One of the more touchingly pathetic images of the current political scene in Europe is the misplaced faith that finance ministers place in the possibility that Germany will one day soon reduce its interest rates so that everyone else can relax fiscal policy a little. All the very substantial evidence is that the Bundesbank will see the entire German economy laid waste before it with ravening hordes at its door before it will change its policy of doing everything in its power to get inflation back under control by one iota. Germans are like that: they have a certain tramline mentality, and once they are fixed on a course, they pursue it come hell or high cliche. There was once a bureau joint venture in which Siemens AG and Nixdorf AG were major investors. It lost money for 17 years before the partners finally decided it might be time to call it a day.

Accumulated losses

AEG AG accumulated losses every year from 1973 to 1982 before anyone decided that something really had to be done about it and lost even more money over the next couple of years straightening things out. AEG’s own Olympia turned in a similarly untroubled chronic loss-making performance under its laissez-faire parent. So what’s up at Siemens? The company showed surprising flexibility in saying that it would not build its own plant to make the 64M-bit memory chips it was developing with IBM Corp. It therefore came as a bit of a surprise that Siemens was committing to what is presumably a one third share of $1,000m to develop 256M-bit memory chips with IBM and Toshiba Corp – and perhaps to its share of another cool billion to equip a plant to make the things when the time comes. Now in general, it has to be said to be a Good Thing for a big chip user like Siemens to make its own parts rather than be dependent on the ups and downs of the market – so long as it can make money out of them, at least in the good years. What is Siemens’ performance? It is neatly encapsulated in the opening of a big piece in the Wall Street Journal. The year is 1984, writes Richard Hudson. With much hoopla, German electronics giant decides to plough $1,670m into a partnership with another semiconductor maker (remember the Mega Project with Philips Electronics NV?) Two years pass. It’s 1986. Siemens, its memory chip business losing money, announces a second partnership to catch up. Still losing money in 1990, Siemens unveils a third chip partnership. In 1991, yet another semiconductor partnership.

Losing money

Still losing money. Now it’s 1992, so you can guess what the hottest news in the European electronics industry is. The paper quotes an analyst that supports the strategy saying that one might say it looks like throwing good money after bad. Other analysts, less sanguine, have been writing increasingly critical notes on the company’s chip strategy, pointing out that Alcatel NV and L M Ericsson Telefon AB seem to do all right without making their own chips. Lehman Brothers estimates that since 1986, Siemens has invested $2,300m in semiconductors, and even in devalued dollars, that’s a lot of lettuce. The return so far? A loss last year estimated at $335m on sales flat at $1,335m. And that loss compares with a profit for the whole $48,750m Siemens group of $1,130m. Shareholders must be beginning to wish the company had simply paid the money out in dividends instead of investing in chips – they probably reckon they could have put it to use more profitably. It is difficult to spot just why Siemens can’t make money out of chips even in the good years most everybody has a hard time when the chip cycle is in one of its regular downturns but for most manufacturers, the good years more than compensate – but the reason is almost certainly the well-nigh unsustainable burden placed by the German state on employers to keep the country’s lavish and wasteful welfare state in the manner to which it has become accustomed, plus absurd restrictions on working hours, which make processes like chip manufacture painfully uneconomic – shut the plant down at lunchtime on Saturday and start it

up again after the weekend and you spend all Monday morning making dud chips while the line gets itself back into shape again. The answer is almost certainly to move all its chip plants into the new Lnder of the east, if not over the border in Poland or Czechoslovakia or way down in Hungary, where they would be welcomed with open arms. So Siemens’ semiconductor operations are in a bad shape: is the Siemens Nixdorf Informationssysteme AG computer subsidiary in any better shape? It should be remembered that Siemens Data Systems lost money steadily for about 20 years before Siemens finally got it into profits for a few years in the 1980s, but the rescue of Nixdorf Computer AG put an end to what was effectively an unnatural state of affairs for Siemens’ computer business. The business was to have returned to profit last year, but didn’t. It was to have broken even this year, but won’t. Now the company isn’t even making a forecast profits for next year. The Sword of Damocles hanging over the company is its dependence on the 7.000 family of BS2000 mainframes, which are way behind ICL Plc’s Series 39 VME machines in incorporation of open systems standards – the Siemens approach has always been to follow closely all that IBM does in mainframes and do it a little better and a little more cheaply with machines that are architecturally similar but not quite compatible.

Still making losses

Siemens has been shielded from the rush to open systems by the fact that the majority of its BS2000 customers are in the German public sector, coupled with the fact that Germans are temperamentally deeply conservative and averse to change. But in the present climate where the public sector is heavily constrained to cut costs wherever it can, the start of the migration from mainframes to open systems cannot be postponed for much longer, and all the signs are that Siemens will just have got Siemens Nixdorf onto an even keel when the body blow that hit Nixdorf with its proprietary 8870 series of System/36-class business computers will hit the Siemens mainframes in spades. Has the company even begun to tread the hard path that ICL trod in the early 1980s, and Groupe Bull SA and IBM Corp are mired in now? Never mind, one thing of which we can be fairly certain is that Siemens will still be in the computer business, still making losses on it, 20 years from now.