In an economic crisis, stock markets fall, interest rates inflate, and the economy collapses. Worst of all, most of us don’t even see it coming. We know that crises happen from time to time, and we know that sooner or later another one will come. This is what our common sense tells us. But if we know that a crisis will come, why are most of us are in shock when it arrives? Because we never know exactly when the financial skies will fall.

We compare economic crisis with natural disasters like forest fires, hurricanes, or an earthquake. Most of us don’t know when these will hit, but animals usually feel subtle environmental changes in advance. Wild animals are intuitive enough to recognize the signs of an earthquake before our sophisticated electronic machines register the first tremors. Sometimes, however, our animal instincts awaken and we start to feel something we cannot explain. We feel the fine matter of the universe and other human beings in ways we cannot describe. We call this a “sixth sense.” Our sixth sense often guides us when we don’t have enough information to make a rational or emotional decision. The sixth sense, developed as intuition, helps us make the right decisions.

We often ignore our sixth sense, and suppress it with overthinking and our doubts. We definitely don’t trust the sixth senses of others or the recommendations they make based on their intuition. Our mind likes facts. We want information written in bold, and tables with numbers, charts and scores. But what if we could record sixth-sense signals as solid data? What if we could use it predict the next crisis? This may sound like science fiction or a baseless speculation, but it is not. Not at all. We have some facts to share here.

Over a course of years, a web service called collected the stock preferences of non-professional traders. FAStocks was a free online service for stock portfolio optimization. It offered a method for building a balanced portfolio with a maximum expected rate of return and minimum price volatility. The service used up-to-date stock prices of 1,200 companies listed on the Fortune 2000 and S&P 500 exchanges. FAStocks was fully automated, charged with Harry Markowitz modern portfolio theory and used a genetic algorithm for portfolio optimization. Users entered the website, selected up to 10 companies they were interested in from the list of 1,200 companies and clicked a button to receive a balanced portfolio, or, literally, the optimal balancing of the assets they chose. After the optimization, the user received a free financial report while the website recorded the portfolio and the creation date. FAStocks recorded no personal user data and didn’t require users to register. This was an anonymous and fair deal between the user and the website. In 2014, the Scandinavian Institute of Business Analytics (SCANBA) acquired FAStocks and analyzed its data.

What SCANBA discovered was intriguing. Over three years, users optimized about 20,000 stock portfolios. The company users chose most frequently for their portfolio was Apple. This was no a surprise. After all, is, well, Apple. The most striking result was the users’ timing and frequency of their selection of Apple for their portfolios. We filtered out fluctuations in the number of website visits and called this frequency a ‘SCANBA index’. Formally, the SCANBA index of a company for a certain period is the percentage of mentions of the company during that period. For example, if on a radio show the hosts mention 12 companies and mentions a particular company three times, the SCANBA index of that company during that radio show for that date would be 3 / 12 = 0.25 or 25%. If, in certain months, a website had a thousand visitors who created stock portfolios and two hundred of them selected a particular company, then for that month the SCANBA index of that company on that website would be 200 / 1000 = 0.2 or 20%. In general, the SCANBA index reflects people’s interest in a certain object. This is not a citation index at all; it is vague and subtle, implicit. People may not even recognize that they are expressing their interest. The SCANBA index is a numeric record of the interest shown by the crowd in an object.

Look at the chart. The bold line is Apple’s stock price (NASDAQ:AAPL day close), and the dotted line is Apple’s SCANBA index calculated using FAStocks data for every month between September 2011 and September 2014. The curves behave similarly during the first two-year period, when Apple stock experienced two major rises and falls. Some may even notice similarities down the line, although they are less pronounced. The surprising fact is the similarity of the curves – between their major behavioral features, like peaks and valleys.



The second surprise is the one- to two-month lag between the curves. The real mind-blowing observation is that the SCANBA index changes ahead of the stock price! In other words, if you know the SCANBA index, or have data to calculate it, then you can predict changes in a stock price one or two months in advance. You can know whether the price change you see is a correction or a real turn. The SCANBA index indicates that interest of the public in a stock acts as a sort of sixth sense for traders.

This discovery is hard to believe – until you see the data. When we saw this chart for the first time, our rational minds resisted adopting the data, although our reason for acquiring FAStocks was to confirm this counter-intuitive theory.  The SCANBA index may feel like magic – even to me, its author – and it still retains a shade of mystery. However, traditional forecasting is not magic. It is a real thing called predictive analytics.