Next Tuesday, at its annual meeting in Tampa, JPMorgan (JPM) will reveal the results of a shareholder vote on whether the bank should split the role of CEO and chairman. Right now, Dimon holds both. Many shareholders think he should give up the chairmanship. The push for the move has gained momentum in the wake of last year’s $6 billion trading loss and a litany of other run-ins with regulators in the past 12 months. The vote is expected to be close.

The main argument that Dimon and his supporters make for why the CEO should keep the chairman role as well is JPMorgan’s bottom line. Bank analyst Dick Bove recently noted that only three companies in the U.S. netted more than JPMorgan last year. And that’s in a year in which the bank had a $6 billion trading loss.

MORE: Jamie Dimon’s successor short-list will not impress critics

Dimon has indeed steered his ship toward the money waterfall while others nearly capsized and have teetered since. But at least part of the bank’s profit gusher is an illusion.

Take the first quarter of this year. JPMorgan said its income rose 31% to $6.5 billion. But more than two-thirds of that gain can be attributed to some nifty accounting coming out of the bank’s mortgage division and not any real improvements in JPMorgan’s business.

In fact, in the first three months of the year, JPMorgan’s losses from potential bad home loans was negative $198 million. That means the bank booked a near-$200 million gain based on a prediction of how much it would lose from future bad loans.

How’s that possible? Accounting, as you may already know, is more art than science. Even so, JPMorgan’s books are more Jackson Pollock than Norman Rockwell.

Banks aren’t forced to write off bad loans when a borrower stops paying. Instead, they are allowed to wait until they have made the determination that they will never get paid back. That typically happens weeks or, more often, months after a borrower ceases mailing in checks. Banks get a lot of leeway in making that determination. Until they do, the bad loan sits on the bank’s books just like any other, delaying losses.

MORE: 3 things Jamie Dimon might have meant when he said he was richer than you

JPMorgan has long been slower than rivals to recognize bad loans. In the first quarter of last year, for instance, JPMorgan wrote off one dollar for every 90 the bank had in bad loans. That compared to an average write-off ratio at the nation’s 100 largest banks of 34 at the time. This year, that ratio at JPMorgan has shot up to 178, meaning it is writing off bad loans at nearly half the rate it was a year ago. Writing off loans at that rate rather than the industry average appears to have saved JPMorgan $697 million in the first three months of the year alone.

Other big banks have slowed their write-off rate as well. But none is nearly as slow in recognizing losses as JPMorgan. The ratio of bad loans to charge-offs at Bank of America (BAC) and Wells Fargo (WFC) was 124 and 113, respectively.

That’s not the only boost JPMorgan got from its accountants. On top of taking fewer loan losses, the bank also pulled money out of the fund it has set aside to cover future bad loans. As a result, an additional $650 million made its way to JPMorgan’s bottom line. The result: What could have been a loss of $1.15 billion from bad loans turned into a gain of $200 million.

MORE: Why only half of Americans gain from the stock market rally

JPMorgan spokeswoman Kristin Lemkau says the bulk of the company’s earnings gains have come from true improvements in its business. “We have had three years of record performance driven by market share gains in every line of business,” says Lemkau. She says that Dimon has regularly told investors that he does not consider loan reserve releases as quality earnings.

Of course, the economy is improving, and house prices are rising. So you would expect fewer losses. Still, the bank’s nonperforming loans have dropped by only 11%, far less than the 50% drop in write-offs. What’s more, the volume of home loans at JPMorgan on which borrowers haven’t made a payment in more than three months suddenly jumped $3.5 billion in the first quarter. That’s the first jump in seriously delinquent loans the bank has seen in more than three years and could be a sign of more problems ahead.

Even without these accounting moves, though, the bank’s bottom line would have been up $600 million or 10% in the first three months of the year. So it’s hard to make the case that Dimon should go. But earnings at Citigroup (C) and Wells Fargo rose 31% and 22%, respectively. That puts the indispensability of Dimon in perspective.