K Ramkumar, Executive Director and Head of Operations & Human Resources at ICICI Bank, loves examining the other side of the traditionally accepted views. He examines everything that comes in his way. That has helped him to broaden his perspectives, something that he would have otherwise never done. Even while reading, he debates with the author by writing on the margins. Ram, as he’s popularly known to his colleagues and friends, believes that there is nothing more joyous than having an open debate with an equally passionate and experimental individual. The outcome is not important. What matters is a counter point - the other perspective to every viewpoint. A science graduate and a post graduate in Personnel management and Industrial relations, Ram is an ardent sports fan. He prefers to be in the game as it keeps him engaged with others. He also enjoys making short documentary films.
Early last month, the Washington Post linked to a remarkable mea culpa by Olivier Blanchard, chief economist at the International Monetary Fund (IMF). He admitted their predictions on which way the Greek economy would go had gone awry. [See the IMF paper.]
To tide over the economic crisis Greece is now going through, the European Central Bank and the European Union (EU) mandated the Greek government to cut spending. Econometrics wonks at the IMF got down to work and deployed mathematical models at their disposal to predict what could possibly happen. When the number-crunching was done, their models indicated if the Greeks cut spending by 100 percent, the economy would shrink 50 percent. As things turned out, it shrunk 150 percent. And that is the problem with mathematical models.
We all have this strange delusion that anything stated in numbers that has a formula or an algorithm must be objective and a statement of fact. Any data or model that is not put through the crucible of human judgement is useless, when it comes to making calls about the future.
The big picture can never be gleaned from just analytics. Models create sameness, while human judgement supported by analytics brings in enterprise. Models, people believe, cut down risks. But because of standardisation they actually amplify risk.
What do I mean by that? Think fixed telephone lines. If a mathematician were to build a model on how the business of telephony would evolve, it would perhaps take in all of the variables and come up with a set of numbers that a business can be built on. But could any model have predicted the emergence of cellular phones? For that matter, a hundred years ago, what mathematical model could have predicted air-planes would be the dominant mode of global travel?
Models anchor us to paradigms and inhibit innovation. However, human enterprise has the ability to bet on the success of a new technology. At worst, it puts to risk a single organisation, never the whole system.
Jean Piaget the legendary French philosopher, in his treatise on thinking, argued that analytics helps in comprehending the relationship between cause and effect, thus helping solve problems. Analytics connect the past to the present consequences, much in the same way as the rear-view mirror of a car. He established that conceptualisation on the other hand, helps in figuring out weak lines that connect the present to the future and thus aids prediction. Conceptualisation also amplifies weak signals in the environment. Analytics is akin to the rear-view mirror.
Nobel Laureate and behavioural economist Daniel Kahneman dwells in detail about the futility of trusting the analytical mind to the exclusion of the intuitive mind in his book, Thinking Fast and Slow. He, through experiments, establishes that the analytical mind is useful in assessing risks under standard and repeatable conditions. He proves it is ineffective in assessing both risks and opportunities, when the environment is non-standard or involves assessing human behaviour or is volatile.
It is here that the intuitive judgement trumps over analytical models. He argues eloquently that markets and humans often behave irrationally. Hence over-dependence on rational models can lead us to missed opportunities or ignored threats.
Only in science can data be extrapolated into the future and reasonably valid predictions be made. Even here scientific models struggle to predict volatile events such as weather and earthquakes. This is where scientists who have a “nose” to support their inferences from the models, win over pure model-based ones. Try predicting the course of a sporting event based on any model, and the perils of model-based thinking will hit you hard.
We will call a few wrong but will also get a few correct. But if we call everything on the basis of the need for certainty of results, then we will call everything late and lose competitive advantage. When business leaders get stuck to models, especially those built by others or from the past, they will have nothing different and concrete to offer. Sticking to the past cuts risks or normalises it, but will not help call the future and break out. History tells us that no one learns from others or from their own past.