Nearly five years after the onset of the financial crisis, we might expect widespread agreement on what went wrong and how to fix it. But there is still a lot to be done, with some new thinking required.

In the UK, the two biggest banks are still in government administration, the prudential regulator judges that the largest banks are still under-capitalised to the tune of £27 billion and one observer says Britain's GDP will be 'one-sixth lower in perpetuity' as a result of the crisis.

The starting point of the Parliamentary Commission on Banking Standards was the 2012 LIBOR scandal, so its focus was on this 'shocking and widespread malpractice'. It identifies the key problem as distorted incentives in salaries and bonuses, and its recommendations are designed to change banking culture through sharper lines of responsibility and heavier penalties (including the criminal offense of 'reckless conduct in charge of a bank').

We can wish them luck. While many failed bankers left their jobs with outrageous severance payments, just about all top management in failed institutions did, in fact, have substantial 'skin in the game'. But as they were already so rich and were often gamblers by nature, they were always ready to spin the dice again. Deferring bonuses won't fix the culture. In any case, when the whole sector is behaving badly, it's hard not to join in. Citibank's Chuck Prince was only reflecting competitive reality when he said: 'as long as the music is playing, you've got to get up and dance'.

In principle, more bank capital will help. This is the buffer that absorbs losses when it turns out that assets aren't worth so much. But we can't agree on how to measure it, or how much is required.

The old prudential rules (Basel II) had already put in place risk-weighted capital requirements (safe assets required less capital). Under these, Deutsche Bank had a simple leverage ratio of 50 ($1 of capital for $50 of assets; see graph below), a patently inadequate buffer.

The response to egregious 'gaming' of these risk-weighted ratios has been to add complexity. Basel III capital requirements are described by a former US regulator as 'hopelessly complicated'. The Basel regulators have so little faith in risk-weighted capital ratios that, in a belt-and-braces approach, a simple leverage ratio (without risk adjustments) will also be required. This, however, puts the incentives in the wrong place: a simple ratio encourages banks to do higher-margin risky lending. And the ratio will be only 3%, so a fall of 3% in asset values would extinguish all capital.

Spain provides a reminder of just how much capital would be required to make a system safe: before the crisis, Spain was the poster-child of capital ratio discussions, having forward-looking dynamic capital provisioning. Not enough, it transpires, to save the banks (or the Spanish taxpayers).

The European economy is so weak and dodgy assets so ubiquitous that it would be hard to have any confidence in banks. Tentative steps have been taken to adopt euro-wide banking supervision. Meanwhile the German Constitutional Court is debating whether Germany can stand behind the ECB's existing commitment to 'do whatever it takes' to save the euro. Unable to come to grips with today's problems, Europe has turned to fantasising about a far-off world.

In America, the Dodd-Frank Act is three years old but is still awaiting implementation. Wall St failed to prevent its passing, but is now making the implementation so complex that lawyers will grow rich inventing loopholes. In an ineffectual bleat of protest against this, the Brown-Vitter Bill now in Congress would take the capital requirement to 15%, but this has no chance of passing.

The UK, Europe and America each have proposals to separate conventional retail banking from the risky business of investment banking. Thus the core parts of finance would once again be boring mundane tasks performed, we might hope, by boring mundane people. This would be a more effective way of changing the culture than warning high-rolling gambler/managers they might lose their bonuses.

Even if these various forms of ring-fencing can pass their legislative and operational hurdles, this leaves an intractable problem: the residual outside the ring-fence (the shadow banking sector) may be so big, complex and connected that it, too, cannot be allowed to fail.

We need a new starting point, less influenced by an ivory-tower notion of the optimality of free markets. With deregulation, we mistakenly believed that financial markets could be enhanced by competition. Innovation such as derivatives would 'complete' markets. More trading would result in better 'price discovery' and deeper liquidity which would still be there, even when the market turned down. Discipline would be provided by a combination of the market and self-interest: all that was needed was transparency.

We forgot the fallibility of human nature, the pro-cyclicality of the financial sector and the self-exacerbating dynamic of markets.

We have now reached a dead end where further layers of regulation, modulated by powerful vested interests, will not restrain the over-expansion of the financial sector.

The better starting point is to see the core products of finance – safe deposits, efficient payments system, effective intermediation through simple loans and transparent instruments – as a relatively small industry, closely and intrusively supervised by powerful, confident regulators with a gimlet eye for asset quality. Outside this core anti-fragile sector, the 'buyer beware' sector still needs regulation to limit its excesses. When the vested interests complain that this will lead to a substantial shrinkage in their business, the policy-makers should nod and say 'exactly as it should be'.

Photo by Flickr user Oolong.