FRM Part II · FRM Exam Part II · The Vasicek and Gauss+ Models
A risk manager compares two Vasicek calibrations that share the same θ and σ. Calibration A has k = 0.10 and Calibration B has k = 0.80. Which conclusion is correct about the short rate's behavior?
Calibration B has the faster mean reversion and therefore the lower long-horizon variance. Stationary variance equals σ² divided by 2k, so raising k from 0.10 to 0.80 shrinks it from 5σ² to 0.625σ².
- AB's short rate reverts faster, so its long-horizon variance σ²/(2k) is lower than A'sCorrect
- BA's short rate reverts faster, so its long-horizon variance is lower than B's
- CBoth have the same long-horizon variance because σ and θ are equal
- DB's long-horizon variance is higher because faster reversion increases the drift
Explanation
Stationary variance is σ²/(2k), which falls as k rises. A: σ²/0.2 = 5σ²; B: σ²/1.6 = 0.625σ². Higher k means faster reversion and a tighter long-run distribution, so B is lower.
Did you get it right without looking?
One question tells you little. A timed set on The Vasicek and Gauss+ Models shows your real accuracy, how long you take and where you lose marks.
More The Vasicek and Gauss+ Models questions
- A model validation team reviews a one-factor Vasicek implementation used to price zero-coupon bonds in a low-rate environment. Which stateme…
- A desk calibrates a Vasicek model and then re-estimates it with a higher σ, leaving k, the risk-neutral long-run mean rate and the current s…
- In a Vasicek model, k = 0.25, θ = 6.00%, σ = 1.00%, and the current short rate is 4.00%. What is the expected change in the short rate over …
- A desk calibrates a Gauss+ model by fitting model-implied swap rates to the market curve and fitting volatilities to historical rate changes…
- In a Gauss+ framework, the short rate is the sum of two mean-reverting Gaussian factors, x1 with a slow mean-reversion speed and x2 with a f…
- An analyst calibrates the Vasicek mean-reversion speed to k = 0.35 per year using historical short-rate data. What is the half-life of a dev…