger idea had been imposed upon the three participants more or less by force a few weeks earlier.

  On the face of it, the problem is the information technology system chosen by the new bank. Automated teller machines across the country malfunctioned almost immediately. Some 2.5 million automated transfers from clients to utilities went unpaid -- a big deal in Japan, where "auto-pay" is the prevalent system for settling household and industry accounts.

  Indeed, considering Mizuho's size, to impose a cease-business order on the bank while it cleans up its act would adversely impact something like a third of the Japanese economy, by the estimates I hear. The pols and the authorities have been breathing fire over this mess, but they are essentially stuck. "It's so big they can't do anything about it," a banking friend in Tokyo tells me.

  You would think that the three banks of which Mizuho is comprised would now be fighting over the IT system and how best to fix it. Not so. The internal arguments, I'm told, are over which of the banks should take responsibility, how many managers from each bank should fall on their swords, and how the newly opened positions shall be apportioned.

  Old habits die hard

  There is a lesson in this of fundamental importance. Japanese banks may have a grand strategy for remaking themselves into global competitors able to take on foreign institutions in their domestic market, but they have yet to outgrow habits that extend back to the Meiji era (1868-1912), when Japan began to modernize.

  The grand strategy, to put a complex matter simply, dates back 10 or 15 years, when the Finance Ministry set out to create three or four -- depending upon whom you talk to -- large universal banks based roughly on the German model. German banks may not strike readers as an efficacious model for anything now, but as the architecture of a system that would suit Japan, there is sense in the design.

  Turf battle

  But what problems -- what distance between concept and execution. And the IT fiasco at Mizuho is revealing in this respect. As a friend close to the situation puts it, "We've actually got, by way of an IT accident, a clear light to shine on the details of a classic Japanese turf battle. For a moment, at least, all the sociological behavior is in plain view."

  Turf battles are old and notorious among the Japanese. In corporate Japan today, they resemble nothing so much as the castle-to-castle warfare characteristic of the days of the daimyos and their han, their small fiefdoms.

  Consider briefly the chronology of the Mizuho mess, for it is in the progress of events since that August 1999 press conference that the reality of Japan's bank restructuring is made plain.

  At the time of the merger announcement, all three banks had their own computer systems -- DKB's was from Fujitsu Ltd, Fuji's was from IBM Japan Ltd, and IBJ's from Hitachi Ltd. In one way or another each bank's technology tie-up reflected old keiretsu relationships. A few months later, the initial idea of developing a completely new system was scrapped in favor of the DKB system -- the argument being that Fujitsu was an important DKB client.

  More arguments ensued. The folk at Fuji -- savvier than the others in banking technology -- were infuriated because they felt they were being forced to accept a system inferior to their own.

  Ignoring the problem

  Then what happened was most interesting indeed: Nothing happened for a year. At the end of 2000, a decision was made to run all three systems in parallel for the first year of operation, at which the point the DKB system would be expanded to serve the new bank. In the meantime, more changes in Mizuho's structure.

  While the three banks would merge, the new entity would split into a retail bank and a corporate bank for large clients.

  More nothing took place. Branch numbers changed, sorting codes changed, and a certain outside-inside schizophrenia came into being: Outside the bank, it would appear that clients were dealing with a single entity. Inside the bank they would actually be dealing with three.

  "To senior management, which knows nothing about IT, all this seemed perfectly logical," my banking friend says. "In fact, it was crazy."

  By mid-March of this year, it was clear by way of test runs that there were bugs in the system. By the weekend of March 30-31, pre-processing exercises made it clear that the system would not be able to service the 3 million transactions anticipated for the April 1 launch.

  At that point, the new bank's senior management was taken up entirely with the quintessentially Japanese matter of face. "They couldn't go back, and they wouldn't go forward," one of my sources says. "So open their doors they did."

  Now the plan has changed again, it seems. The DKB system is out -- or will be by 2004-05 -- and a wholly new system is to replace it. The climate inside has changed, too. Reports now are that an exodus of Mizuho staff has begun, prompted by concerns over the bank's stability.

  Anyone with knowledge of Japanese banking history will recognize the story. When Dai-Ichi Bank and Kangyo Bank merged in the early 1970s, there wasn't one personnel structure in the new entity but three: Dai-Ichi's, Kangyo's, and the one created for DKB.

  We can't end this tale by concluding merely that senior management at Japanese banks is (1) in certain respects incompetent and (2) stuck in historical identities and a corporate mentality wholly unsuited to the restructuring exercise Japan has embarked upon. We can't leave out the authorities.

  The Financial Services Agency, under Minister Hakuo Yanagisawa, was aware of serious technology problems in the Mizuho merger as early as March 2001, when the FSA began a three-month review. But once again, nothing was done. And now an agency that is supposed to be policing the financial sector much in the Anglo- Saxon manner is left with its hands tied -- nothing can be done.

  There are many books on my Japan shelf explaining "modern Japan." We will have to wait, it seems, for one to take up "modernized Japan." We're not quite there yet.
• Most Popular

  • 1

    UK navy unveils plans near Taiwan

  • 2

    US involvement in Iran could affect Taiwan: expert

  • 3

    31 charged over KMT-led recall campaign forged petitions

  • 4

    Chikungunya fever cases rise to 13, CDC says

  • 5

    Lai begins 10-stop tour with speech on ‘nation’