There is no question that IBM’s 1988 figures marked a substantial recovery in the company, which has been struggling since 1985 to get its business back onto the high growth tack that it enjoyed with the odd one- or two-year blip between 1964 and 1984. And the company had plenty of good news wrapped up in the figures even the downside was up so to speak, because although one of the charges it took against profits was the result of more people than it had budgeted for taking early retirement incentives, that will reduce the company’s cost base for the coming year and enable it to achieve a better return on equity even if turnover stagnates. On the downside, IBM has been advising analysts that its tax rate this year is likely to be two percentage points higher than in 1988, which caused a rush of pencil sharpening as those IBM watchers who had been the most bullish on the outlook for the company rushed to lower their per share earnings forecasts a few cents.

Plus-plus-plus Nevertheless, the consensus of the US analysis fraternity is that IBM is set fair for at least one and very likely two good to very good years as momentum resumes in large systems sales particularly fuelled by all the new disk drives expected shortly from the company as well as from a migration over to Enterprise Systems Architecture. Another plus-plus-plus point seen for the company is the superb early successes of the AS/400. This is most exceeding expectations in areas where it was thought likely to do least well – new users; it is doing relatively least well in one of its two key roles, that of preserving and deriving a lot more money from the System 36 base. Among the most bullish of all on the progress of the AS/400 is Steve Cohen of SoundView Financial Group Inc – Gartner Group that used to be – who says that before the AS/400 came out last summer, he had pencilled in shipments of 25,000 for 1989, but went to 46,000 a couple of months later, is now at 65,000 and may raise his estimate to 80,000, tellingt the Wall Street Journal that I can’t raise my numbers fast enough. The PS/2 is still a worry for IBM, not doing paticularly well in the US, doing rather better in Europe, fulfilling its role as the workstation of choice for large users who vote the IBM ticket, only an also-ran at the retail end of the market. But in all the exaggerated euphoria, there are one or two aspects of IBM’s overall performance that seem to be being disconcertingly overlooked. The figures out yesterday from IBM Credit Corp – see page five – come as a reminder of some remarkably generous deals the finance arm has done over the past four or five years, which in the current state of the leasing market could well start coming home to roost within the next year or two. And none of the US analysis we have seen so far has highlighted the fact that in 1988, IBM did a whopping 57.6% of its business outside the US market.

Good news That’s good news to the extent that IBM is able to repatriate yen and marks and pounds to the US, but it does little for the dismal US balance of payments, because in general, what IBM sells abroad it makes abroad. The corollary of the healthy foreign business is that IBM, almost unnoticed, did conspicuously badly in 1988 in its largest single market, the US itself, where, when one picks one’s way through the minefield of charges and extraordinary credits, the company appears to have seen net profits down 10% on turnover that fell about 1%. That rather spoils the AS/400 story, because as you go down IBM’s product line from the very top, you find that while the company sells about as many large mainframes abroad as it does in the US, the US starts to take a larger proportion of the total System 38s and a much larger proportion of the System 36s than the rest of the world does. The figures for personal computers slew back into IBM’s favour to some extent, but the broad trend is very, very clear. And the reason is very simple: at the mainframe end of the market, IBM has no real competition except in Japan, whereas at the System 38 and 36 level, as well

as the Wang Laboratories and the DECs and the Hewlett-Packards that it meets in the US, there is a local champion such as Nixdorf Computer AG, Olivetti & Co, Groupe Bull SA, ICL – and in Japan there are about 10 local champions in the so-called mid-range, with the result that IBM Japan’s share at that end of the market is dismal. And if the US market is really in such a miserable state as IBM’s figures seem to suggest, the sustained demand for the AS/400 now that many of the easy sales – all those System 38 users panting for more power – have been made, may well start to wane sooner than anyone is expecting. Moreover IBM is being congratulated on the success of a machine that almost by definition would be successful: it had a ready-made market within IBM, and the early acceptance there made it look attractive to the uncommitted. But while it may be hurting Wang’s VS, there is very little evidence that it is doing any harm at all to the company IBM really needs to wound – DEC. Not least because that is the role of the 9370, and the 9370 just is not man enough for the job. And that is another deep worry for IBM: two of its five key computer lines are well-nigh universally regarded as failures – the second is the RT – and the jury is still out on the third, the Micro Channel PS/2s. And that gives cause for concern because in its recent history, IBM has never succeeded in rescuing a major computer line once it has been branded a failure. The Series 1 minicomputer never recovered from its early lack of software, the 8100 stumbled along for the whole of its benighted life, never generating a spark of real excitement once the market got to grips with it and found it wanting. The System 23, intended to lead users gently to System 34 or 36, never had a chance once it got mercilessly clobbered by IBM’s own Personal Computer and other people’s software. IBM can certainly rescue the 9370 and the RT to make them as successful as the Series 1 and the 8100 ultimately became – but something rather more than that is needed to stop DEC’s charge in its tracks. But IBM’s growing dependence on the foreign market for its success is bad news for the company in its heartland, the top-end mainframes, as well.

Remarkable Those remarkable figures on moves in the 370 base in France (CI No 1,097), suggesting that in 1988, the installed base of 3090s there rose by just three while the 308X base rose by 113, point up the fact that in many of IBM’s major European markets, users are much less ready to move to the latest releases of IBM’s mainframe hardware and software at the time IBM tells them they should than are its US users. Coupled with rising nationalism in both Europe and the Far East, the increasing proportion of its business that IBM does outside the US is likely to be a growing cause for concern at Armonk, needing an ever greater application of diplomacy and finesse to counter a rising tide of protectionism.