::: nBlog :::
The methodologies of EU banking stress tests left me somewhat puzzled, as it seems that the worst case scenarios were carefully engineered so that most banks would actually pass. JP Morgan and MarketWatch offer a sobering view.
My understanding of stress tests stems from the military, where we used to perform them often in order to keep up with technological and knowledge development. Most Finnish conscripts remember yellow and blue teams, which are at imaginary war during yearly excercises. Leadership of both are usually given quite some freedom of strategy and means to defeat the other side.
Every so often the winning side invents something totally unexpected, resulting a quick victory. It is this ‘unknown unknown’ (according to Donald Rumsfeld), which in my opinion should form the basis of any realistic stress test, whether in energy security, telecommunications, military – or banking.
Most utilities in the developed countries are actually quite good when facing unexpected events – even Tokyo Electric Power Company (Tepco) was quickly able to restore electricity supply to most customers even though one of their primary nuclear plants was nearly destroyed. Yes, it was a major disaster and we’ll be cleaning up the radiation spill for a long time – But Tepco passed a horrible real life stress test – their customers do get electricity now.
This preparedness is something we must retain when we embrace the Smart Grid. Traditional IT is not used to stress tests or high resilience; it has always been easier to start a project to upgrade to a new version and talk about the benefits ‘when the new system is up’.
When the Smart Grid is finally implemented for entire societies, it must inherit and improve upon the fault tolerance of existing power networks. With highly distributed generation and crowd sourced demand response, data management becomes even more important than securing enough generating capacity. That is why we must expect a new level of reliability – and stress tests – from all of our systems.