A few days before Christmas, IBM brought back from pasture a couple of its former big shots, Paul Rizzo and Kaspar Cassani, to help chairman John Akers. Akers seems to be having a little trouble, although you cannot tell this from his official statements. This unprecedented manoeuvre was disclosed shortly after IBM’s December 15 admission that it would institute further job cuts, plant shutdowns and the like. These actions would cost an estimated $6 billion. When IBM announced these write-offs, Akers sent a note to his company’s shareholders. In the letter, Akers concedes that the owners of IBM stock might be a bit concerned about recent events, such as the halving of IBM’s share price during the past six months. Because IBM has been much in the news in recent days, I expect you have questions about the company’s business performance and direction, Akers wrote. Along with the brief letter, Akers enclosed a thin flyer recapping IBM’s restructuring plans. The leaflet’s theme, stated in its second sentence, is that, The computer industry is in a time of fundamental transition, and the velocity of change is increasing.

Folklore

We can’t recall when the computer industry wasn’t changing, but it is possible that Akers thinks otherwise, which would explain certain matters. Oh, well. Akers was obligated to tell his shareholders something, even if a few of them would have preferred a message that was shorter and more to the point… something like Goodbye. There is a body of folklore that surrounds Paul Rizzo’s career at IBM that may be more relevant now than it has been in more than a decade. According to the central myth, Rizzo might have been made chairman – he became vice-chairman – were it not for the discord surrounding a crucial strategic decision he is believed to have made. The decision on which the history of IBM, the mainframe business and Rizzo’s fate is believed to have turned surround the pricing of IBM’s 4300 series of mid-range systems, which were announced in January 1979. The machines were priced very aggressively, too cheap in the opinion of critics, producing some unexpected and mixed consequences. The most important impact of the 4300’s low cost was a dramatic resurgence of IBM in the small mainframe and mid-range arenas. Not only did the 4300 machines attract orders from existing IBM customers, they also brought Big Blue into many new accounts. And the impact on IBM’s competitors in the mid-range market was nothing short of devastating. Minicomputer vendors, principally Digital Equipment, had been nibbling at the low end of IBM’s 370 customer base; the 4300 stopped them cold. Makers of small IBM-compatible mainframes suffered a big setback, too. Leasing companies that had overextended themselves in the 370 business found themselves unable to realise the residual values on which they and their backers had formerly counted; Itel, the largest lessor of the era, had to file for bankruptcy. Not all the effects of the 4300 announcement were good for IBM. During the middle and late 1970s, IBM had amassed buckets of cash. For reasons that remain unclear (but Morgan Stanley, IBM’s outside investment bankers at the time, may have played a role), IBM used its cash to buy large blocks of its own shares in the open market. Even after its stock repurchase, IBM was left with what looked like plenty of cash… but, it turned out, this was incorrect.

By Hesh Wiener

As soon as IBM announced the 4300s, customers with large mainframes – 303X machines – figured that a new line of large mainframes was imminent. This product line, the 308X family, was two years in coming. But at the time nobody outside IBM (and possibly nobody inside IBM, either) realised how long it would take to get the 308X out the door. Expecting new and improved systems, most customers who needed mainframes decided to rent rather than buy 303X machines – the only ones in the IBM line at the time. They acquired the computers under a long-since-abandoned scheme that made rental very attractive. To fill its orders, IBM had to build hundreds of 303X c

omputers. This was so enormously expensive that IBM ran low on cash and was forced to borrow hundreds of millions of dollars by issuing bonds. Also, when IBM completed its bookkeeping for 1979, corporate profit had dipped. (In those days, even a small downturn in IBM’s profit was considered shocking.) This tale of the 4300’s inception, which we are sure is not entirely correct but seems to be widely believed, ends on a sad note. Morgan Stanley’s role as IBM’s investment banker was diminished, and other Wall Street firms picked up the business that Morgan lost. Paul Rizzo was diverted from a path that might have led to IBM’s top job. In London, Lloyd’s, which had provided a peculiar type of insurance coverage for computer lessors, took a huge bath: numerous clients in the leasing business filed claim after claim in the wake of losses on their 370s. And IBM never again officially priced a commercial computer at the lowest levels permitted by technological progress… although it didn’t necessarily charge whatever the market would bear. (The first 3081s were considered so valuable that a number of customers paid premiums to lessors and traders for options to acquire them early in the cycle. In 1981 a Technology News publication, Computer & Communications Buyer, reported a payoff to one leasing company of $1m by a bank eager to get an early start in the 308X generation.) So IBM is bringing back Paul Rizzo. And if folklore is any guide, he will persuade IBM to slash mainframe list prices… with numerous consequences. Imagine if the list prices of IBM mainframes were a third of what they are today, if IBM made these prices consistent the world over, published these prices and ceased offering discounts. (The 4300s were, on a price-performance basis, about a third the price of the 370s they directly replaced.) IBM would almost certainly find itself deluged in orders. Despite the company’s statements that it had excessive production capacity, particularly in the mainframe division, we believe IBM would find itself unable to meet demand without reversing some of its planned plant closings and layoffs. The cheer in New York’s Hudson River Valley, where IBM makes large systems for the US market, would be heard all the way to Wall Street. If IBM’s decision initially forced it temporarily to build engines that were less powerful than those in the 9021 (in order to reduce processors to one or a few chips), all the better.

Only a myth

Coupled with a new low-cost systems software pricing scheme that favoured multiple mainframes rather than single large ones, IBM would no longer have to worry whether Amdahl or Hitachi built a faster computer. After all, in mainframes IBM still sets the rules… whenever it has the presence of mind to create them. Sufficiently cheap mainframe MIPS also would allow Big Blue to economically provide new software to support attractive and demanding applications, including those involving graphical user interfaces and very high transaction rates. Customers’ investments in programming and, more importantly, know-how, would be not merely preserved but increased in value. This reward would more than offset the inevitable losses of residual value that would beset owners of current mainframes. (Although they have been known to grumble about unexpected change, IBM customers have never objected to paying a fair price for progress. Why should they?) Lastly, IBM might want to offer its machines on a monthly rental basis once again. This would affirm the company’s confidence in the lasting value of its products just as surely as its change to a policy of outright sale or long-term leasing told the market the opposite – that IBM was afraid to invest in its flagship machines and that the company was shifting the burden and financial risk to customers. It’s rather unfortunate that the tale of Paul Rizzo and the 4300 is only a myth. Or is it? Copyright (C) Technology News Ltd.From the January 1993 edition of Infoperspectives International.