One of the growing problems I have noticed over the years with trading is the increasing amount of data that is available to traders. In the good old days we had the reverse problem, getting any information was difficult. Even finding out the closing prices in the US could be tortuous. The internet has completely reversed this problem with not only every world market available to traders at an instant but the powerful analytics of tools such as Bloomberg terminals have begun to filter down to the masses. This presents a unique problem because increasing the amount of information does not increase the fidelity of decisions being made. As an example I got bounced the graph below this morning for comment.
Source – Dave Wilson
The graph shows that US pension funds are increasingly beginning to rely upon cash as a means by which to fund their obligations. The interpretation that was put to me was that this meant that Pension funds were becoming disinterested in equities as a growth or funding mechanism for their obligations. The extension of this was that there was less and less cash being dropped into the market and the market would respond by falling. This is a valid narrative but a narrative is only a story used to explain historical data. An alternative narrative might be simply that it reflects the ageing population of fund members and as the demographics of the fund shift so to does the strategy that funds that. However, that too is merely a story.
The point that troubles me is that that this information conveys in some a strategic advantage when in fact it is simply data and this is the issue that bedevils traders – sorting data from information. However, there is a secondary issue and that is whether the information you are receiving in some way adds fidelity to your trading decision. It doesn’t matter whether the information is of a macro nature or is perhaps more tactical in nature, the overwhelming question is whether it adds to your decision making. Decision making is a bounded utility, it is bounded by time, the quality of the information you receive and your cognitive ability. These can never be infinite but implicit within this is that decision making is a somewhat quick and dirty process that has to managed and the information upon which you are making a decision has to be managed. This is why models and systems tend to perform better than people do. More information does not make for a better decision. As a real world example consider the case of the heart attack model developed by US Navy cardiologist Lee Goldman. Goldman built a simple visual model that enabled particularly submarine medics to decide whether a heart attack was taking place. As you can imagine getting someone quickly off a submarine is not an easy task. This model might have sat in relative obscurity if it had not been taken up by Cook County Hospital in Chicago. Previously diagnosing a heart attack with the hospital had relied upon a battery of tests (data) combined with the opinion of whichever cardiologist was on duty. Short of funds and looking at ways to streamline treatment options the hospital implemented Goldman’s simple model. There was naturally concern that the outcome for patients might be compromised by using such a simple linear model with only four metrics. Intriguingly health outcomes for patients didn’t change. More data or noise did not mean better outcomes.
So before you go plastering your chart with indicators each with supposedly magical ability or you fork out for a Bloomberg terminal ask whether what you are adding actually adds to the fidelity of your decision making or whether it is fulfilling another need.