Speaking at the British Bankers' Association's annual meeting in London on June 30, Turner said capital and liquidity should act as "shock absorbers" in future periods of market stress.
There should be higher capital ratios for banks, higher-quality capital, and more loss-absorbing equity, he said. Also, capital should be made more countercyclical by using buffers built up in good times to be drawn down in bad times.
He continued: "What is striking, looking at the debates over the Basel II capital regime, is how much intellectual energy was devoted to producing a system which was more sophisticated in its definition of relative capital requirements against different classes of assets, and how little to the fundamental issue - what is the overall optimal level of bank capital ratios?"
The area where the existing regime was "most clearly and dramatically inadequate" was in relation to the capital held against trading book exposures, he noted.
However, while higher capital levels are already required of UK banks, which have a minimum core Tier I capital ratio of 4%, regulators need to be "very careful of the transition path to any higher capital ratio regime", he cautioned.
Changes in the capital regime must also be matched by major changes in the approach to liquidity, he continued. Liquidity is another area where regulators across the world "took their eye off the ball". Supervisors were slow in reacting to changes in the nature and scale of liquidity risks, and changed patterns of maturity transformation.
Turner recommended banks move away from a dependence on liquidity through marketability and from a reliance on wholesale interbank funding, including funding from abroad. Larger buffers of "undoubtedly" liquid assets, primarily government bonds, as well as tighter controls of overall maturity mismatches, will be required going forward.
Another area where regulators need to focus their attention is the systemic importance of banks, and whether a firm can be 'too big to fail' - an issue not explicitly discussed in the Turner Review of global banking regulation, published in March this year.
One solution could be resolution arrangements, whereby haircuts could be imposed on some of the non-insured creditors of large failed banks. "There are certainly attractions" in warning the non-insured creditors there is no certainty of a bail-out in failure, Turner said. However, if bank failure is more the result of systemic fragility than excessive risk-taking, the regulator would act to preserve the system.
The more drastic option of breaking up larger banks should not be excluded, but "we might have to make banks very much smaller before we could really accept failure with equanimity", he said. However, he conceded that, on a practical level, it is unclear that absolute limits on bank size can be agreed at a European or global level.
Turner's preferred solution for managing systemic risks is higher capital requirements, and in particular equity capital, because "equity capital can be allowed to suffer loss without systemic danger". The FSA would even consider a capital surcharge for systemically important banks, as has been proposed by the Basel Committee on Banking Supervision.
He also addressed banks that were "too big and too cross-border to save", such as the Icelandic banks, which were large relative to their home countries and operated internationally, making crisis co-ordination more complicated and leaving the home country to bear the brunt of the cost in the event of default.
Measures restricting the ratio of bank liabilities to the gross domestic product of the home country would be of some use, but they would "gloss over an extremely important distinction between the liabilities of a bank's overseas branches unmatched by ringfenced assets and the position of those banks whose foreign operations are fully capitalised," he said. "The more important response to the problem lies again I suspect in capital and liquidity requirements, but in this case looking not just at the group level but at the capital and liquidity residing in national entity balance sheets," he concluded.
Changes to regulatory rules should be supplemented with improvements in identifying system-wide risks. However, one concern would be whether macroprudential tools, and in particular countercyclical capital adequacy, "should be varied in a discretionary fashion or hardwired, through, for instance, a Spanish dynamic provisioning type approach".