Systemic risk is not caused by just a few institutions. It does not result from any of them being "too big to fail."

Systemic risk is driven by the aggregate balance sheet and operations of the industry and cognate non-bank entities. The Fed's most important safety and soundness responsibility is for the banking system, not for individual banks.

When the financial industry as a whole gorges itself on a new class of assets, or when the overall quality of an asset class drops precipitously, then true systemic risk is created. A phenomenon to which, up to right now, the Fed has been blind.

That lesson could have been, but evidently wasn't learned from the role of mortgages (not just the much maligned "sub prime mortgages") in the past decade. Whether or not housing prices dropped in a particular year, bad mortgages were far more prevalent in 2006 than in 2001. If the bubble had burst later, it would have been an even larger and more destructive bubble.  The fact that no one expected housing prices to decline explains nothing.

We also should have learned that bad ideas and practices, especially those offering the hope of quick profits, are highly contagious and can proliferate dangerously rapidly through the industry. Money is fungible, lenders are fungible, and weakened lending standards invite emulation. "The others are doing it" may be good enough for the boardroom. It shouldn't be good enough for regulators.

This could have happened even if there were 100,000 banks each having equal market shares. In fact, it could have happened more easily in that case, because regulators would have had to expend their resources less efficiently and examine assets in much less depth. Common asset quality weaknesses could easily be passed by examiners who judged that "This bank is not particularly bad."

Since U.S. legislators and regulators failed to learn how banks' shared delusions generate systemic risk, it's not surprising that European banking regulators also missed the point, which for them is:

Sovereign euro debt of Italy is not, for example, risk free to a French bank. Total holdings of Italian debt by French banks can pose systemic risk to the French banking system.

Sovereign euro debt quality sank, even though ratings lagged. Dangerous accumulations were built up by banks that, finding themselves in similar circumstances, made similar decisions. Sound familiar?

The resulting systemic risk arising from Italian debt does not depend on a couple of huge institutions going hog wild buying it. Rather, it is proportional to the total amount of this debt in the portfolios of financial entities in one country or in this case across a group of countries.

If just one "systemically important" institution with 10% of national deposits had choked on Greece or Italy, the outcome would be much less grim. What's systemically important is the aggregate assets and expenses of the whole system, not 10% of it.

This systemic risk, which now shakes the financial system, is all the more inexcusable because anyone who was able to see that Greek or Italian debt could be even a little bit more risky than German debt should have been able to see also that Greek debt could potentially become a whole lot riskier. Historians may well say that the decision to promote a partial default by Greece sparked the detonation of this entire rotten structure. (Evidently, in the end Greece was not TBTF. It was too big to be saved from failure. Ha!)

Design and implementation of stress tests embraced the same error, by focusing on individual institutions. Stress test content is primarily political, with their details indecipherable. Stress testing of the whole system — with aggregate risk judgments based on sampling — might be a sensible idea, involving a lot less guesswork and inconsistency. This would enable a focus on particular asset classes or other activities causing fulminating systemic risk.

It still might have been possible for the euro to fail without causing international banking systems to fail — if the total risk to banking generated by cross-border investments had been controlled. But a sappy emotional preference for promoting further European integration, compounded by gross negligence, blocked that exit.

This isn't about having 2% more capital. There can never be enough capital if assets are contemptible. However, the measurement of aggregate risk trends could lead to more rational assessment of aggregate and individual bank capital requirements.

To prevent the next disaster, focus on the banking system as a whole. Locate and address its vulnerabilities.

Andrew Kahr is a principal in Credit Builders LLC, a financial product development company, and was the founding chief executive of First Deposit, later known as Providian. He can be reached at