This month the banking world has seen the emergence of a new set of rules, loosely called "Basel III". At the same time, we've marked two years since the global financial crisis peaked in September 2008, with the collapse of Lehman Brothers. Over the past two weeks, Basel III has frequently been described as aiming to prevent future financial collapses. It won't. In fact, quite the opposite: it specifically acknowledges that there will be more financial collapses in the future. You might say that Basel III is like road regulation, not like building regulation. Building regulation helps ensure buildings almost never collapse. In most of Australia, you have a negligible chance of being hurt by a falling building. Road rules are different: many of us will suffer some sort of road accident during our lifetimes. What's more, we know and accept it. The costs of stopping all road injuries and deaths are more than we're willing to bear. We've made an uncomfortable trade-off: we accept some deaths in return for cheaper cars, cheaper roads and faster travel times, and we try to limit the inevitable damage. Banking regulation, Basel III acknowledges, is like our road rules. We're going to accept the occasional accident in exchange for being wealthier overall. In the two years since the collapse of Lehman Brothers, this has arguably been the single biggest change in thinking about bank oversight. Officials around the world now accept far more than before that finance is an inherently dangerous business - that it's more like a road than a building. In Basel III, regulators acknowledged the near-certainty of future accidents. Perhaps the strongest signal of the change in thinking came in January's speech by US Federal Reserve chair Ben Bernanke; and in a supporting Fed staff working paper Bernanke and his staff painted a picture of a financial system that gets carried away with new ideas, in this case lending against residential mortgages. (Economists have a name for this: it's Hyman Minsky's "Financial Instability Hypothesis".) In this view, instability is inherent in the system. So with Basel III we have not just a tightening of the old Basel II rules but also the financial equivalent of police radar, speed bumps and traffic blitzes: •     Rules that react when things look like getting out of hand - in this case the "countercyclical buffer" designed to push against a once-in-20-years credit boom. •     Rules that admit regulatory imperfection - the special rules for "systemically important banks" acknowledge the problem of banks that are "too big to fail", such as UBS, Citibank and the Royal Bank of Scotland. •     Rules designed as backstops - the Basel III leverage ratio is designed to make harder the gaming of Basel II rules undertaken by the likes of UBS. The new Basel rules, even if properly implemented, will not guarantee the system is "safe". Quite the opposite: they start from the implicit