Let's say I offer you a bet on the flip of a fair coin. I'll pay you $2 for every $1 you bet if the coin lands on heads. If it lands on tails, you lose your bet. But we'll do this only once and you have to bet your entire net worth. Would you play?

This is a bet with very large positive statistical expectation to you per dollar wagered, but it's so risky you'd be justified in taking a pass. And you'd be crazy to take the bet if I paid you only 75 cents for every $1 you bet. You'd be making a "sucker bet," one that offered insufficient odds.

It's easy to calculate the expectation of the above bets, harder in games of imperfect information, like poker, and even harder in business, where success can be framed as inevitable in retrospect, and failure framed as impossible to have foreseen. But people who gamble for a living are adept at thinking in probabilistic terms. If they bet on a long shot, they expect to make multiples of their bet should they win. A trader is just another type of professional gambler. And for all their complexity, the models traders use are trying to answer one conceptually simple question: is the trader getting a fair or rich price to take on some risk? If not, he or she should take a pass.

A lot of ink has been spilled on JPMorgan Chase (JPM)'s $6.2 billion "London whale" trading loss since it occurred just over a year ago, but as far as I can tell, no one has written about its apparent cause: a large, crazy bet JPMorgan made in 2011 that just happened to pay off. And that means no one has considered the troubling implications.

JPM's Synthetic Credit Portfolio, housed in the bank's Chief Investment Office, made $453 million in 2011, and this trading gain made the CIO eager to employ its winning trading strategies on a greater scale. How did it do this? According to the exhaustive report the U.S. Senate's Permanent Subcommittee on Investigations released March 15, in the fall of 2011, one CIO trader, Bruno Iksil (later to be known as "the London whale") spent about $1 billion buying credit protection that would expire worthless in four months or less unless a corporate default occurred among 100 higher-risk companies. Three weeks before the bet's expiration, American Airlines declared bankruptcy, and the portfolio made $400 million. It would have lost money for the year otherwise (see pages 53-56 in the report).

So to recap, the CIO bet $1 billion and earned $400 million. In expectation terms, that means unless the CIO believed there was at least a 72% chance of default, or unless multiple defaults would have led to meaningfully bigger profits, its sizable bet had negative expectation. A sucker bet. Did Iksil think a default was more than 72% likely? And even if he thought default was that likely, shouldn't he still have feared the giant loss he'd incur if he were wrong, like in the coin flip discussed above?

An internal memo from the Senate report's exhibits shows CIO head Ina Drew downplayed the riskiness of this bet when she explained it to JPMorgan Chairman and CEO Jamie Dimon (Exhibit 84a). That buys him political cover. With JPMorgan's 260,000 employees and 2011 earnings totaling $19.0 billion, Dimon shouldn't be held responsible for the actions of a single employee. Or should he? Few individual employees can affect JPMorgan's earnings number on their own. Arguably Dimon can't in the short run. But Iksil could; his "winning" bet comprised around 1% of 2011 earnings. Which implies Dimon should have been watching. But Drew, who was paid $14 million in 2011, and whose job it was to watch Iksil, wasn't angry about this trade. She was a fan of it (report, page 55). She pushed Iksil to replicate it in 2012, and was quoted as saying that the CIO "likes cheap options" (report, page 63). That ultimately drove the CIO to lose $6.2 billion, or 17% of what JPM earned in 2012, in a benign trading environment.

Why did JPM make enormous high-risk bets, and bad ones at that? In poker, as in trading, luck plays a role. Professional poker players know to separate skill from luck. A pro who misplays a big hand but is saved by a miracle card might breathe a sigh of relief, but he doesn't pat himself on the back, because he knows that if all he has to rely on is luck, he'll be a long-term loser. Were Drew and her team familiar with this concept? Is Dimon?

There have been rumblings that Dimon might resign should JPMorgan shareholders vote next week to strip him of the Chairman role. JPMorgan shareholders should take that risk seriously. JPMorgan's stock price has grown 28% since Dimon became CEO, whereas it fell 17% during prior CEO William Harrison's six and a half year tenure. But JPM's all-time high of $65.67 occurred on Harrison's watch, in March 2000, when the bank had $391 billion in assets, versus its current $2.39 trillion. On Thursday, JPMorgan shares closed at $50.97, 22% below the high of 13 years ago. Wells Fargo and U.S. Bancorp, both large, high-performing banks, are far closer to their respective all-time highs – highs which occurred in 2008.

The Chairman/CEO split issue is a sideshow. Dimon is unquestionably among the best CEOs in financial services, if not the best. But he hasn't been able to get JPMorgan's stock price above where it was in 2000. Is this a reason for why the firm was (or is) gambling?

To put it another way to those who are especially nervous about Dimon's potential (although at some point inevitable) departure: what big opportunities can only Dimon help JPMorgan exploit, and what big mistakes can only Dimon keep JPMorgan from making? He's spent over seven years improving JPMorgan as CEO. After all that time and effort, is he still the only person capable of running the bank optimally?

Harvard Winters, a former investment banker, writes research on banks.