‘Forecasts may tell you a great deal about the forecaster; they tell you nothing about the future.’
– Warren Buffett

People often ask me – “What’s going to be the price of bitcoin at the end on 2018”. Without any statistical basis, my unremitting retort is – “I don’t know.” While pundits forecast bitcoin to reach $60,000 at the end of this year, their guess is as good as yours. That is true for most forecasts.

At the onset of every year, we always search and consume a multitude of headlines such as “Forecasting the world in 2018,” or “Predictions in 2018 and beyond.” Companies attend various seminars on economic forecasts and technology predictions in the current year and beyond only to understand what lies ahead – they forecast market demand, exchange rates, raw material prices, and others. Then based on these, build new production plants, create new entities, or launch new products.

The problem with these predictions or forecast, if we take them at face value, is that they may be either self-serving to those who purvey them, such as vendors or even cryptocurrency operators, or they may be tinged with the forecaster’s overconfidence and bias. Some of the most spectacularly wrong predictions made by erudite industry leaders are that of IBM Chairman Thomas Watson who said in 1943, “I think there is a world market for maybe five computers,” and former Microsoft CEO Steve Ballmer who in 2007 predicted that, “There’s no chance that the iPhone is going to get any significant market share. No chance.”

Numerous studies reveal the failure of the forecasts of even so-called experts. As Wharton professor Philip Tetlock, co-author of “Superforecasting: The Art and Science of Prediction,” showed in a landmark 2005 study, even experts’ predictions are only slightly better than chance.

That’s why organizations, together with their executives and planning departments, need to be critical of forecasts and predictions, and instead, build competence in forecasting themselves with accuracy.

In general, forecasts are based on quantitative analysis, also called objective analysis, and quantitative forecasting, also referred to as managerial or judgemental analysis. The extensiveness of the use of these types of forecasts depend on two variables – time horizon (short to long term) and data availability (low to high).

Quantitative or objective forecasts operate mainly under short to mid time horizon with sufficiently available data, primarily used for operations. This is where the use of big data analytics (predictive and prescriptive) and artificial intelligence (AI) become extremely useful.

Predictive analytics can identify future probabilities and trends by providing information about what might happen in the future, such as how the business process outsourcing companies in the Philippines use predictive analytics to understand causes of employee absenteeism and low productivity, and predict who and how many will eventually leave the company.

Prescriptive analytics, coupled with AI, is dedicated to finding the best course of action, given the certain parameters, and suggest decision options to best take advantage of a future opportunity or mitigate a future risk, such as the ones used by the oil and gas industry to optimize operations on where to explore new oil sites to predict performing and non-performing oil wells.

Managerial or judgemental forecast, on the other hand, is employed under mid- to long-time horizon with little information, such as in strategic planning and policy-making. This is where much of human error comes into play such as cognitive biases, desire to influence others’ thinking, lack of domain knowledge, concerns about reputation, and so on, resulting in forecasts that are no better than anyone’s guess.

To improve judgemental forecast accuracy, Professor Tetlock, in his landmark well-researched book, prescribes training among forecasters to be aware of psychological factors that lead to biased probability estimates, such as training exercises to quiz participants and uncover how well they know what they don’t know. Another is to assemble a team of forecasters “who are cautious, humble, open-minded, analytical—and good with numbers. In assembling teams, companies should look for natural forecasters who show an alertness to bias, a knack for sound reasoning, and a respect for data.”

Combining the power of analytics and AI with sound judgemental analysis can greatly improve forecast accuracy. But companies will seize this advantage only if business leaders champion the effort, by recognizing the value it brings and accepting “what they don’t know.”

The opinions expressed here are the views of the writer and do not necessarily reflect the views and opinions of FINEX. The author may be emailed at The author is president of The Engage Philippines, a digital customer engagement solutions company, and co-founder of Caucus Inc, a data privacy consulting firm. He teaches strategic management in the MBA Program of the De La Salle University. He is also an adjunct faculty of the Asian Institute of Management.