courtrigrad
- 1,236
- 2
Hello all
Let [tex]\delta t[/tex] be a timestep. Then the mean is equaled to [tex]\mu\delta t[/tex] where [tex]\mu[/tex] is a constant. Assuming a nornal distribution, [tex]\frac{S_{i+1}-S_{i}}{S_i} = \mu\delta t[/tex]
[tex]S_{i+1} = S_i(1 + \mu\delta t)[/tex]. Hence after M timesteps we have:
[tex]S_m = S_0(1+\mu\delta t)^M = S_0e^{Mlog(1+\mu\delta t)} \doteq S_{M}=S_{0}e^{[\mu M(\delta t)]}= S_0e^{\mu T}[/tex] How do we get the last part (the approximation)?
Thanks
Let [tex]\delta t[/tex] be a timestep. Then the mean is equaled to [tex]\mu\delta t[/tex] where [tex]\mu[/tex] is a constant. Assuming a nornal distribution, [tex]\frac{S_{i+1}-S_{i}}{S_i} = \mu\delta t[/tex]
[tex]S_{i+1} = S_i(1 + \mu\delta t)[/tex]. Hence after M timesteps we have:
[tex]S_m = S_0(1+\mu\delta t)^M = S_0e^{Mlog(1+\mu\delta t)} \doteq S_{M}=S_{0}e^{[\mu M(\delta t)]}= S_0e^{\mu T}[/tex] How do we get the last part (the approximation)?
Thanks
Last edited: