Forecasting

 The late Sam Goldwyn was a major film producer in the golden age of Hollywood studios. He was famous for making very funny comments about a variety of topics. One of the most frequently quoted is: “Forecasts are difficult to make—particularly about the future.”

As witty as that comment seems to be, the great producer put his finger on a very serious topic. 

The standard definition that I have seen for the last 50 years goes something such as this: “Forecasting is the systematic process or creative estimates about future conditions, relying on historical and current data.”

Short term forecasts rely on historical patterns and current data and are used to forecast the stock market over the next week, outcomes of imminent elections and sporting events. It amazes me that media outlets such as CNBC and Bloomberg Business give airtime to “analysts” who provide remarkably short-term estimates covering only a week or a month. They are generally wrong. I am always willing to hear forecasts of a financial heavyweight looking out over a few years, but very near-term estimates are worth very little.

Long term forecasting is a VERY tricky business. Some look at past trends and do direct mathematical extrapolation which is generally not reliable. The more sophisticated players add qualitative issues to the mix and often provide several multiple futures. Others use the Delphi method which solicits opinions from a panel of experts and often publishes a consensus of the opinions and promises anonymity to participants. Such averages have a rather weak batting average. 

Whenever I have looked at long term projections I have noticed they do not consider external shocks which forecasting models cannot see. Such Black Swans, popularized by Nassim Taleb in several books, upset the applecart of almost all projections.

Today, in finance, many people have put their savings in index funds. They do not try to pick individual winners among securities. The whole market is purchased. Your never do better or worse than the total market performance.  And, fees are extremely low which over a lifetime make a huge difference to long term performance.

Another huge factor in my view that limits long term forecasting is that the rate of change in today’s world is much faster than in the 20th century.  Look the media business where many MR readers have spent their careers. It was clear that advertiser supported cable channels were going to grow but streaming video caught many of us alleged futurists flat footed. I remember spending quarter after quarter breaking down the balance sheets and income statements of Netflix and wondering how and when they were going to become very profitable. Google’s You Tube has become something of a stealth bomber taking up more usage than futurists could have seen a decade ago.

With the emergence of Artificial Intelligence (AI), change will be more sweeping than we have seen in the past. Yet my feeling is that the growth may not be as quick as some people are now forecasting. It will, however, come and it will be huge.

Okay, I just made a forecast using the Delphi method. Check back in 5 years and see if I was correct!

I started this post with a Sam Goldwyn quote about forecasting. Here is a great one to end with— “Ninety percent of the art of living consists of getting along with people you cannot stand.”

If you would like to contact Don Cole directly, you may reach him at www.mediarealism.blogspot.com or leave a message on the blog.