Mad Dog 21/21: Leverage
September 4, 2007 Hesh Wiener
Archimedes was talking about leverage, one of his favorite topics. He said, “Give me a place to stand on, and I will move the Earth.” The Sicilian from Syracuse probably knew more about mechanical leverage than anyone of his era, which was the third century BC, but he didn’t know about financial leverage. Some people are still finding out about financial leverage, particularly on Wall Street, and doing it the hard way. What they cannot learn from Archimedes they might try to garner from IBM, which looks like it knows a little more about the topic than most companies.
IBM uses financial leverage to strike a suitable balance between debt and equity in its finances, and knows that a proper ratio for its leasing business is different from the one for its other operations. Evidence of this can be found in Big Blue’s financial reports, and in the way its share value has held up lately compared to that of so many other enterprises. Still, the best anyone can say about IBM in these uncertain times is that if IBM turns out to be wrong about leverage, nobody has yet found any evidence of impending difficulty. Big Blue hasn’t bet its boodle on risky mortgages, and it derives a considerable portion of its revenue and profit from persistent deals in software and services. It has gotten out of some business segments that inherently have a lot of volatility, like the personal computer trade. While the company and its investors might wonder how its reseller financing will hold up during the coming months, IBM has done a great job avoiding that kind of risk in the past and then managing it carefully when it couldn’t be avoided.
Nevertheless, IBM has a lot of big customers in financial services, such as banks, insurers, mortgage companies, and brokerage houses. If they have a bad year, they might cut back their computing budgets. IBM is not completely immune to external conditions, and it never was. Twenty years ago, in the aftermath of the stock market meltdown of 1987, IBM entered a challenging period. The upshot was a big change in management that culminated six years later when Lou Gerstner became the company’s first boss drawn from outside the IBM empire.
Archimedes suffered from circumstances around him, too. A Roman soldier killed him during the Punic Wars. This was probably a mistake. Historians believe the forces laying siege to Syracuse were ordered to spare him. Sometimes empires get it right, sometimes not.
In the years before he met his tragic end, Archimedes completed exhaustive studies of levers, pulleys, screws, and other mechanical devices, discovered a number of principles in physics and advanced the state of mathematics. Among the many things for which he is famous, Archimedes proved that the surface area and volume of a sphere is 2/3 that of the smallest cylinder that can surround it.
IBM’s business used to be more like a sphere in a relatively large cylinder. IBM got revenue from the one-third of user budgets that went for hardware and from a bit of another third that was spent on software. The last piece of users’ budgets went for personnel. During the 1990s, IBM went after that final third, offering services to replace in-house personnel. The services business took off quickly and now accounts for more than half IBM’s revenue. IBM is also learning how to pick up more of the middle third, the software budget, mainly by acquiring companies that have staked claims on software sales in the IBM customer base and in areas where IBM does not have a lot of influence, but wants to. IBM has learned how to use its established position to gain various kinds of leverage and by other means as well.
Archimedes extended the world’s understanding of mechanics by expanding his concept of leverage and by discovering other principles through keen observation and insight. One principle whose discovery has been credited to Archimedes is that involving the concept of buoyancy. Archimedes expressed with perfect clarity the idea that an object would displace a mass of water equal to its own mass, and if it sank, because it was heavier than water, displace a volume equal to its volume. In a famous (or perhaps apocryphal) tale, Archimedes applied this principle to the investigation of a crown given to a King Heiron II of Syracuse. Concerned that the crown was not really gold but instead made from an alloy of lesser worth, Heiron asked Archimedes to discover the facts without melting and thus destroying the crown. Archimedes pondered the problem and then arrived at, so to speak, a solution: He dropped the crown in a container full to the brim with water and then measured the volume of water that overflowed. Armed with the volume of the crown and its weight, he was able to calculate its density and certify that it was indeed gold. It is said that when he first came up with the idea he was so excited that he ran naked into the streets of Syracuse shouting “eureka,” which is Greek for “I have found it.”
There is no comparable story about Lou Gerstner and his discovery that the services business could bring growth to a faltering IBM, nor of any IBM CEO contemplating the history of the PC running into a street, such as Wall Street, naked except for a suit costing a couple grand, and shouting, “I lost it.”
In all its endeavors, IBM depends very heavily on the mainframe base to gain leverage with customers. IBM’s big iron users have vast investments in mainframe software and know-how; the much larger i5/OS customer base (in terms of customers, but not money, is also a key lever). Their data is stored in mainframe and System i formats, in EBCDIC code, and organized about mainframe file systems. Even small shops that want to migrate to an alternative platform have to do quite a bit of work to make the transition, and even then the outcome in terms of total cost of computing is hard to predict even if the new platform uses hardware and software with smaller sticker prices.
Nevertheless, mainframe and System i servers are not interactive or more accurately event-driven the way Unix, Linux, and Windows systems are. And they are not quite as adept at many tasks employees and customers want to perform over the Internet. The result is that IBM mainframe users have used alternative architectures (or Linux running on mainframe or System i hardware) to manage email, support Web sites, and run other applications that have grown up in other environments. Whatever their views of these systems, they generally gain one thing that the mainframe no longer offers: Access to the largest pools of technical talent.
IBM’s answer is its services business. IBM has the talent it takes to run mainframe shops and it keeps enlarging its roster of trained professionals, mainly by staffing up in India. This has opened vast opportunities to Big Blue, but it has at the same time exacerbated the shortage of mainframe talent available to users in advanced countries. The effect has taken a while to appear and its full bloom will take longer still, because IBM’s shift to India has put a lot of mainframe professionals on the street, where they become available to end users (and rival services companies). But the trend has also reduced the flow of young computing professions into the mainframe world. There are fewer opportunities there than in Unix, Linux, or Windows, and the opportunities are concentrated in and around the locales where mainframes persist.
IBM has built a talent pump, and while at the moment it appears to be creating new reservoirs of software and operations specialists available to its customers, it is also draining the large pool of mainframe professionals that formerly made it relatively easy for mainframe users to enlarge their in-house teams. Archimedes is famous for a pump, too, one that used a screw inside a cylinder to raise water.
It took Western science the better part of 2000 years to get from Archimedes to Newton and to appreciate Newton’s Third Law, which says that for every action (in mechanical systems) there is an equal and opposite reaction, a concept that also forms the basis of many useful social metaphors. In social systems and economic systems, the reactions don’t occur immediately, as they do in physics labs, nor are they always precisely equal and opposite.
So we might have to wait a long time to see how IBM’s shift to services, now in its second decade, ultimately plays out. Similarly, we might not discover how financial leverage inside and outside IBM affects the company for a while, or its big competitors in the IT arena. Even as the worlds’ central banks try to pour oil on troubled financial waters, it is clear that some fundamental aspects of the world economy are changing. Mathematical models that for years have helped investment institutions manage risk seem to be failing. And established business models that helped executives–including those at IT vendors–best use leverage could fail during inflections and discontinuities, too.