In five years, the company's revenues soared over 8,000%. Yet it has never sold equity and rarely borrows money.
California's Silicon Valley is a featureless sward of high-priced turf spreading outward from the steps of Stanford University in Palo Alto much as Route 128 does from the Massachusetts Institute of Technology in Boston. It is to West Coast technology as Beverly Hills is to movies: the home of stars. Hewlett-Packard, Apple Computer, Intel, National Semiconductor, and scores of other high-tech luminaries are housed there.
Most of the neighborhood consists of public companies with large capitalizations, impressive amounts of long-term debt on their balance sheets, and infusions of venture capital at various stages of their rise to preeminence. One cheeky newcomer, however, shares none of these taken-for-granted marks of making it.
Among the country's largest privately owned manufacturers of microcomputers (and, it claims, the world's biggest manufacturer of S-100-based microcomputers: Invented in 1975, the S-100 bus is a sophisticated component that shunts data from one part of a computer to another), Cromemco and its 450 or so employees moved into a brand-new 200,000-square-foot plant in Mountain View in 1980. (What the mountains are is academic now: The view is steadily vanishing into megalopolitan smog.) The company rents these quarters not because it can't afford to own them, but because buying real estate and paying off long-term notes runs counter to the financial strategy that had brought it this far, according to Cromemco's owners, "without even being close to getting into trouble." Renting also gives the company flexibility; it has moved five times in five years. Barely six years old, Cromemco is expanding at a compound rate of 216% per year -- a giddy pace that demands that cash be plowed into growth, not tied up in grounds. Liquidity -- the ability to move in and out of cash quickly -- has been Cromemco's financial cornerstone since its founding as a partnership in 1975.
In and of themselves, Cromemco's revenue figures are nothing short of astonishing: $50,000 in 1975, $600,000 in 1976, $4 million in 1977, $11 million in 1978, $20 million in 1979, $30 million in 1980, and an expected $50 million in 1981. But they are all the more impressive in light of the fact that the gains have been accomplished without a nickel of outside investment. Cromemco remains owned wholly by its two co-founders, who have engineered its spectacular ascent to the top of the microcomputer market solely through internal profits. Aside from what shares of stock Cromemco contributes to a profit-sharing plan, ownership is undiluted.And though it has some short-term bank financing of its receivables, the company, incorporated in 1976, has no long-term debt.
At the outset the owners, both steeped in the traditions of closely held family businesses, were determined to bring Cromemco along conservatively. Their heroes were William Hewlett and David Packard, who remained active in their business through their retirement; their fear was to end up like Gene Amdahl, whom they consider a "macroexample" of show ownership can be wrested away through financial exigencies. What could have been accomplished -- or mishandled -- had they gone for broke in the just-dawning microcomputer industry of the mid-'70s is anyone's guess. And guess some of it must remain, since one of the privileges of private ownership is that you don't have to wear a P&L heart on your corporate sleeve. Cromemco's financial strategists are willing to talk -- but only to the point where someone else's financial strategists might be listening.
One of the keys to Cromemco's remarkable ability to disdain venture capital infusion and to remain free of long-term loan obligation can be found in when and how the company began. Cromemco started almost as humbly as a mom-and-pop corner store, and was swept along in a rising technology tide. It was a self-launch that seems to have happened, and to have succeeded, by preordination.
The corporation's two, and only, shareholders, president Harry T. Garland and vice-president Roger D. Melen, both 34, received their doctorates from Stanford in 1972. Garland's degree was in biophysics with an electrical engineering minor; Melen's was in electrical engineering. In those years the pair wrote several articles for Popular Electronics, offering in the text to supply the components of the described kits (one a capacitance meter, another a TV camera). Almost grudgingly, they were in business.
At the same time, both were invited to join the Stanford faculty. With this steady income, plus the free magazine "advertising," they were able to put together enough capital -- the grand sum of $5,000 -- to enable them to buy the parts to manufacture a small computer. Adapting the S-100 bus to modular "add-on" circuit cards of their own devising, they kept on selling computers. The units were not what could be called aggressively marketed: Recalls one observer at the time, "It was almost word-of-mouth." Their research lab was a room in Melen's apartment; the production facility, Garland's garage. "I like the way we did it," Melen reflects. "No rent, no salaries. We packed the stuff ourselves, we soldered it ourselves. We didn't sap the operation when it was too small to support us." On December 31, 1976, Cromemco was incorporated.The name was an old-grad gesture in honor of their Stanford dorm, Crother Memorial Hall.
It turned out to be a classic case of being in the right place at the right time with the right product. From the start, Cromemco went for the high end of the microcomputer field, emphasizing quality and reliability over price. (A full top-of-the-line base unit sells for about $8,000, and a complete six-user system for over $18,000.) Only once has the company had to lower a price to meet competition. But if the capital aspect of the business was played close to the vest, it was innovative in other areas. The company holds a number of patents (it plows back some 8% of revenues into research and development), and pioneered marketing arrangements with independently owned retail computer dealers whom Cromemco trained at its factory -- an arrangement to which it still holds today.
Cash flow in the start-up days was considerably enhanced by the fact that neither owner drew much of a salary. Indeed, there were 15 employees on the payroll before Garland and Melen joined as No. 16 and No. 17. But as exhilarating and easy to come by as its early growth spurts must have felt, the conservative instincts of the owners led them to seek outside business guidance. They went of Keith Argabright, a C.P.A. Melen and Garland's "only" directive was that their progress be totally self-financed. As inhibiting as a constraint against venture capital may have been to an ordinary business, it didn't seem to hamper Cromemco: Its sales leapt nearly 600% from 1976 to 1977.
Argabright, now regional managing director of Main Hurdman, a large international accounting firm in San Francisco, has been closely involved with Cromemco's game plans ever since. In 1977, the company went whole hog into manufacturing computers. By the end of 1981, it had carved out a 40% share of the S-100 high-performance micro market. Not surprisingly, Argabright is a strong advocate of CPA-firm involvement with any aspiring business.
Given the owners' aversion to outside investment, Argabright's immediate concern was to maximize bank credit. Cromemco, he insisted, had to have an adequate credit line. As a matter of business principle, growth could not be left entirely in the hands of internally generated cash; there were bound to be some soft periods. The company's first credit line, from Bank of America, was for $250,000. As it turned out, Cromemco has never had to tap its full line. But to make sure it can stay out of trouble if slow times hit the market, Argabright still insists on keeping open maximum lines (which today exceed $5 million), even though they may never be deeded.
To maximize the generation of internal cash that was to be the fuel for growth, the company turned to the Internal Revenue Code. Three strategies were devised that would take whatever advantages the Internal Revenue Service allowed. The first was to elect accelerated depreciation schedules for plant and equipment. A second involved a profit-sharing plan for employees that provided a tax write-off while costing the company nothing in the short term. A third tactic pushed Garland's and Melen's salary and bonus arrangements against IRS ceilings for closely held companies.
Sooner or later, of course, accelerated-depreciation deductions run their course. But -- so far, at least -- Cromemco's growth rate is such that each year's cessation of depreciation is insignificant in terms of new capital investment. In year seven, for example, Cromemco expects to purchase more equipment than in the previous six years combined. "A good strategy for growth," Garland believes, "is always to buy more equipment than is becoming fully depreciated. When growth slows -- to 25% or 35%, say -- then we'll reexamine our position. At that stage, we may not need the tax advantage." Right now, as a private company, the formula is simple: Intentionally show low profits through maximum depreciation and the resultant minimum tax payments.
Explains Garland, "We take fast depreciation when allowable. That saves tax dollars today that essentially are reinvested in the company and will earn us more in the future. Some companies may elect not to, and are willing to pay more in taxes. In our opinion, that wastes the opportunity to use it."
Cromemco can pursue its diminished paper-profit maneuvers because, unlike public companies, it does not have to showcase real profits. Freed from quarterly reports and earnings-per-share concerns, Cromemco's growth avidly consumes profits by financing additional inventory and equipment. "We always have a use for profits," says Garland, "since a private company can look to the long term. You have an option to choose fast depreciation schedules, for example. But in a public company, the tail often wags the dog. A public company will be willing to trade more taxes for higher visible earnings. That's short-term window dressing vs. long-term health."
One of the conservative tacks that dictates Cromemco's pace is the constraints imposed by the banking industry, itself fiscally conservative. Even at that, the company uses only about half of Bank of America's commitment. "We want to have a cushion within the very cautious nature of a bank loan," says Roger Melen, who now teaches a course in entrepreneurship at Stanford. "Cash flow can kill faster than profits can cure. Running out of cash is the biggest threat to a small company." Even their own banker, Rory MacLean at Bank of America's San Jose office, deems their approach reserved. "Cromemco's leverage is conservative, especially in view of their industry and marketplace, and the type of credit arrangements we've established," observes MacLean, who, like Argabright, works very closely with the company. Nonetheless, MacLean describes Garland and Melen as "real whiz kids with the numbers," and counts Cromemco as one of the bank's preferred customers.
Two areas where cash flow troubles can kill most quickly, Melen can Garland agree, are in inventory control and receivables -- both monitored carefully at Cromemco by, of course, Cromemco computers. Acting on Argabright's urging that "anticipating needs was of utmost importance," they established a materials planning system at an early stage to conform to projected growth and production rates. Argabright insisted on keeping to a growth rate that could be controlled, arguing that most companies get into difficulty by letting the growth rate itself make the rules. If you leap ahead of your projected growth, he coached, many sectors of the manufacturing process go haywire; accounting systems break down, overtime pay soars, parts shortages occur, and so on. An inviolable cash flow dictum was laid down: Cromemco's projected growth rate could be no greater than its return on equity. Following this stricture unswervingly, Cromemco's net after-tax profit margin has ranged from a high of 11% to a low of 4% -- resulting in an average that is standard in the industry.
Despite the company's widespread use of its own computers in such areas as product testing, dealer/OEM files, production reports, inventory tracking, accounting, and sales, an unexpected problem arose: Its people, not its computer models, were overambitious. At an early growth phase the purchasing department became too optimistic and assumed the company was growing faster than the computers were predicting. So they ordered extra parts. It's the nature of purchasers, Garland and Melen discovered, to fear shortages rather than overages. "That's something we had to learn," Garland admits. "If you're not careful, the tendency is to order too much. Not that there's something terribly wrong with it, except that it ties up cash. If you have to go to the bank with inventory, money is apt to have become more expensive."
Argabright, too, was concerned with inventory control. The trick, he advised, was to cut back on inventories as soon as possible when orders dropped off. "I continually preached that we ought to keep inventories down. High-tech companies get into trouble by building inventories too fast."
To help those involved keep inventory in line with growth projections, the company now shares its financial planning with the purchasing department, feeding it as much data as possible. "Information can get buried in a system," Garland explains. "It's not that it's being hidden on purpose; sometimes it's simply not provided." With Cromemco's fact-supplying guidance, purchasing people now can measure their own performance against management goals. The approach is working: Cromemco's 1981 inventory turn was a healthy 3.3 times.
Cromemco's approach to receivables, however, contains only one bare fact: Pay up within 30 days of shipment. The age of the company's receivables, financed through its revolving bank line, has rarely exceeded this interval, since its products are highly sought, backed now by national advertising. In turn, Garland and Melen are happy with their network of independent dealers and original equipment manufacturers (some 10% of Cromemco products go into OEM machines). Predictably, they enjoy the idea of working with entrepreneurs like themselves.
Though all seems tranquil in Cromemco's corner if Silicon Valley, like the mountains in the distance the company has had some ups and downs. Curiously, as a breather, management would just as soon have it the latter on occasion. "After all," explains Garland, "if the market softens, that just helps. To keep growing requires funding." On the other hand, when the growth rate exceeds return on equity (by Cromemco's measure, equity is simply the amount of money in the company), management cuts back on expenses, rather than seek new financing or extend borrowing. "Since it's our money," the owners say, "we tend to be more conservative. It's no one else's money that we're going to lose."
Cromemco's unusual, ESOP-like profit-sharing plan has also been a factor in cash-flow control, inasmuch as it deals in corporate shares rather than direct cash investiture. The company "buys" stock from its own treasury, then "sells" it to its profit-sharing trust, which in turn holds it for employees. Though for tax purposes the company has made a deductible contribution, in effect, no money has actually been used.
Another advantage of using stock instead of cash for vestment is that the employees like it better. An amount based on 15% of their salaries is used to acquire stock, whose value is independently appraised; after 4 years of employment they are 40% vested, and after 11 they are fully vested. Thus employees not only share profits, but as a result are able to participate directlt in the fortunes of a dynamically growing company. An employee who cashes in after 11 years is apt to come out very well indeed, and still be young enough to start a new career -- or company. Not surprisingly, employee turnover at Cromemco has been low.
And Garland and Melen remain faithful, too. They have been wooed by venture capital firms and underwriters from coast to coast, but have spurned all offers. Yet Cromemco is at a crossroads. If it continues to reject outside capital, its growth will slow; if its growth slows, Cromemco may lose its market share within a still-exploding industry. A decision has to be made: more capital with dilution of ownership, or just another closely held business whose owners are doing well enough, thank you.
Melen and Garland would like to stay private. It's all under their control, and that's the way they like it. "We see it as a lifelong career," they say. Going public, Garland argues, would only cause a distraction. "We just want to deliver a product and keep our employees happy," he insists. "From the beginning we didn't want other shareholders." Cromemco likely can maintain a reasonably strong pace -- 30% to 50% annually -- for the near future, somply by operating under the philosophy that a small company can be run more efficiently than a large company. To this end, Cromemco recently has opened markets in Europe (where, to avoid currency disparities, payments are made in U.S. dollars) and is among the first computer manufacturers to move aggressively into China.
Argabright, however, would rather see much more rapid expansion. "I think they're growing too slowly," he says, "and I've told them so. They're on the threshold of dramatic growth within an industrial revolution. To grow rapidly, you need outside capital." He recommends that they at least consider a private placement of equity "with somebody who would complement their product line, not a pure venture capitalist." He feels Cromemco's weakness now is in marketing, and is urging them to bolster this area. The private placement, he points out, might well be with a firm that has strong international marketing outlets.
But Argabright is still apprehensive. "To grow rapidly," he insists, "you need outside capital for dynamic growth. You've got to go public. I'm sure they will." If they do, he foresees, Cromemco could become another Hewlett-Packard; if not, he rues, "they'll just have to find a niche in the market and stay there."
Whoever wins out in the developing battle between conservative management-owners who want to grow within the parameters of available cash and an activist adviser counseling additional financing, both agree on one thing: The Cromemco story could not happen again today. "The industry is too mature now," says Garland. "Cromemco already has millions of dollars invested. No one could self-finance to this stage again." To be sure, Garland and Melen happened to be there with the right product -- an expandable microcomputer that allowed customers to buy only what they could afford and add on to it later -- at the right time. "But in any venture," Argabright reflects, "luck is a factor. Nonetheless, you have to take advantage of that luck."