Budgeting for future performance: Pricing new information is where the trick lies

Some professional experts do prediction of all kinds of things. The most recent one was by an investment bank that predicted India will get 8 medals in Olympics. I would be the happiest person if this happens, but unfortunately this will be the most bizarre forecast.

Planning & budgeting activities start with future forecasts of performance; experts go through past trends and study future potential and factor-in the environment where strategies play out. The plan is then argued by some constituencies before it is cast in stone.

The real world works on large number of variables which interact with each other. It is like a design of experiments which has N number of factors and n number of levels. There could be first level interactions and second level of interactions. This could change with new variables getting introduced as the environment could be very dynamic.

Imagine the price of Oil. Experts can do very elaborate design of experiments to find out which factors are dominant and which are not and do multiple regressions. The entry of one single variable would make all this null and void, like the mild discussion of a supply shortfall which could be only 0.01% of the world’s supply. How much that could impact the price of Oil can hardly be computed accurately. So even with test of significance using any sophisticated statistical tool, we could be very wrong as forecasters of Oil price.

So we have new variables coming in every day. The prediction experts must re-plot and change their bearings every day. It is like plotting the ship’s position every day on a chart when the ship is on a free move.

No wonder you have commentaries from experts on economic future events on a daily basis.

But the prediction markets are far more efficient than this. They take in new information every moment and that gets incorporated in the prediction. Exactly like the stock market, the sum total of all information gets incorporated in the prediction. There are no arbitrage opportunities in prediction markets. They are impervious to manipulation. They have lower statistical errors than the professional forecasters. The economic models behind forecasts can be uncovered by these markets.

Exchange-traded markets are created for the purpose of trading the outcome of events. The market prices can indicate what the crowd thinks the probability of the event is. A prediction market contract trades between 0 and 100%. It is a binary option that will expire at the price of 0 or 100%.

Let me imagine an exchange traded market is trading the outcome of U.S. Presidential poll. There are no arbitrage opportunities. If Hillary wins, Trump loses. The everyday position of the predicted outcome will be changing with new information. When large hordes of people will be participating in such trades it has been shown prediction markets are at least as accurate as other institutions predicting the same events with a similar pool of participants. But more importantly probability of an event is close to its market price but significantly closer than the average single subjective estimate.

In planning and budgeting for the year, we start with an initial position. As new information is gathered the initial position changes. As we progress through the year, the changed position could undergo complete reversal.

With rising volatility, it is currently impossible to create a prediction for the year on an average basis. It is like the average depth of the river, which could be low but you could drown trying to cross, because the actual depth could lie between high extremes.

So what could we learn from prediction markets that could be used for planning and budgeting?

To use a rubric that takes out the obvious biases at the initial part of the journey. This could be done when large participants get involved and there is a healthy debate.

The ability to take in new information and factor in that variable which is dominant is the most important part of the whole puzzle. To price any new information is where the icing lies.

Many companies use the concept of latest estimate. But the latest estimate can only be of significance if the impact of every change can be taken through the entire financial artifice. The linkages and sensitivities to the final outcome is one that would make this possible.

The prediction markets do this through trades; each new information thus gets priced automatically. There no chance of a misprice as the price is dynamic.

Making a rolling plan dynamic is where the risk lies. It has all the perils of getting biased to an outcome or it could well be that choice of factors could be such that the dominant ones are lost in the search for insignificant ones. The test of significance is thus the only respite.

But companies would do better than the Olympics medal forecast by the investment bank, I am sure. As they would not make the mistake of making a static forecast.

The ability to price new information is where the trick is.