Two weeks ago, I gave four reasons why you should avoid banks. All of them. Anywhere. Simply, accounting gimmicks and industrywide disclosure shenanigans often mask banks' true health. What you as an individual shareholder see is just a fraction of reality.

Lehman Brothers' use of so-called repo-105 transactions to reduce its balance sheet during the days it reported to shareholders is a prime example -- but hardly unique. In the article, I quoted a former senior investment banker as saying:

Before the [leverage] ratio was published at the end of each quarter, investment banks would take the necessary steps to sell enough of the assets to get the leverage down to a more "acceptable" ratio.

In the meantime leverage "often approached 50:1 during the middle of the quarter."

According to the Wall Street Journal, data from the Federal Reserve Bank of New York proves the prevalence of this practice. A study found that 18 banks, including Goldman Sachs (NYSE: GS), Morgan Stanley (NYSE: MS), JPMorgan Chase (NYSE: JPM), Bank of America (NYSE: BAC), and Citigroup (NYSE: C), "understated the debt levels used to fund securities trades by lowering them an average of 42% at the end of each of the past five quarterly periods."

They intentionally (albeit legally) misled shareholders, in other words.

To explain this further, the balance sheet -- far and away the most important piece of data when analyzing a bank -- is only reported to shareholders on four days each year -- the last day of each quarter. During the other 361 days, investors have no idea what's going on inside the company. Banks (now probably) utilize that blackout to leverage up to juice returns. That makes a bank's return on assets look higher than it really is, because net income is compared with the reported balance sheet figure.

Here's a simple example. Say a bank is levered 20-to-1 during most of the quarter, and earns $100 million on $4 billion of assets. That's a return on assets of 2.5%. But if that bank then sells assets just before reporting to shareholders, bringing leverage down to 10-to-1 and total assets down to $2 billion, then suddenly it achieves a 5% return on assets. Managers look like geniuses (and pay themselves accordingly), but in reality, the reported results were far riskier and unimpressive than shareholders perceived them to be.

None of this surprises me, and shouldn't surprise you, either. More than any other industry, banks have proved that what you see isn't always what you get.