Nuclear power is scary, because of the fear of radiation (radiation itself doesn’t seem to be as bad for you as you might think) escaping and contaminating all around. But is it as scary as banking? There was an absolutely fascinating piece by Tim Harford in the Financial Times last weekend. It was called “What a nuclear reactor can teach us about the economy“, and it draws parallels between the way engineers build safety systems for nuclear reactors (broadly speaking, by applying science and learning from mistakes) and the way that regulators build safety systems for the banking system (broadly speaking, by making things up and not learning from mistakes). The key observation is that the banking system is complex:
It might seem obvious that the way to make a complex system safer is to install some safety measures. Engineers have long known that life is not so simple.[From What a nuclear reactor can teach us about the economy]
What Tim is saying is that financial products such as Collateralised Debt Obligations (CDOs) and Credit Default Swaps (CDSs) appeared as safety systems (for spreading risk) and then, just like the coolant filter that got dislodged and jammed the coolant flow thus causing a partial meltdown of the Fermi reactor in Detroit, they blew up the system they were supposed to stabilise.
So what can the financial sector learn from the nuclear reactor sector, given this analogy? Well, using the example of Three Mile Island, Tim explains that one of the key reasons that the reactor came close to meltdown was that the operators couldn’t understand all of the dials, lights, warnings and other signals. As a consequence
since Three Mile Island, much attention has been lavished on the problem of telling the operators what they need to know in a format they can understand.
When the financial system started to melt down, regulators were faced with the same problem: given all of the warning lights flashing, given all of the alarms sounding, what was actually going on?
Andrew Haldane, director for financial stability at the Bank of England… argues that the same technologies now used to check the health of an electricity grid could be applied to a financial net- work map, highlighting critical connections, over-stressed nodes and unexpected interactions.
This analogy is imperfect in a couple of ways, of course, because banks can create money from nothing whereas electricity costs money to create and because electricity substations don’t lie to the national grid in order to get a bigger bonus, but you can see his point.
There’s one aspect of this that Tim didn’t explain though. Engineers don’t forget things, but financiers do. Once engineers have learned, for example, how not to build a bridge, then they stop building bad bridges. But bankers don’t work that way. They would stop building bridges that way for a short time, and then simply go back round and starting building collapsing bridges again a few years later.
What does this have to do with payment systems? I’d like to highlight two points: complexity and decoupling. We need to beware of complexity, to treat it as an enemy (the current case study of EMV illustrates this perfectly), and we need to decouple so that parts of complex systems can fail without bringing down to whole of the system. It seems to me that this prescription provides a pretty clear manifesto for payments: separate the payments systems from the banking system and have a lots of simple payment systems instead of a small number of complicated ones.