The Butterfly Effect is the notion from chaos theory where small changes in initial conditions can result in large differences in the outcome. The term was coined when a mathematician and weather scientist was describing a tornado, and noted that the flapping of a butterfly’s wings a week before might seem to change the path of the storm. In deterministic computer simulations, rounding some of the initial conditions can create a totally different outcome.
The Butterfly Effect is a challenge to any predictive model. In a complex nonlinear system, conditions are changing all the time. It’s why we can’t predict exactly the path of a hurricane. Some models had Hurricane Irma going up Florida’s Atlantic coast and hitting South Carolina, while others had it going into the Gulf of Mexico.
Actual path of Hurricane Irma. Source: Wikipedia
The Butterfly Effect has made its way into popular culture, but we get its lessons wrong. In the 2004 movie of the same name, Aston Kutchner travels back in to time change the circumstances of his life, with dismal results. In an earlier gambling flick, Robert Redford tells Lena Olin that a butterfly can flutter its wings in China and cause a hurricane in the Caribbean.
But that’s backwards: we can’t track the connections. There are too many butterflies in China for us to know which one altered Irma’s track to go over Tampa. The future is indeterminate. The lesson of chaos theory is that tiny, infinitesimal changes in one of a host of innumerable variables can have an outsized effect. We can’t know which one, nor can we predict the results. All we can do is observe the trends and calculate the odds – making allowances for fat tails and black swans.
Douglas R. Tengdin, CFA