With the banking stress tests now completed, investors are finding it more sensible than it's been in months to invest in banks. Clarity, they'll tell you, is returning to the industry. We can now make informed estimates as opposed to the blind guesses we were stuck with in months past. Not only do we have a roadmap of a bank's future, but heck, the stress tests prove that no big bank will completely bellyflop and go under, right?

Kinda. Sorta. Not really.
Some will say yes. I still think the recent rally in bank stocks is grossly overcooked. Contrary to the litany of comments and e-mails, it's not because I'm a "paid basher trying to cover my shorts." It's because the stress test data that's been behind banks' explosive run is blindly running in the face of two objecting points:

  • Investors misinterpreting the data as meaning that the worst is over.
  • The forward-looking assumptions used in the stress tests may have been far too optimistic to begin with.

Here are two points you should consider before assuming banks are out of the woods.

1) What did the stress tests really show?
Last week I wrote, "Raising capital is designed to help [banks] through the 100-year storm, but that storm will still make their lives miserable for years to come." 

In other words, the stress tests demand that every major bank wear a bulletproof vest, but just the fact that they're wearing bulletproof vests should tell you something. They're marching straight into a war zone, and it ain't going to be pretty.

To see what I mean, look at the estimated losses the stress tests assume that major banks could face over the next year and a half:

Bank

Estimated Losses Through 2010

Citigroup (NYSE:C)

$104.7 billion

Bank of America (NYSE:BAC)

$136.6 billion

JPMorgan Chase (NYSE:JPM)

$97.4 billion

Wells Fargo (NYSE:WFC)

$86.1 billion

Goldman Sachs (NYSE:GS)

$17.8 billion

Morgan Stanley (NYSE:MS)

$19.7 billion

American Express (NYSE:AXP)

$11.2 billion

Source: Federal Reserve.

Note that the estimated potential losses over the next 18 months are considerably larger than the losses taken over the past 18 months. In fact, the cumulative loss estimates of the 19 banks that underwent the stress test is $599 billion by the end of 2010, compared to less than $400 billion that the same banks have recognized since mid-2007.

Don't confuse what that's saying: In terms of losses and writedowns, the next 18 months are expected to be worse than the preceding 18 months.

Take that any way you wish. I don't see how you can put a positive spin on it, regardless if the said banks now have enough capital to avoid death. The ability to avoid death doesn't mean you're healthy. Nor does it make you worthy of an investment. Not by a long shot.

Granted, some banks were pricing in death before the rally began. But that's far from the case today: Wells Fargo trades where it was in late 2007; Bank of America trades where it was before we realized what a mess Merrill Lynch is. Banks aren't pricing in death -- they're pricing in perfection. That's scary.

2) Besides, who says the stress test assumptions will be accurate?
A more pressing question might be whether the assumptions -- and that's all they are -- were stressful enough. Bank CEOs will tell you they were. Bank of America CEO Ken Lewis said of the stress test: "Never has a test been so aptly named."

But many less-biased analysts feel quite differently. As the Wall Street Journal recently wrote:

… the Fed used a different measurement of bank-capital levels than analysts and investors had been expecting, resulting in much smaller capital deficits … Bank of America's final gap was $33.9 billion, down from an earlier estimate of more than $50 billion … Citigroup's capital shortfall was initially pegged at roughly $35 billion, according to people familiar with the matter. The ultimate number was $5.5 billion.

The old saying about financial modeling, "garbage in, garbage out," should start resonating right about now.

Another, more shocking example of how optimistic the stress test assumptions are comes from estimated unemployment rates. The Treasury's "worst-case" scenario used in the stress tests assumes that unemployment could hit an average of 8.9% in 2009. (Have a look.)

This should sound quite peculiar to anyone who saw last week's labor report, which shows unemployment currently sitting at 8.9%.

When the "worst-case" scenario is already happening, it's disturbingly obvious that the stress test aimed low in order to achieve the desired results. This should be unsettling for anyone taking the stress test's word as gospel.

And judging by banks' recent performance, many people are.

For related Foolishness:

Fool contributor Morgan Housel doesn't own shares in any of the companies mentioned in this article. American Express is a Motley Fool Inside Value recommendation. The Fool owns shares of American Express. The Fool has a disclosure policy.