Last month, I wrote a short piece about the demise of Northern Rock. I wasn't planning to return to the subject but the latest wave of epitaphs, triggered by the 10 year anniversary of the run on the bank that prompted the Government to step in, highlights an important problem that this quote - from Lord McFaul - also flags. The strength of the consensus as late as August 2007 is still unnerving.

"What escapes categorisation can escape detection altogether."

And then, in a different context, I read the above (in an essay by Rebecca Solnit), which struck me as relevant to understanding the lead up to Northern Rock, and the decade since. Both periods seem to suffer from the tendency to over-rely on categorisation, breeding an unhealthy degree of certainty in the frameworks and models used.

The diagnosed causes of the Rock's failure and the financial crisis generally - too little capital and liquidity, weak regulation, unaccountable executives, casino banking, insufficient powers to resolve failing firms etc. - have been neatly categorised over the last decade, with solutions designed for each. And the success of these solutions has in several cases already been declared, though as the attached article makes clear, this is very much contested ground. Either way, several risks remain, not least that the solutions, with the best will in the world, may reflect too neatly the tools to hand and the mindset of those involved.

A standard bearer of this approach is the Senior Manager & Certification Regime (SM&CR), which is now being extended to all regulated firms. Two of its key principles are the concept of "Overall Responsibility", which captures all the firms activities, and that Conduct Rules should apply to everyone apart from ancillary staff, and. Both seek to be all-encompassing, but in doing so they create the temptation to stop looking for new activities, or for elements exist  at the perimeter of, or outside, the defined model. SM&CR therefore, like the other purported solutions to the problems that caused the crisis, should be implemented in the knowledge that its frameworks and requirements will almost certainly not solve completely the problems it targets - in this case around accountability and culture. 

If we are to really learn from Northern Rock, we need to create and preserve a space to appreciate that uncertainty and the unknown are an unavoidable part of the mix. Categorisation is important, but will never be all of the answer, either for firms or regulators.