Here was the proposal: A railroad in Montana wanted to take advantage of stratospheric electricity prices by generating lower-cost energy from idle diesel locomotives. A power company, Commercial Energy of Montana, would put up $1 million -- half the cost of retrofitting the engines to produce power -- as well as handle sales and schedule transmission of that power to customers. The power company's founder and CEO, Ron Perry, saw three potential gains: a 50% share in the venture's profits, a deeper relationship with an important customer (the railroad), and a chance to try generating his own power, a cornerstone of the company's long-term business plan. But Perry knew one other thing about the plan: It was very risky.

The energy market was volatile, the application of trains to power creation unorthodox. Still, Commercial Energy -- which honors prices in customer contracts even while electricity prices yo-yo madly -- lives and breathes risk management. The company uses Crystal Ball, sophisticated software that maps Commercial Energy's financial performance under thousands of energy-price scenarios and calculates the likelihood that those scenarios will occur. Mulling the new venture, Perry fed the software every variable he could imagine, including the range of uptimes and operating costs for the engines, possible fluctuations in the price of electricity and diesel fuel, the likelihood that those prices would move in opposite directions, and potential customer demand and contract prices. "We determined that there was a small chance of a $10 million upside if electricity prices stayed high, with the greatest probability of making $2 million to $3 million in six months," says Perry.

Of course, spreadsheet cells can't accommodate the vagaries of human nature or government fiat. Diesel conversion is a messy affair, and some neighbors balked at having a generation site near their property. Local environmental groups joined the choir, the issue wound up on NBC Nightly News, and the railroad -- spooked by bad PR -- wavered. Then the Federal Energy Regulatory Commission capped electricity prices at $100 a megawatt, further undercutting the potential profits. Ultimately, the project collapsed, and Perry lost a little over half his investment.

Nonetheless, Perry believes he did everything right. "We knew there was a chance we could lose up to $1 million, but we also knew we weren't going to lose more. I could afford to risk $1 million for the opportunity to make 10 times that."

Risk experts would probably agree that Perry did the right thing. It was a good decision not just because the potential upside was high but also because he got comfortable with the potential downside up front. The ability to recover from loss should be the first consideration for entrepreneurs wrestling with uncertainty, says Peter Bernstein, author of Against the Gods: The Remarkable Story of Risk (John Wiley & Sons, 1996). "It's important for small-business owners to distinguish between risks that are reversible -- liquidity is a key element here, for example -- and risks that are difficult to reverse," says Bernstein. "In an uncertain life, being wrong has to be expected on occasion. Therefore the consequences of being wrong must dominate the probabilities of being wrong. If you are betting the ranch, you had better be able to pick up the pieces if things turn out differently from what you expect."

"It's important for small-business owners to distinguish between risks that are reversible," says risk scholar Peter Bernstein, and risks that are not.

Bernstein's advice echoes Pascal's 350-year-old rationale for believing in God: If you wager against His existence and are wrong, the inferno is a tough spot to bounce back from. Today the profile of risk, overall, has grown comparably portentous. Shadowed by war and terrorism, the risks of once-trivial bets -- that your plane will stay up until it's time to come down, for example -- appear disproportionately dire. The possibility of catastrophic loss nests in the back of our brains even as we pursue the most mundane daily tasks. It's like being constantly aware of your tongue.

Business folk, for their part, see the fields of commerce liberally sown with mines. But in this riskier world, it's important to remember that not all risks are alike. There are "intrusion" risks -- risks of ungovernable disruptions by God or man -- and they seem potentially greater and less calculable than ever. The other kind of risk, though -- call it "decision" risk -- is both more useful to think about (because we affect it directly by our choices and actions) and, better, is actually getting easier to govern, thanks in part to the intensified attention risk is getting from software developers, political economists, psychologists, and others.