Black Swans vs. Crystal Balls: Considering the Unknown Unknowns

When it comes to forecasting, FP&A professionals have the unenviable job of trying to predict the future — as well as being held accountable for failing to predict it. As we've discussed in previous blogs, simply monitoring trends and reviewing historical data can still leave you vulnerable to adverse events, especially those that were not predicted because they lack historical precedent. It requires imagination and the right corporate culture to attempt to mitigate the impact of events that cannot be predicted simply by looking into the past.

Yet some events are, by their very nature, seemingly unimaginable. They appear to come out of nowhere and can have a devastating impact. No amount of historical data can predict them since they are outliers, events that have not previously occurred. This type of event is sometimes called a "Black Swan" — referring to huge, highly-improbable events we usually only feebly attempt to rationalize after the fact.

"Just because an event is seemingly impossible doesn't mean it's inconceivable."

Knowing the Knowns and the Unknowns

One such event was the terrorist attacks on 9/11. After the fact, Secretary of Defense Donald Rumsfeld spoke to the difficulty of forecasting such an event based on the knowledge that was available at the time:

"There are known knowns - there are things we know we know. We also know there are known unknowns - that is to say we know there are some things we do not know. But there are also unknown unknowns – the ones we don't know we don't know. And if one looks throughout the history of our country and other free countries, it is the latter category that tend to be the difficult ones."

Imagining the Impossible

Rumsfeld's statement cuts to the core of why forecasting for huge, disruptive events can be so difficult. We simply cannot know the things we don't know. To this, our only defense is to turn our imaginations to the "impossible" and recognize that even these are possible. This may be seen by some as a waste of time and resources, but as Aaron Brown in his book "Red-Blooded Risk" puts it:

"It is small harm if you assign a nonzero probability to a scenario that is in fact impossible. You might give up a little profit, but that's survivable. It can be fatal to assign a zero probability to a scenario that is in fact possible."

A Look in the Crystal Ball

As forecasters, our job is to consider the things we haven't yet seen, the things that managers caught up in the day-to-day likely don't think about, and to conceive of the inconceivable. This requires looking into our crystal ball and trying to anticipate futures where our models could be disrupted or rendered obsolete.

For instance, take driverless car technology. While not yet a feature of everyday life, we've already seen prototypes on the road. While it may initially seem that the car manufacturers and their suppliers are the only ones who are directly affected, a variety of industries should be taking note. Driverless cars are only a few steps away from driverless trucks, which could spell a very different future for the trucking industry than what we see now.

In this way, we attempt to account for all the knowns and unknowns: Driverless cars are the known knowns, because we know they exist. Driverless trucks are the known unknowns, a technology that doesn't yet exist but is the natural extension of the known.

But the forecaster's responsibility is to consider what comes next, i.e. the unknown unknowns. This is the art of the crystal ball.

Forecasting Into the Distant Future

In the long run, driverless cars could have a massive impact in many segments of the economy. Public sector revenue models would have to contend with the loss of income from issuing fewer drivers' licenses, as who needs a license if the car is driverless, as well as from fewer traffic tickets as computer-driven cars will be programmed to operate 'by the book'. In turn, the taxi industry – and maybe even Uber – will be impacted as those unable to drive due to illness, handicap, drinking, etc. simply summon their computer-driven vehicles. Fewer police officers will be required to enforce road and highway laws, while fewer insurance adjusters and car repair facilities will be needed as the number of accidents declines.

Similarly, in industries with long-lived assets, one is forced to consider what could be, in addition to what is. Just because the genesis of disruption hasn't occurred, or hasn't been thought of, doesn't mean it won't happen.

Better Forecasts Anticipate Black Swans

An effective forecast incorporates scenarios that might be just a bit 'off the wall', such as modeling a mortgage portfolio with a 20 percent decline in housing prices or a government budget without drivers' license and traffic ticket revenues. In this fashion, even if the exact event or situation is not foreseen, the general circumstance can be modeled, its impact assessed, and mitigation plans developed. Simply put, better forecasts see more and see further, thus 'seeing' black swans before they fly into the picture.