Last week, we introduced the idea of having a selling discipline for small-company stocks, as opposed to solely a buying discipline. Actually, the major "discipline" at all in investing is associated with selling rather than buying. If you do your research correctly and have sound judgment when looking at a stock, reaching a conclusion whether to buy it is relatively simple. Determining when to sell, however, tends to be a more complex question.

To form some sort of selling discipline for small-company stocks, it was suggested that a good starting place might be to look at how small companies themselves come to selling decisions on the corporate level. There seem to be two main rationales for selling part or all of a small company -- deteriorating growth prospects or economic returns from the business, or when a buyer offers a great price. We'll examine the first rationale here, using the recent sale of former small-cap high-flyer 3dfx Interactive (Nasdaq: TDFX) as our main case study.

Long-time readers of The Motley Fool are most likely familiar with 3dfx, the developer of the popular Voodoo line of PC graphic cards that used to be a holding of the old Fool Portfolio (now the Rule Breaker Portfolio.) While our records show that the company never actually made it onto the Rule Breaker Portfolio monthly Foolish 8 list during its glory days in late 1997 and early 1998, it passed through most of the old Foolish 8 screens (they have since been updated) when the Fool Port bought the stock in January 1998. At that time, the company sported a market capitalization of $260 million. On Dec. 15, 3dfx said it would sell most of its assets, essentially intellectual property and chip inventory, to rival NVIDIA (Nasdaq: NVDA) for a total consideration of $112 million in cash and stock, and dissolve the rest of the company.

What motivated the company to sell its major assets to NVIDIA, and how did it reach its final sell decision? Management pointed to deteriorating business quality and growth prospects in announcing the news to its shareholders. "[H]igh inventory expenses, decreasing margins, and slowing demand have done irreparable harm to 3dfx," the company stated. Investors could have tracked these elements and seen that 3dfx was running itself into the ground, but there was an even easier way to measure the company's decline. The key area to look at was the business returns, or the amount of money being generated in relation to the amount being expended. Here's how these two variables have looked over the past few years for 3dfx.

Voodoo Economics

($ thousands)  Oper. Cash Flow    Cap. Ex.

2000 YTD          (31,463)         13,711
1999              (33,456)         25,733
1998               16,763          13,844
1997                 (255)          4,730

Total             (48,411)         58,018

As you can see, $58 million was poured into the business over the past 3 3/4 years, only to rack up $48 million in negative operating cash flow, both of which are elements included on the company's statement of cash flows. In fact, 1998 was the only year operating cash flow exceeded capital expenditures, and that was somewhat suspect since the firm's cash flow benefited from a large one-time legal settlement. With its business returns not improving and its growth prospects deteriorating, salvaging the company's remaining valuable assets through a sale appears to have been the rational conclusion for 3dfx.

The 3dfx cash situation is exactly opposite of what a small-cap investor wants to see. Ideally, investors want to own companies that regularly have money left over after all business-related spending is taken into account, including operating costs such as salaries and R&D that are expensed on the income statement, and capital expenditures for future growth that are recorded on the cash flow statement. The idea is to look for companies that are lousy with cash -- lousy in the sense of being "amply supplied" with cash rather than "miserably poor or inferior" with the cash they have.

Focusing on business returns is important to a small-company investor in particular, since small businesses with bad economics don't stick around for long. Most small companies don't have the financial stability or track records needed to garner high credit ratings and tap the debt markets for a low-cost source of cash. That leaves the equity market and commercial bank lending as the primary financing windows for small firms, and both of them will slam shut if a company's returns stay poor for more than a handful of quarters.

So, when should an investor focused on business returns have sold 3dfx? It became rather apparent in 1999 -- if not sooner, if the one-time item in 1998 is eliminated -- that the company wasn't making the grade in terms of its business returns. That insight would have correctly justified a sale practically any time over the past three years. In fact, adopting a business-return discipline from the outset would have probably kept a small-cap investor from buying a company with 3dfx's cash economics in the first place.

The upshot of the 3dfx example is that small-company investors need to think about their companies in terms of cash-on-cash returns at all times. This means evaluating not only the money that is coming out of a business as cash flow, but also the money that is being regularly put into it. In this way, a focus on business returns becomes a simple small-cap selling discipline, as well as a useful guide to initial buying decisions.