When Forecasting Fails


Every few years or so there is a crisis of confidence in the ability to forecast. Yet organisations still dedicate large amounts of time and talent to developing very detailed forecasts of the future that have little chance of coming to frutition.


More dangerous is that these works of fiction are used to make material decisions about which strategies to pursue and where to allocate resources. The usual excuse is to blame a so-called 'Black Swan' event, an unpredictable event that is beyond what is normally expected and has severe consequences, for the obsolescence of detailed plans, budgets and forecasts.


Unfortunately just like in nature where the global poplulation of black swans is estimated at more than 500,000, the occurrence of Black Swan events seems to be increasing. Over the last thirty years, all of the following events have been characterised as excuses for the inability to forecast the future: collapse of the Soviet Union, dot.com bubble, 9/11, Asian Tsunmai, SARS, Enron, Greek debt crisis, Global Financial Crisis, election of Trump, Brexit and, of course, Covid-19. In each instance traditional forecasting methods based upon extrapolating past data to the future proved to be not just inneffective but danagerous to an organisation's health.


Despite mounting evidence to the contrary we seem unable to abandon what is a clearly broken process. Inertia is a clearly a factor, after all we have always done it this way - well at least since Alred P. Sloan commented in 1926 that, 'There is no single thing that has done as much to instil confidence into the minds as has the forecast.' Unfortantely, times have changed, Sloan operated in a time where planning was largely a supply-side problem largely unaffected by global events (at least until World War II).


There is also a huge investment of time, talent and technology vested in forecasting and over the years we have introduced tweaks to the process to try and improve accruracy. Driver-based foecasitng, Monte-Carlo simulations, regression analysis and the like but in total their impact has been minimal. In a fascinating article in the 22nd October issue of the Wall Street Journal, United's Pandemic Strategy: Stop Trying To Predict It, CEO, Scott Kirby commented that, 'We're not going to pretend we know what demand will be.' United abandoned the budget and relied solely on a scenario-based approach that addressed both a rapid return to normal travel patterns but also envisioned a future where a much smaller ailrine would be needed for success. Resources and dependencies were then mapped to each scenario.


Perhaps the crucial step was that actions were set in motion to allow the company to respond with speed in the event a more attractive scenario presented itself. For example, refilling the pilot pipeline and speeding up the security clearance process would allow the airline to respond quickly to a surge in demand. There was a price to such preparedness but it placed the airline in the best position to react to an uncertain and volatile future.


There are five key lessons here for finance professionals:


  1. Black Swan events are more like London buses, you wait a few years and then three come at once!

  2. It is time to abandon traditional trend (or even modelled trend) forecasting. It consumes too much time and resources, rarely provides the right answer, provides a false sense of security (‘its ok, it’s in the budget’) and actively hinders speed and agility. This does mean running a secret parallel process on a series of spreadsheets--just in case!

  3. Risk and uncertainty need to be implicit in all planning processes

  4. Plans need to translate into actions that create the ability to rapidly flex to alternate course of action as needed. For example, securing access to additional capacity, building redundancy into supply chains, ensuring flexibility in labour supply and cost structures.

  5. In short: spend more time preparing than planning.

We, as finance professionals, have a duty to lead our organisations to a more dynamic and realsitic way of planning that recongnises the reality of today's uncertain and volatile world.


Additional Reading


Articles


Books

  • The art of the long view, Peter Schwartz, Currency Doubleday, 1996

  • My years with General Motors, Alfred P. Sloan, Jr, Currency Doubleday, 1963