One of the most awkward and embarrassing things that can happen to an analyst is to get a call spectacularly wrong. The cynical view on this is that analysts have an incentive to make predictions that are similar to those of other analysts, so that if they are wrong, at least they aren’t the only wrong one. On the other hand, if your calls are too pedestrian, it will be hard for you to find a market interested in reading your research, and you will ultimately match or, if really unlucky, underperform the index. So there are some counterbalancing motives to be contrarian.

These incentives are something to keep in mind when considering what happened to commodity experts in 2014. I clearly remember a popular BBC news program in which the journalists interviewed an oil and energy analyst at a very large investment bank about her expectations for oil. This analyst asserted with great confidence that oil would likely increase in price at a faster rate throughout the year.

The interview was in April 2014.

Of course, if you have followed energy at all, you know that the second half of June is when oil prices collapsed. Oil went from over $100 per barrel to less than $50, and would stay below $50 more or less until the start of 2018. The analyst’s wrong call was so bad that I remember it many years later.

If nothing else, this is a warning to analysts.

It is also something to consider about the health of the financial profession. Analysts tend to be lumped together by people who do not work in finance, and the assumption is if one analyst is wrong, they all are, and enough famously wrong calls in the public media will make the public increasingly hostile to Wall Street—a trend that is already in full force with the intense and growing popularity of passive index investing.

However, this story is also a reminder that there is an opportunity to be made if you can discover a data-driven methodology to predict asset or commodity price changes with greater accuracy than others, whether that data is public or private. This, after all, is how Alan Greenspan began his career; by looking at obscure data on commodity and goods flows, he was able to accurately predict future demand, and thus price, of several assets.

There is a lot more data out now, which makes the analyst’s job easier and harder. It is easier because there’s more data available, of course, but it is also harder because there are more people analyzing the data (thus difficult to have an edge above the market) and because much of this data is spurious or misleading. Analysts must be more selective and, well, analytical when deciding which data sets to incorporate into their models.

Now, oil is rising swiftly, having breached $70 for the first time since the middle of 2014. The economic implications of this are difficult to summarize, and that is why analysts should be focusing on it right now. Likewise, commodity analysts should take this opportunity to try to figure out why this is happening now, so that they can find the data that is truly predictive and ignore the embarrassing error that the analyst made on BBC those years ago.