It has been four years since Lehman Brothers failed, AIG was rescued and credit markets seized up. Can we avoid another crisis and, if so, how?
Banking is now a system too complex and interlinked to survive the failure of anything but a small, insignificant bank. So if we are to avoid another global financial crisis, we need to find a way to let a bank fail without it bringing down the whole shebang.
In his book Adapt: Why Success Always Starts with Failure, Tim Harford recounts how a domino-toppling record attempt was ruined when a pen dropped from a person's pocket and knocked one over. Domino topplers then introduced safety gates, "removed at the last moment, to ensure that when accidents happened, they were contained".
The banking system needs such safety gates. As Harford argues, complex rules managing complex systems increase rather than reduce risk. A complex, interconnected system requires simple rules.
The two main policy responses to the crisis fail this test.
In 2010, the Dodd-Frank Bill was passed, designed to "reform" the US banking system. The draft ran to 1300 pages. The bill was 2600 pages long and spawned 65 studies, 243 new rules and 100 committees.
Basel III is better. It targets a range of measures to protect "systemically important banks" and runs to 616 pages, although none explain what "a systemically important bank" is. Banks will be forced to hold common-equity capital of at least 7 per cent of their risk-based assets (it was just 2 per cent before the GFC), reducing their ability to lend.
This is a more useful effort than Dodd-Frank, mainly because Wall Street lobbyists were less effective in hobbling it, but it is still not a true safety gate.
Economist E.F. Schumacher is believed to have said: "Any intelligent fool can make things bigger, more complex and more violent. It takes a touch of genius - and a lot of courage - to move in the opposite direction."
Basel III forces banks to hold more capital and contains a mechanism discouraging credit growth in boom times and encouraging it when times are tough. These are sensible, simple measures. But they are not enough. An enforced separation of critical retail banks, which hold depositors' money, from more risky but less systemically important investment banks, is also necessary.
The Glass-Steagall Act did this well for more than 60 years but thanks to bank lobbying, was repealed in 1999. Even Sandy Weill, former head of Citibank and once a lead campaigner against it, has called for its reintroduction.
We also need a far quicker way of resolving bank failures.
Sweden, during its bank crisis of the early 1990s, showed how important it was for regulators to be able to take over banks and quickly restructure them. It emerged from a bank crisis quite quickly.
Thus far, there is not much evidence of genius or courage.
Worse not better
There are plenty of other ideas. A banking crisis starts with cheap credit. Crises would be less likely if central banks did not make credit so readily available at the first sign of trouble.
Some call for reform of the short-term bonus culture that encourages risk taking by the banks but leaves those risks to be carried by someone else down the track. Others want to address the problem of "too big to fail" by breaking the banks up into smaller entities.
As for regulators, it would help if those charged with overseeing the banks were not previously employed by them, and vice versa.
And if political parties could not raise money from the banks, we would stand a better chance of getting the regulations we need.
Right now, the problem is worse rather than better. Before the GFC, banks were not sure that if they made stupid loans governments would bail them out. Now they do.
Dodd-Frank and Basel III may lower the probability of a systemic collapse a bit but nowhere near enough. That is why we recommend you don't have more than 10 per cent of your portfolio in bank shares, including hybrids such as ANZ CPS3 and CBA Perls IV.
If you own insurance stocks and investments in other financial companies, make sure the total does not exceed more than 25 per cent of your portfolio.
This is not about being scared of what might happen, but planning for the possibility of it. We have not fixed this problem and your portfolio's structure should reflect that fact.