Scribd
Upload
10 views9 pages

MSE245B Lecture06 2025

The document outlines the agenda and content for MS&E 245B, focusing on discrete-time models of asset dynamics, including binomial lattice, additive, and multiplicative models. It discusses the mathematical foundations and applications of these models, emphasizing the lognormal distribution and its relevance in financial engineering. The lecture aims to provide insights into asset price fluctuations and portfolio value changes over time.

Uploaded by

Jeeho Choe
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
10 views9 pages

MSE245B Lecture06 2025

The document outlines the agenda and content for MS&E 245B, focusing on discrete-time models of asset dynamics, including binomial lattice, additive, and multiplicative models. It discusses the mathematical foundations and applications of these models, emphasizing the lognormal distribution and its relevance in financial engineering. The lecture aims to provide insights into asset price fluctuations and portfolio value changes over time.

Uploaded by

Jeeho Choe
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 9

MS&E 245B

Apaar
Sadhwani

Modeling
Assets

Binomial
Lattice Lecture 6: Discrete-time models
Additive
Model

Multiplicative
MS&E 245B (Spring 2025)
Model
Lognormal
distribution
Apaar Sadhwani
Principal Instructor
Management Science and Engineering
MS&E 245B
Agenda
Apaar
Sadhwani

Modeling
Assets

Binomial
Starting new chapter ("Models of Asset
Lattice
Dynamics") today!
Additive
Model

Multiplicative
Model
Agenda:
1 Modeling asset dynamics
Lognormal
distribution

2 Binomial lattice model

3 Additive model

4 Multiplicative model and lognormal

distribution
5 Random Walk (start)
MS&E 245B
Modeling asset dynamics
Apaar
Sadhwani

Modeling
Assets
• Intersection of financial engineering and economics
• Questions: How do asset prices fluctuate? How do
Binomial
Lattice

Additive
Model
portfolio values change over time?
Multiplicative
Model
• Objective: Mathematical foundations in Chapter 12
• Two approaches
Lognormal
distribution

• Binomial lattice
• Discrete-time
• Simple: suffices for many applications
• Numerical solutions only

• Ito processes
• Continuous-time
• Advanced: realistic and very generally applicable
• Analytical solutions possible
MS&E 245B
Binomial lattice model
Apaar
Sadhwani

Modeling
Assets

Binomial • Discrete-time model


Lattice

Additive • At each time n


Model
• Toss a coin which turns "H" w.p. p and "T" w.p. q = 1 − p
Multiplicative
Model • Asset price Sn changes by a factor u or d depending on
Lognormal
distribution
"H" or "T" resp.
• Sn+1 = uSn if "H" (w.p. p) or Sn+1 = dSn if "T" (w.p. q)
• (draw lattice diagram)

• Here u > 1 and typically 0 < p < 1, d < 1 and d = 1/u


• Get more realistic by dividing [0, T ] in finer intervals
• Simple, suffices for many applications (e.g. pricing)
MS&E 245B
Binomial lattice model
Apaar
Sadhwani

Modeling • Multiplicative nature


Assets
• Price never negative, log(price) fundamental variable
Binomial
Lattice • ν = E [ln(ST /S0)]= expected annual return
Additive • σ2 = var [ln(ST /S0)]= expected annual variance
Model

Multiplicative • Choosing parameters of the model u, d and p given ν


and σ
Model
Lognormal

• Fix ∆t as small period (i.e. divide year in 1/∆t periods)


distribution

p
• p = 12 + 12 σν ∆t
p
• u = 1/d = e σ ∆t
• (sketch proof; as N → ∞, ∆t ↓ 0 ⇒ ∆2 << ∆t )
t

• Example: ν = 15% and σ = 30%


⇒ u = 1.04, d = 0.96, p = 0.53
• Computational aspects: finite set of prices, linear
growth, how to simulate
MS&E 245B
Additive Model
Apaar
Sadhwani

Modeling
Assets Sn+1 = aSn + un
Binomial
Lattice • Noise/shocks un are i.i.d. (e.g. un ∼ N (0, σ2 ))
Additive
Model • Markovian nature of price
Multiplicative
Model
Lognormal
• (prove) Sn = an S0 + an−1 u0 + an−2 u1 + . . . + a0 un−1
distribution
• S (n) normally distributed
• E [Sn ] = an S0 if E [uk ] = 0

• Pros
• Geometric growth when a > 1
• Continuous prices (e.g. assuming Gaussian uk )

• Cons
• Negative price values possible
• Variance independent of current price
MS&E 245B
Multiplicative Model
Apaar
Sadhwani
Sn+1 = un Sn
Modeling
Assets
• Noise/shocks un are i.i.d.
Binomial
Lattice

Additive
• ln Sn+1 = ln Sn + wn where wn = ln un ∼ N (ν, σ2 )
Model • un = e wn > 0 ⇒ prices are positive!
Multiplicative
Model
Lognormal
• Sn = S0 u0 . . . un−1
• (prove) Sn has lognormal distribution
distribution

• E [ln Sn ] = ln S0 + nν
• var [ln Sn ] = nσ2

• (sketch proof) Estimate ν, σ from single trajectory


P € Š P ” € Š —2
• ν̂ = 1
N n ln
Sn
Sn−1 , σ̂2 = 1
N −1 n ln Sn
Sn − 1 − ν̂

• Pros: positive prices, close to real distribution


• Cons: not appropriate for tail event prediction (draw
PDFs)
MS&E 245B
Lognormal distribution
Apaar
Sadhwani

Modeling
Assets

Binomial
Lattice

Additive
Model

Multiplicative U = e W where W ∼ N (µ, β 2 )


Model
Lognormal
distribution
• E [U ] = exp(µ + β 2 /2) – why more than e µ ?
• var [U ] = exp(2µ + β 2 )(e β − 1)
2

• (draw PDF)
MS&E 245B
Summary of Discrete-time models
Apaar
Sadhwani

Modeling 1 Binomial lattice: Sn+1 = un Sn


Assets §
• i.i.d. un =
u , w.p. p
Binomial
Lattice d , w.p. q = 1 − p
Additive
• Discrete-time, discrete-space
Model

Multiplicative 2 Additive model: Sn+1 = aSn + un


Model
Lognormal • i.i.d. un ∼ N (0, σ2 )
• Discrete-time, continuous-space
distribution

3 Multiplicative mode: Sn+1 = un Sn


• i.i.d. un = e wn , wn ∼ N (ν, σ2 )
• Discrete-time, continuous-space

4 Lognormal distribution: U = e W , W ∼ N (µ, β 2 )


• In multiplicative model, Sn ∼Lognormal(ln S0 + nν, nσ2 )
• In binomial lattice, as #periods in [0, T ] = N → ∞, we
D
have ∆t = T /N ↓ 0 and ST −
→Lognormal(ν, σ2 )

You might also like