A readable and well-argued article in the Financial Times last week by Wolfgang Münchau explores the risk that we all face in the media and metal-buying communities in allowing our bias to influence our interpretation of data.
Even as past and current practitioners, we at MetalMiner have to be constantly vigilant to the possibility that our own expectation of market trends is not influencing our interpretation of data. Münchau uses the recent announcement of an EU-Japan trade agreement — an announcement conveniently released on the eve of the G20 summit — to illustrate his point.
The other critical section of the EU-Japan trade story, he argues, is a good example of what psychologists refer to as confirmation bias, or the tendency to filter out everything that is not consistent with our beliefs. We believe in free trade, hence we want an EU and Japan deal to be true, therefore we accept the announcement as fact even though most of the detail has yet to be sorted out and the timing of the announcement is a shameful attempt at media manipulation by the G20.
By way of illustration, Münchau uses the debate about the future of the euro to show how opposing sides have fervently talked up developments that support their belief.
On the one hand Euroskeptics who would relish nothing more than the death of the euro have consistently underestimated the political cohesion of the Eurozone, Münchau suggests. Meanwhile, euro enthusiasts have shown a consistent complacency about failures of the banking union and instability caused by the rise in macro-economic imbalances between regions to the sustainability of their single-currency totem.
Biases Forward and Back
Münchau quotes Daniel Kahneman, the psychologist and author of “Thinking, Fast and Slow,” who identifies two other forms of bias we are all at risk of committing: hindsight bias and outcome bias.
Hindsight bias is looking at issues, like the financial crisis, and interpreting them with hindsight as if we knew they were bound to happen, when in reality often only a minority had the prescience to see it coming and the vast majority were happy to follow the herd.
Outcome bias, meanwhile, is what happens when we judge decisions to be right or wrong by whether they are successful or not. The example he cites is those few hedge fund managers who bet against the sub-prime market in the U.S. and earned billions. Yet if the sub-prime crisis had been delayed by just a year, many of those bets would have expired and/or lost their managers millions if they exited rather than hung on.
Münchau is the first to admit that avoiding our bias is hard, as he says we like to see our prejudices confirmed and our forecasts come true. He likens confirmation bias to a drug – it feels great while you are confirming your beliefs, but somewhat different if you wake up in the morning to find your country has left the EU!
So, does he have a solution?
Sensibly, he does not suggest that simply fact-checking is the answer, as he says we are not lacking in facts or access to facts — the problem is we filter out inconvenient ones.
What we should do, both as media analysts and as metal buyers, is to try to argue the opposite case, try to use those same facts to test the probability that our initial reading could be wrong and, whether we like it or not, an alternative state of affairs could be just as or even more likely.
Here at MetalMiner, we would like to think that that is part of our role with our readership, as we strive every day to make a balanced assessment of the markets in which we all operate.