An $11.98 plastic contraption, the Super Soaper Pet Washer, and a $25,000 De Lorean car sheathed with stainless steel may not have advanced the frontiers of technology very far. But the two disparate products -- along with any number of high-technology devices that have advanced them -- grew in part from the same source: capital provided by a limited partnership formed specifically to finance research and development projects of American companies.

R&D limited partnerships aren't exactly brand new. The concept first emerged in Silicon Valley's fertile environs about eight years ago. But recent events, including John De Lorean's well-publicized financing efforts and Gene Amdahl's successful public offering of $55 million through Merrill, Lynch, have raised interest in the technique to dramatic levels.

As the label implies, R&D limited partnerships allow investors -- usually individuals in the highest tax brackets -- to invest money directly in the work of inventors and inventive companies.Since a favorable Supreme Court ruling in 1974 (see "The Drift of Snow," page 73), the twists and wrinkles of such arrangements have proliferated. But all spring from a basic concept.

Simply speaking, investors put taxsheltered money into a project. If the project is marketed profitably, the partners are bought out by the inventor; if not, they lose their investment. For a rudimentary illustration, let's say a modernday Thomas Edison wants to improve the light bulb, but his company doesn't have enough capital to pay for the necessary R&D. So Edison turns to a limited partnership. He enlists 10 well-heeled acquaintances. Each of them gives him $100, which they deduct on their tax returns. In exchange for the $1,000, Edison gives them ownership of the technology he is about to create. He makes arrangements to buy it back later if it's a smashing success, expecting to have to pay $400 to each partner to make it worth the risk that the new bulb may not work. All shake hands, and the partners leave Edison to his labors. In a year or so, Edison's brilliant idea becomes a marketed product. He sets the price at $1 and starts to sell bulbs on behalf of the owners. As a means of payment for buying back his technology for himself, Edison pays the partners 10? for each bulb sold. Eventually each owner gets his appointed reward and the partnership turns back everything to Edison.

But that's hypothetical. In the real, more complicated business world, the technique of R&D partnership funding came along just in time for publicly held Hadron Inc., a maker of computer-controlled laser production tools. In 1978, when an investment group including now-president Dominic Laiti purchased the Vienna, Va., company from Xonics Inc., its prospects seemed bleak. Hadron stock, which had traded as high as $7 over-the-counter, then was selling around 30?. In the company's 14-year history, sales had never gone over $3 million. It was Laiti's hope that new management could play on the reindustrialization and capital investment themes being trumpeted by the current administration. One thing Hadron did have going for it was 100 patents in laser technology. What the company needed to help make it competitive was some additional products.

But Hadron's problem was that one of the most promising products -- a laser market that could etch code numbers into tiny objects such as silicon chips -- was only an idea. The idea needed much more development, and development cost money. Money that Hadron didn't have.

"We'd bought a bad balance sheet," Laiti admits. "The company had unusual debt for its size, with a current ratio of one to one, and we had a lot of creditors. We couldn't even get near a bank to apply for a loan."

His partner, venture capitalist Earl Brian, realized Hadron didn't need a bank. The situation was tailor-made for an R&D limited partnership. Laiti and Brian wrote a 60-page private-offering memorandum that included an extensive list of the risks involved in developing the invention (competition, marketability, the possible need for additional financing, and many more), arrangements for assigning any new patents to the partnership, royalty payments of 7% of sales, and an option to buy back the technology at a future date for shares of stock. If the project was successful and the option exercised, Hadron ultimately would be able to reap the fruits of its labors, and, possibly, strike it rich.

Because they were offering the deal privately, Laiti and Brian became their own underwriters. Their goal was to sell 14 units at $50,000 each. Within two months they virtually had made it: 13 units were placed. In September 1979, the partnership -- called Laser Associates -- came into being. Laiti and Brian were its general partners; for legal and accounting expenses, they kept 1% of the $650,000 raised.

Hadron's wholly owned subsidiary, Compulaser, which was in the process of developing the marker, turned over its patent licenses to the partners. Before a year was out, Compulaser had manufactured four laser markers and sold them for $100,000 each. Laser Associates received $28,000 in royalties, providing instant cash flow to the partners.

But there was a hitch. To keep the product competitive required some refinements, and there still weren't enough earnings. Undaunted, Hadron went back to the well. With the necessary majority vote from the original 13 limited partners, in September 1980, a second private offering was made as an extension of the first. This time, however, it cost investors $100,000 each to get in. Fourteen (three of them from the first partnership) signed on.

"We figured people would pay more per unit because we now had a product," says Laiti. "We were building on something. It's like running out of money when you're drilling for oil, and toward the end you find a few traces of oil. That oil hole will attract a lot more investors, and you can charge more because there's less risk." The second partnership was sold with the help of some independent brokers (who took a 7.5% fee from the proceeds), and Hadron had a cool $1,295,000 to push onward with.

What did the partners get out of their commitment in addition to their tax write-offs and the $2,153.85 each original investor received earlier? Nine months after the second offering, in June 1981, Compulaser exercised its option to buy back its patent licenses and technology from the second partnership. The payout for each of the 27 units -- whether $50,000 or $100,000 -- was 20,000 restricted shares of Hadron common (which couldn't be sold for two years), plus 20,000 shares of Compulaser common, which had no public market. In early 1982, no investor had yet hit pay dirt. Hadron stock was selling at 2 1/2 bid. The fact that it was down from 5 in six months was academic, since it wasn't tradable until June 1983, in any event. Compulaser had filed for a public offering and issued a preliminary prospectus indicating a $3 selling price.

Thus the best a unit could do -- even if all stock could have been sold in January -- was a payback of $110,000. That's a far cry from the fortunes of the partners in an R&D deal with Metricon, a private Silicon Valley manufacturer of lasers. They put up $750,000 and were bought out by Johnson & Johnson within 18 months for a reported $3.8 million.

Hadron's R&D investors may see greater profits if the stock goes up, thanks to their having kept Hadron in the laser business. "This is not a free lunch," says Laiti, referring to the dilution of his company's stock, "but we couldn't have stayed in the laser market without R&D financing. It gave us the chance to become a growth company without bank loans or in-house R&D financing. To me, that's the crux of government support of R&D through write-offs. It has allowed us to be a viable competitor in a multimillion-dollar market."

Though Hadron expects to post sales of $22 million in its current fiscal year, it hasn't stopped using R&D partnerships. Hadron is going after even bigger money, this time through a public offering. The advantage of the public offering is that Hadron can thus attract more than 35 investors, the limit for smaller private placements. "The disadvantage, though," comments Laiti, "is the amount of time it takes to get the offering okayed by the Securities and Exchange Commission. When we did our private offering, it took three months from the time we wrote up the memorandums to the time we'd sold the units. We've already spent six months on this public offering and we're still not selling."

Hadron's funding saga illuminates one direction that research and development partnerships may take. There are any number of other maneuvers that can be devised by the ingenuity of lawyers and financiers. But all must adhere to the fundamental conditions that define research and development partnerships.

An R&D partnership, ranging from a few people to a few hundred, pools its funds and "hires" a separate company, called the "sponsor," to carry out research and development work on a specified project. Often the sponsor is also the general partner and assumes all obligations of general partners in any partnership; sometimes the general partner is independent of the sponsor. In return for the investment, the partnership owns all the sponsor's output, which usually culminates in a new technology or advanced product. The limited partners have a tax deduction for the dollar amount of the R&D contract. In a separate agreement, the partnership grants the sponsor company an exclusive license enabling the sponsor to market the technology on the partnership's behalf. The sponsor then pays royalties to the partnership out of sales.

At the same time, the partnership grants a purchase option that allows the sponsor to reacquire the technology in its entirety or, alternatively, to walk away from the project. The option can be exercised by paying sufficient royalties to the partnership, by paying straight cash, by supplying stock in the corporation, or through a combination of the three. If the technology fizzles, the sponsor simply lets the purchase option lapse. The partnership swallows the failure lock, stock, and barrel.

But the partnership is not generosity incarnate. Against the risk of being wiped out are weighed potentially big returns if the invention is successfully marketed. A partner stands to make as much as 50% per year, compounded annually, net of capital gains. There aren't that many investments anywhere that will yield as much these days. The partners often can establish long-term capital gains even if the sponsor elects to pay royalties rather than a flat buy-back fee. This, plus up-front tax sheltering, make the risk acceptable to the right (read "rich") investor.

Because royalty rates paid to the partnership are often steep, sponsor companies usually choose to buy back the technology if it's promising. Some deals are now being written in which there is a minimum royalty payment due the partners regardless of sales. This "encourages" the sponsor to market the product aggressively. At the other end is sometimes a maximum aggregate royalty, or cap, beyond which the sponsor is considered to have discharged all obligations to the partnership. If the sponsor decides not to purchase the technology, the partners should then be free to try to market the technology somewhere else.

Many arrangements -- "equity partnerships" -- give the partners a shot at getting equity, rather than cash.Depending on the risks of the project, the number of shares involved will range from around 65% to 100% of the amount a venture capital firm would expect to receive in an ordinary financing. But the partners' total return is enhanced by the tax deductions accorded partnerships through Section 174 of the Internal Revenue Code. The after-tax cost of such shares to the partnership, which can deduct most contributions when made, is at most 50 cents on the dollar. The key to this favorable tax treatment is that the limited partners have a choice in the matter and do not automatically receive stock. Otherwise the Internal Revenue Service may apply its "disguised equity" doctrine to take away the Section 174 deduction.

Occasionally the sponsor will buy back the technology in cash. the payout to the partners in after-tax dollars is often calculated on the basis of a 50% return compounded annually. In many cases, Section 1235 of the Revenue Code -- a special provision that applies primarily to R&D partnerships -- can be utilized. This section allows the inventor (i.e., the partnership) to get long-term capital gain treatment on the sale of a patentable invention even if the inventor has held it for as short a time as one day and no patent has been applied for. Typically a patent attorney will furnish an opinion that the device is patentable before the project is undertaken.

But if such heady rewards seem out of line with usual returns, John J. Hentrich, a Los Angeles attorney whose firm, Riordan, Caps, Carbone & McKinzie, has been actively putting together both sides of R&D deals, points out that "the partnership is assuming the risk that the purchase option won't be exercised." And even if a sponsor pays out in equity, it still can be mutually advantageous. Says Peter F. Zinsli, vice-president of private Rexon Business Machines Corp. of Culver City, Calif., which recently closed a $1-million R&D deal to finance the development of an expandable business computer: "We would continue with our development program even if we hadn't funded it with an R&D partnership. The value of the R&D deal comes when we can finance at a lower price than we could with venture capital. We think we've done that. We're offering to sell [the partnership] stock at a fixed price that is beneficial to us. We hold the cash without interest for two years, and stock dilution is two years out as well."

But an eventual dollar-convertible buy-out such as Rexon's is not necessarily the goal of all successful R&D projects. Partnership tacticians are now taking advantage of Revenue Code Section 351, by which the taxable event of the buy-back can be postponed. Section 351 provides for a reorganization under which it is possible for the partnership to incorporate itself. "If properly structured, the partners may not be obliged to recognize the gain until they ultimately sell the stock of the sponsor company," advises attorney Hentrich.

At least one venture capital concern, Crosspoint Financial Corp. of Palo Alto, Calif., is pushing this situation to a logical conclusion by using it to found a business from scratch complete with top management and marketing experts. Last year Crosspoint forged what its president, John Mumford, claims to be the first financing of a prenatal-stage business that melded a partnership with straight venture capital. In September 1981, a $125,000 partnership was devised specifically to develop a digital voice data switch for the telecommunications industry. The partners were high-level executives and engineers who had been wooed into the formative dealings by the promise of shares in the business once it was set up. Within nine months, the $125,000 had been absorbed in defining the product.

At that juncture, partnership shares were converted into corporate shares and the key employees had their corporation. To fund the development of the product line, Crosspoint arranged a second R&D partnership together with ordinary venture capital. The amounts this time were considerably heftier: $3.5 million came from the partners, and another $1.5 million in working capital from venture investors. The infant Irvine, Calif., corporation was named CXC. Within its first quarter of existence it was employing 50 people -- solid testimony to the efficacy of Tax Code Section 174 as broadened by the Snow decision.

The partner-venture capitalist combination, which many underwriters see as the new wave in R&D deals, gives venture capitalists, Mumford explains, "tremendous leverage provided by Uncle Sam." Otherwise, venture capital would shy away from so unformed a situation. And the corporate owners are left with a substantially greater proportion of stock -- up to 2 1/2 times as much, according to Mumford -- than if a typical venture capital deal had been struck. Furthermore, as a raw start-up device, some CPAs consider an R&D partnership to be a more effective tax set-up than Subchapter S. In a Sub S, tax advantages are proportioned to shareholdings. If, for example, only half the shareholders put up all the capital, they will be able to deduct only their half of the losses.

When it comes to selling the R&D deals to investors, most legitimate partnership creators do not view the technique as a tax shelter so much as a business investment structure that happens to use before-tax dollars.Thus, they tend not to leverage the deals. One CPA -- Nicholas G. Moore, a Coopers & Lybrand managing partner who co-authored the definitive paper on the subject -- says, "The only thing I'm worried about is charlatans in the marketplace. There are some bad deals coming out, particularly in highly leveraged packages. I'd be in favor of eliminating the ability to leverage altogether." And the brief history of R&D limited partnerships already shows that the IRS is looking very closely at leveraged arrangements.

Like any specialized financing evolving from the complex meshing of disparate tax and investment provisions, research and development vehicles will require a number of years before all their ramifications -- positive or negative -- become clear. Right now, they seem mostly good. Extols Ronald Moskowitz, president of Ferrofluidics Corp. of Nashua, N.H., which recently has been investigating partnership financing of new products based on magnetic fluid and crystal-growing technologies, "We considered a variety of sources. My feeling is you match different sources of capital with different purposes. We had several exciting projects in basic technology and development, and the R&D partnership seemed like a super concept. If oil and gas can do it, why not research and development? It's a terrific opportunity because it allows us to expand early without pushing us into a possible loss."

Venture capitalist Charles Kokesh, whose Technology Funding Inc. of San Bruno, Calif., has been writing R&D deals for three years, shares the enthusiasm. "The R&D limited partnership is an incredibly powerful and flexible tool. Within the next four or five years," he predicts, "it will become a normal part of the financing tool kit of most high-tech chief executive officers."

But Kokesh sounds the warning that not all developing products should be funded this way. For one thing, if the deal is not carefully entered into by the sponsor company or entrepreneur, the net received might not be so cheap after all. An outside general partner is apt to rake 20% off the top, leaving only 80% for the company to spend, but 100% to pay back to the investors. Moreover, says Kokesh, very few manufactured products can withstand a 10% royalty rate -- not uncommon in today's market. Kokesh insists that if the gross margin of profit isn't at least 45%, it should not be funded through a royalty-based limited partnership. Further, if a product saddled with large royalties and small margins has to compete with another, unfettered, product in the same company for promotion and advertising, it most likely won't get them.

Unless the various provisions of a deal are structured unassailably, a sponsor company can get itself into deep trouble. One such instance involves a medical electronics manufacturer that was inadvertently done in by the SEC. In 1979, the company entered into an R&D deal that provided for stock payout in lieu of royalties. Two years later, when the company filed an S-1 prefatory to new equity financing, the SEC asked for more information about the partnership. After the agency got the details, it ruled that the agreement really constituted a sale of stock, and demanded that it be restated on the company's books. Under the new accounting, its income statement showed a loss. And with a loss on its books, the stock flotation was doomed. Worse yet, when the S-1 offering was postponed, the company's bank withdrew its line of credit, which was contingent on the stock float.

An inventor himself, Boston's John Taplin, the Super Soaper savior, casts a jaundiced eye at some of the big-money R&D deals being executed on the West Coast for high-tech companies. He thinks investors are so mesmerized by the technology mystique that they hand over their money like robotic sheep. "People who put their money into the Trilogy deal to build a supercomputer are taking more of a flyer than the people who invested in the Super Soaper," Taplin asserts. "Just because it's high-tech doesn't mean they'll end up with a marketable product." Indeed, some analysts of big high-tech R&D financing anticipate that such companies might relegate their more questionable product lines to R&D partnerships, while keeping predictably profitable projects for themselves.

Taplin, of the Taplin & Montle Development Fund, finds that certain entrepreneurs don't need the hefty sums that venture capitalists insist they take. Rather than inflate the requirements to take in the half-million dollars that Michael Kennedy, the dog washer's inventor, was initially offered through a venture firm, Kennedy brought Taplin & Montle a 48-page business plan he had prepared. The financiers were impressed, and matched Kennedy with some interested partners. Thus Kennedy proceeded without a hitch -- and without giving away any business ownership. (He since has raised another $50,000 through an equity placement with friends.) Of the 17 area banks that had turned him down, Kennedy only remarks, "The banks in Boston are too conservative."

Taplin & Montle represent what might be called the unglamorous path that some R&D partnership financings have taken. As unlikely as the products they're considering -- a reclosable bag for potato chips, a higher-protein-yielding bean -- may strike technocrats, their trend is definitely up as well. In 1981, Taplin & Montle arranged 2 deals for a total of $350,000. In 1982, they expect to enter into 10 more deals totaling $2.5 million.

While a properly constructed research and development limited partnership can be what statisticians term a win/win situation -- and at the same time foster jobs and productivity in the overall economy -- a carelessly done arrangement can end up with both sides losing. Investors must be alert to possible tax complications even if there is no leveraging involved. The IRS is currently examining 1,800 tax returns claiming R&D partnership deductions. A number of deductions have been disallowed "because it is obvious that the money was not being used to develop anything new," an IRS spokesperson reports. "We call this reinventing the wheel." Tax accountants recommend that the sponsor company keep detailed records so that R&D expenses can be substantiated. And the IRS, as well as the SEC, may look askance on certain prearranged buy-back provisions, and may require that the income realized from them be treated as ordinary income, rather than capital gains. For these and other areas, such as stock option conversion, it's hard to tell how tolerant the IRS will remain. The most sensible precaution for an investor in this regard is that he have a sharp tax adviser.

Most partnership analysts, however, point out that the best investor protection comes from focusing on the merits of the project itself, rather than its taxsheltering aspects. Since R&D deals invariably are more complicated than, say, initial public offerings, an investor should read the prospectus all the more carefully, should thoroughly investigate the risks involved, and should assure him- or herself that the general partner (an independent, rather than sponsor-affiliated, general partner is preferable in this regard) has carefully appraised the prospects for success.

Despite proliferating precautions, the formula for a foolproof R&D deal from either side of the aisle has yet to be discovered. For both, the sagest piece of advice comes from Kokesh: If you don't know what you're doing, don't do it.