Strategy

Finance, strategy and the three-body problem

Two forces you can solve exactly. Three and the maths gives up on you. Most finance decisions are the second kind.


In physics there is a problem called the three-body problem. Predict the motion of two bodies pulling on each other and you can solve it exactly, on paper. Add a third and the exact solution disappears. The system becomes chaotic and enormously sensitive to where it started. You are left with simulations and numerical methods rather than an answer.

I keep coming back to it because it describes the work better than most things written about the work.

Why plans behave like this

A company is balancing internal resources, external market conditions and competitive dynamics at the same time, and all three move. The relationships between them are not simple, and a small change in one shows up somewhere you were not looking. A new regulation, an economic turn, a change in leadership, a shift in what the culture will tolerate. Any of these can move the trajectory of a business.

The same is true inside the numbers. Interest rates, exchange rates and investment returns are all pulled by a web of factors, most of which are outside anyone's control.

Slight changes in one area ripple through the others in ways that are difficult to predict and easy to explain afterwards.

What that changes about the work

Physicists did not respond to the three-body problem by pretending it was a two-body problem. They built tools that work under chaos. That is the part worth copying.

In practice it means a financial model that carries a range of scenarios rather than one set of assumptions. It means a strategic plan with contingencies attached to conditions, so that a pivot is a decision already thought about rather than a panic. Scenario planning, better data, and now AI, all help. None of them make the system simple.

Prediction is still possible

The three-body problem has no exact solution, and physicists still make useful predictions about how the system behaves over time. That is the encouraging half.

We can rarely name the outcome. We can usually describe the range of plausible ones, and we can improve that description as new information arrives. That requires understanding the forces underneath, and a willingness to keep revising. Risk assessments and projections updated as data comes in. A competitive picture that is monitored rather than filed. The point is to stay responsive instead of committed to a course set with old information.

I have never found a way to make this feel comfortable. The best I have managed is to stop expecting it to.


Comments

Corrections, disagreements and better examples are all welcome. Comments are read and approved before they appear, so there is a delay.