Risk & Modeling

Degrees of Freedom (DOF)

TL;DR

Degrees of freedom (DOF) is the parameter that controls how fat the tails are in a Student's t-distribution. Lower DOF = fatter tails = more frequent extreme market events. At DOF=5, extreme events occur roughly 4x more often than a normal distribution predicts.

Degrees of freedom (DOF) is the key parameter of the Student's t-distribution that determines tail thickness. It provides a single, interpretable dial for controlling how frequently extreme events occur in a Monte Carlo simulation.

How It Works

The relationship between DOF and distribution behavior:

DOFExcess KurtosisPractical Meaning
4Extremely fat tails — DOF must exceed 4 for finite kurtosis
56.0Close match to empirical equity return data
83.0Moderate fat tails
150.5Mild fat tails
30+~0Nearly indistinguishable from normal distribution

Critical constraints:

  • DOF > 2 required for finite variance
  • DOF > 4 required for finite kurtosis
  • Retirement Lab uses DOF values between 4 and 10

As DOF increases toward infinity, the t-distribution converges exactly to the normal distribution. This makes DOF a smooth control — you can dial from "realistic fat tails" to "textbook bell curve" with a single parameter.

Why It Matters for Retirement Planning

DOF directly affects success rate estimates. Lower DOF generates more extreme returns in the simulation, which — combined with sequence-of-returns risk — typically reduces the percentage of scenarios where the portfolio survives.

The difference is meaningful: the same retirement plan might show a 92% success rate at DOF=30 (near-normal) and an 84% success rate at DOF=5 (fat tails). The latter is a more honest estimate of real-world risk, which is why DOF calibration matters for retirement security.

Frequently Asked Questions

What happens if degrees of freedom is too low?
Very low DOF values (below 3) produce distributions where the variance is infinite or undefined, making the simulation results unstable. DOF must exceed 2 for finite variance. Retirement Lab uses DOF values of 4-10, which produce meaningful fat tails while keeping the distribution mathematically well-behaved.
How do I choose the right degrees of freedom?
Empirical research on equity markets suggests DOF of 4-6 for monthly returns. Retirement Lab defaults to DOF=5. Higher values (8-10) are more conservative and produce milder fat tails. You can adjust this in the advanced settings to see how tail thickness affects your plan's success rate.