Modelling The Future


I have been building computer models of financial situations and business plans for close to 40 years.  There is just one thing I have learned in those 40 years.  No model is correct.

Which leads to the question, “Why do some people behave as if modeled answers are facts?”  There is another whole crowd who behave as if nothing is fact and apply some discounts to the numbers.  Neither are right.

Models produce ideas about how things interact and what might happen, “all things being equal,” which they never are and seldom specified.

Systems are complex, not all the variables are known with precision and not all variables are weighted correctly.  In financial modelling there is another factor.  There are other people involved and their decisions will influence the outcomes.  You cannot know how they will act in a given future circumstance once they see how you have acted.  You cannot even know how you will react.

Any system that includes other people, “intelligent agents” is the correct term, is not capable of producing answers about the future.  Models rely on global generalities for their variables and their rules, but people respond locally.  People who lose their job behave differently than those who get promoted.  Maybe it all averages out or maybe it does not.  Maybe it averages out a year or two later.

Once upon a time, a 40 year personal financial plan was probably useful.  In North America things fluctuated but, they did not change all that much.  Not so true now.

In finance it is important to note that the global systems include many important variables that are not considered in most models.  You can no longer be certain that the US Dollar will remain a dominant reserve currency.  You cannot know how the industrialization of China will play out.  You cannot know how Islamic ideas will end.  What of Europe?  Will Americans go back to their making money heritage or continue the getting money game of service and packaging?  Who knows?  Who knows if it will matter?

Modelling complex systems is to gain insight not to provide answers.  For some local questions, models can show where the edges of the problem are.  Should I pay down mortgage or invest in an RRSP?  Is an RRSP better than a Tax Free Savings Account?  Should volatile returns influence my decision to participate in a leveraged investment?  In any of these questions, is there a rate corridor that works and others that do not?

Modelling allows you to think about questions better, but it will not provide answers.

Any model must be updated as new information and new relationships become evident.  No plan is complete without a robust review and revise process. Be sure to consider how you will collect information and how you will assess its meaning.

It is like driving.  You cannot know everything that will happen before you leave.  It is those little steering wheel adjustments that keep you moving towards your goal

Subscribe to the daily MoneyFYI article at, Follow on Twitter @DonShaughnessy or Contact:  

Don Shaughnessy is a retired partner in an international public accounting firm and is now with The Protectors Group, a large personal insurance, employee benefits and investment agency in Peterborough Ontario.


One Comment on “Modelling The Future

  1. Great post, thank you. I like right questions — questions that open possibilities in my mind and heart. Models are not going to work in all contexts, and I don’t always ask the best questions — but I keep open to shifting my models and my questions. However, like systems, they show how components can work together — mechanically or organically or perhaps metaphysically. Models help me think about how in certain contexts pieces of a particular puzzle fit together. Models have always inspired me because they helped me look at the interplay of variables. Key components of a model often make that particular system unique as there are often other variables that may be common in many models. What engages our mind/body/spirit to get inspired to create, innovate or solve a particular problem? I like simple systems — and subsystems because I can review the concepts within that model and learn it — top of mind variables to have something that works and I get it.

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )


Connecting to %s

This site uses Akismet to reduce spam. Learn how your comment data is processed.

%d bloggers like this: