A pair of mining investors amassed a $235 million fortune as metal prices tumbled. Now they're risking it all in a last bet that gold--one of the hardest hit metals--will rise again. It seemed like a bum deal. The Stillwater palladium mine in Montana was only partly built, and Russia was expected to release vast stores of the metal shortly. Prices were sure to tumble. Or so went conventional wisdom. Seymour Schulich of Franco-Nevada Mining, who had worked for four years to get a piece of Stillwater, thought the Russian threat was exaggerated. So when a royalty on the mine came up for sale in April 1998, he offered $36 million, topping the next-best offer by 20%. "Opportunity comes to those who are patient and who can write a big check," says the 62-year-old. Over the next four years palladium prices quintupled, to $1,085 an ounce. The mine generated $17 million in royalties, nearly half Schulich's cost, and should be good for $12 million a year over 50 years. It would be easy to dismiss the Stillwater win as sheer luck if the frugal Schulich and his French-Canadian colleague, the 54-year-old Pierre Lassonde, had not pulled off plenty of similar stunts before. Through uncanny bets on commodity prices and shrewd royalty deals, Franco's bosses have managed to return 38% annually to shareholders since taking Franco-Nevada public 18 years ago. What's more, the two pulled this off in gold mining, one of the toughest places to make money in the last two decades. The duo's 9% stake in Franco is worth $235 million; now they're betting it all in a final wager: that history will reverse and gold will rise again. They arrived at that bet after months in which the fate of Franco-Nevada hung in the balance. Last April the Toronto-based company bought 20% of Normandy Mining of Australia. A few months later Normandy became the target of a bidding war between two gold-mining giants. The value of Franco's stake in Normandy has since more than doubled. Now shareholders of the recently crowned winner of Normandy-Newmont Mining of Denver, Colorado-will vote on whether to add Franco to their purchase. If they do, the three-way deal would give tiny Franco, with just 21 employees, a 32% stake in the world's largest gold company, which employs 14,000. Newmont Mining is for gamblers--it's one of the few gold producers that doesn't sell much of its output forward to hedge against a fall in gold prices. This means big profits if gold rises, but if the price drops far enough, Newmont's fortunes vanish. Franco, which went public in 1983 as a gold-exploration company, started inauspiciously. The two investors sifted through 43 deals with little luck. Then Schulich, who had made millions buying oil and gas royalties, had an idea: If royalty deals work for oil and gas, why not gold? The two floated the idea to Canadian and American investors but got the cold shoulder. Royalties in gold were traded almost exclusively among mining companies, and no one had tried to break into the business before. The two decided to make a run at it anyway, using $930,000 from a secondary stock offering in 1985. Their first deal, in 1986, was for the rights to a 4% cut of revenues, plus a few smaller royalties, on land in Nevada where a mine run by Western States Minerals was pulling 44,000 ounces of gold annually from near the surface. The two offered $2 million for the royalties, half the cash on Franco's balance sheet, figuring that the gold already discovered would allow them to break even someday whether there was gold deeper in the ground or not. No one else bid. Within a year more gold was discovered, and annual production doubled. The royalties now generate $23 million annually, and reserves stand at 24 million ounces. "We like to buy annuities," says Schulich, "but with liens attached on anything else that's found." Two years later Franco bought a royalty on a gold mine in California called Castle Mountain. It turned out to be a moneyloser for the operator and was closed. Yet Franco itself did fine--it had already collected triple the $2.8 million it had paid. Franco went on to buy royalties on diamond mines, platinum mines and oil and gas deposits, but gold is where it shines. The two founders complement each other. Schulich instinctively looks for ways to protect the downside in a deal. Lassonde, once Canada's top gold-fund manager, has a good feel for where metal is buried. The contrast extends to their lifestyles. Schulich drives a seven-year-old Lincoln and lives in a house he bought 25 years ago; Lassonde has a $232,000 Ferrari and a new mansion in Toronto's tony Forest Hills. Yes, luck certainly does play a role. The only mine that Franco has owned outright--Midas, in Nevada--is pulling $65 million worth of gold a year from a narrow but long and deep deposit. If Franco had shifted 2 meters to either side of the spot on which it chose to sink a drill bit, it would likely have had no clue that any gold was below. A contrarian instinct helps, too. Shares in Voisey's Bay, a nickel mine in Labrador, Canada, fell from $27 to $4 in the two years after speculators panicked over falling nickel prices and political setbacks to the project. In October 1998, with the stock near its low, Lassonde invested $50 million. He sold the shares for $55 million plus warrants now worth $10 million. "People have visions of sugarplums. You bet against them," says Lassonde. "Fear sets in. You bet against them again." In April, Franco announced the sale of Midas to Normandy for 20% of that company's shares and a 5% royalty on the mine. In September AngloGold of South Africa offered 74 cents a share for Normandy. The deal would have given Franco a 60% gain, but Schulich and Lassonde figured that they could get more. They lobbied Newmont to put in its own bid and buy all of Franco, too. Newmont bid three times, finally offering $1 a share for Normandy. Newmont will also hand over $2.5 billion of its shares to Franco's shareholders for absorbing that company. If all goes according to plan, Lassonde will become Newmont's president and Schulich a director. They intend to sell $700 million of assets and pay down debt. And by choosing not to hedge, they will be betting their fortunes that gold will rise. Schulich bases his bullishness on demand. He thinks that, given the high U.S. current-account deficit, the dollar will fall soon. The price of gold will then rocket upward, as investors seek safe havens. Lassonde traces his bullishness to supply. He estimates that more than a third of the new supply of gold, excluding central bank sales, will disappear, thanks to shuttered mines. Short-sellers will scramble to cover. And what if gold goes nowhere? Schulich and Lassonde won't suffer. They are selling Franco at a rich 35 times estimated earnings this year before taxes, depreciation and capital expenditures on new mines. They are buying Newmont at only seven and a half times. That's not luck. That's just plain shrewdness.