Module 3 Lesson 4
Module 3 Lesson 4
2
The art of calibration
Although we usually follow the same steps, the calibration process can take different forms:
▶ Which market variable are we calibrating to? Which error function?
- Mean Squared Error (MSE) of option market prices (call/put?):
N
1 X ∗ 2
min Cn − CnModel (α)
α N n=1
√
dSt = (rt − rJ )St dt + νt St dZt1 + Jt St dNt
√
dνt = κν (θν − νt )dt + σν νt dZt2
√
drt = κr (θr − rt )dt + σr rt dZt3
So, much like with Bates (1996) calibration, we need to take a sequential path:
5
Calibration results BCC (1997)
6
Calibration results BCC (1997)
7
Summary of Lesson 4
In Lesson 4, we have learned about:
⇒ TO DO NEXT: In the notebook associated with this lesson, we will guide you through the complete
calibration process of the BCC (1997) model.
⇒ In the next module, we turn to Markov processes and chains, which are integral for the development of
reinforcement learning, a field that has been gaining importance in quantitative finance given its broad
applications.