Advanced Fixed Income Analytics

© All Rights Reserved

20 views

Advanced Fixed Income Analytics

© All Rights Reserved

- C A S E 28 Inner-City Paint Corporation (Revised)
- Splash Corporation Background
- Hong Kong Trading Techniques
- 4610383664-461 (1)
- Du Gayy Salaman (1992) The Culture of the Customer
- Leadership in the 21st Century
- Taal View Heights Computations
- SWOT Analysis Template
- EQUITY Morning Bell 091211
- Visual Basic Formulas de Erlang
- The Deindustrialization of Many Markets
- Mergers
- Test Paper 2 CO
- Topics
- 07Simplex3B
- Sales and Shipping - Consumption Data for Items (From 20B) - Order Fulfillment - SAP Library
- 8
- Build It Overview
- 01- Dynamics of Business and Economics
- Business Proposal

You are on page 1of 13

Lecture 6

1. Uses of models

2. Assessment criteria

3. Assessment

4. Open questions and new directions

stochastic volatility (Cox-Ingersoll-Ross)

volatility inputs (Heath-Jarrow-Morton)

multiple factors

pricing kernels and risk-neutral probabilities

6-2

1. Use of Models

Valuation of new structures

{ approach 1: look at comparable assets in the market

{ approach 2: models guarantee consistency with other

(\even a bad model can be tuned to get some prices right")

{ what's the sensitivity to a 20 bp rise in 5-7 year spot rates?

{ models can be used to do \partial derivative" calculcations,

but the answer depends on the model

{ you take dierent short cuts depending on the purpose

{ the key is to understand where short cuts hurt you

2. Assessment

Questions for models:

{ does it reproduce current spot rates?

{ does it reproduce current term structure of volatility?

{ are sensitivities and hedge ratios reasonable?

{ does it reproduce swaption volatility matrix?

{ can volatility change randomly?

{ can it reproduce volatility smiles and skews?

{ does it allow \twists" in spot rate curve?

Summary assessment:

Ho-Lee Black-Derman-Toy Hull-White

spot rates?

yes

yes

yes

vol term str? no

yes

no

sensitivities? no

no?

maybe

vol matrix?

no

no

no

random vol? no

no

no

smiles?

no

no

no

twists?

no

no

no

6-3

2. Assessment (continued)

Ho and Lee

{ the innovation was to match current spot rates

{ didn't do the next step: volatilities

Black, Derman, and Toy

{ matches volatility term structure

{ short rate or log of short rate? (not clear)

Hull and White

{ but where does it hurt us?

{ technical trick: build mean reversion into trinomial tree

{ volatility matrix implied by models, not a exible input

{ no volatility smiles

{ one-factor structures

6-4

6-5

3. Vasicek Revisited

The model

, log m +1 = 2=2 + z + " +1

z +1 = (1 , ') + 'z + " +1

t

Pricing relation:

b +1 = E (m +1b +1)

n

Solution is log-linear:

, log b = A + B z

n

with

B +1 = 1 + B '

(start with A0 = B0 = 0)

n

Forward rates

{ denition is

f = log(b =b +1)

{ Vasicek solution is

f = ' z + constant

n

(think: Ho-Lee and BDT set ' = 1)

6-6

4. Stochastic Volatility

Cox-Ingersoll-Ross model

, log m +1 = (1 + 2=2)z + z 1 2" +1

z +1 = (1 , ') + 'z + z 1 2" +1

=

Comments:

{ conditional variance varies with z : Var z +1 = 2z

{ square root keeps z positive

{ mean reversion

t

, log b = A + B z

n

with

A +1 = A + B (1 , ') , ( + B )2 =2

B +1 = 1 + 2=2 + B ' , ( + B )2=2

(start with A0 = B0 = 0)

n

\stochastically"

6-7

Overview:

{ approach based on forward rates

{ allows input of volatility matrix

{ many variants | we focus on a linear one

Pricing relation for forward rates:

1 = E (m +1 R +1)

X ,1

log R +1 = r , (f +1

,f )

t

=1

Behavior of forward rates (we start to get specic here):

f +1,1 = f + + " +1

Comments:

{ note the volatility input : varies with date t and

maturity n (a matrix!)

{ linear structure common, not necessary

{ Vasicek is similar:

f +1,1 = f + constant + ' ,1" +1

(note the specic relation of volatility to maturity)

{ return is

X

X

log R +1 = r , , " +1

=1

=1

= r , A , S " +1

(A and S are partial sums)

{ other versions have multiple "'s

n

nt

nt

nt

jt

jt

nt

nt

6-8

Arbitrage-free pricing

{ a pricing kernel:

, log m +1 = + " +1

{ pricing relation imposes restrictions:

A , S , S 2 =2 = 0

t

nt

nt

nt

Calibration

{ inputs:

current forward rates: ff g

volatility matrix: f g

{ drift parameters f g chosen to satisfy arbitrage relation

{ open question: set = 0? (more shortly)

n

t

nt

n;t

nt

Implementation on trees

{ at each node we have complete forward rate curve

{ branches don't typically \recombine"

6-9

Approaches:

{ t smooth curve to observed implied volatilities

{ \implied binomial trees" that allow volatility to vary across

states as well as dates (see Chriss's book)

{ continuous models that extend Black-Scholes logic to

non-normal distributions

Gram-Charlier smiles

{ Gram-Charlier expansion adds terms for skewness ( 1) and

kurtosis ( 2) to the normal (where 1 = 2 = 0)

{ Wu's approximation of a volatility smile:

#

"

2

1

2

v = 1 , 3! d , 4! (1 , d )

where

2 =2

log(

F=K

)

+

v

d=

v

{ intuition:

positive skewness raises value of out-of-the-money calls

positive kurtosis raises probability of extreme events

and value of out-of-the-money puts and calls

6-10

7. Multi-Factor Models

Problems with one-factor models

{ changes in rates of all maturities tied to a single random

variable (")

{ shifts in spot rate curve come in only one type (parallel?)

{ correlation of spot rates across maturities restricted

{ spreads typically not variable enough

{ the model:

, log m +1 = + X(2=2 + z + "

t

it

+1

i;t

{ solution includes

= 'z + "

i

it

+1 )

j;t

+1

i;t

!2

X4 2

X

1

,

'

1

f = + 2 , + 1,' 5+ ' z :

{ standard solution: '1 close to one, '2 smaller )

z1 generates almost parallel shifts, z2 \twists"

long rates dominated by z1, spreads by z2

{ generates better behavior of spreads

n

i

it

6-11

We've taken two approaches to valuation

{ a pricing kernel (Vasicek, for example)

{ risk-neutral probabilities (binomial models)

How are they related?

{ q is value now of 1 dollar in down state next period

{ valuation of cash ows (c ; c ) follows

u

p = q c +q c

{ one-period discount factor is b = q + q = exp(,rh)

u

Risk-neutral probabilities

{ dene = q =(q + q ), = q =(q + q )

{ note: ( ; ) are positive and sum to one

u

(they're probabilities!)

{ valuation follows

p = exp(,rh)( c + c )

u

Pricing kernel

{ dene q = m , q = m (\true probabilities")

{ valuation follows

u

p = m c + m c

u

6-12

Where's the pricing kernel in a binomial model?

{ gure it out:

, log m +1 = + r + " +1;

t

with

2 = log( =) , log( =)

u

risk-neutral probabilities (and is zero when the two are

equal)

Is = 0 a mistake?

{ caveat: Black-Scholes pricing doesn't depend directly on

(so maybe it's not a bad mistake in general)

{ bottom line: who knows?

6-13

Summary

1. Models are

simplications of reality

ways to ensure consistency of pricing across assets

only as good s (and
oors) and swaptions

2. Binomial models capture some of the elements of observed asset

prices, but most versions leave some issues open:

stochastic volatility

volatility smiles (again, more work)

multiple factors (possible, just more work)

can we ignore ?

3. Modeling remains as much art as science

- C A S E 28 Inner-City Paint Corporation (Revised)Uploaded byMasud Rana
- Splash Corporation BackgroundUploaded byAllan Francis Villegas
- Hong Kong Trading TechniquesUploaded bycean
- 4610383664-461 (1)Uploaded bysunil295
- Du Gayy Salaman (1992) The Culture of the CustomerUploaded byFernando Francisco
- Leadership in the 21st CenturyUploaded byJohn Sutherland
- Taal View Heights ComputationsUploaded byJayvee
- SWOT Analysis TemplateUploaded byperishz
- EQUITY Morning Bell 091211Uploaded byshailendraresearch
- Visual Basic Formulas de ErlangUploaded byMayra Trigueros
- The Deindustrialization of Many MarketsUploaded byJoshua Ioji Konov
- MergersUploaded bySrinidhi Rangarajan
- Test Paper 2 COUploaded bydarshankaroshi
- TopicsUploaded byLediana Mehmeti
- 07Simplex3BUploaded byPotnuru Vinay
- Sales and Shipping - Consumption Data for Items (From 20B) - Order Fulfillment - SAP LibraryUploaded byyaswanthrdy
- 8Uploaded byapi-373634605
- Build It OverviewUploaded bywaste33
- 01- Dynamics of Business and EconomicsUploaded byMartinus Warsito
- Business ProposalUploaded byLuis Abdul-Melendez
- Velasquez C6Uploaded byZ
- Liberty - September 28 2017Uploaded byTiso Blackstar Group
- 101017 LibertyUploaded byTiso Blackstar Group
- Liberty - October 18 2017Uploaded byTiso Blackstar Group
- Liberty - October 19 2017Uploaded byTiso Blackstar Group
- Liberty - October 20 2017Uploaded byTiso Blackstar Group
- g1123Uploaded byKamal Dave
- Power 2006Uploaded byRongLiu
- 18313-37303-1-SMUploaded byNaisya
- sbd0254.57.98.4803.000010.pdfUploaded byZaphod

- Kecacatan bangunanUploaded byzaman musa
- DILLI HAAT Case StudyUploaded byvijeta93
- Acct 260 Chapter 14Uploaded byJohn Guy
- Windows Shortcut HandoutUploaded bywalor4real83
- Installation GuideUploaded byBryan Yaciuk
- Stress-Strain Creep and Temperature Dependency of ADSS Sag AnUploaded byденис логинов
- Application of Dry Cell Hho Gas Generator With Pulse Width Modulation on Sinjai Spark Ignition Engine PerformanceUploaded byesatjournals
- Effective Management Diverse Teams-ToolkitUploaded byHo Uyen
- Kyocera KM-2560 Operations GuideUploaded bySheri Rosa
- Germantown Express News 080913Uploaded byHometown Publications - Express News
- Amul Projects by Ravi Kant SainiUploaded byRavi Kant Saini
- Learning Needs AnalysisUploaded byJessie
- Mr Satprem Maini_Auroville.pdfUploaded bysubi29
- Fiche 72-600 Septembre 2014Uploaded byMakhou Ladoum
- wsr2013Uploaded byDen Sabando
- DBC Format 2007Uploaded bythoroftedal
- jan09_capdebtFY2009Uploaded byEliot Brown / New York Observer
- F585 Globalisation Essay NotesUploaded byLauren Chua
- DTMPUploaded byChandeshwor Shah
- Software EngineeringUploaded byanisha shetty
- Synopsis Sample YM Showcase 2018Uploaded bySheri Abhishek Reddy
- c++ and javaUploaded byAnand Dhanalakota
- The Economic War Against the Jewish StateUploaded byEriden Estrella
- the oneUploaded bydallexcis
- TCS3P125.pdfUploaded byJuan José Tovar Pérez
- 2017-12-16 - PC Express - Dealers Price ListUploaded byPocholo Malvar
- One Day Workshop on FEMA by CA.Uploaded byTaxpert mukesh
- Peter Dacoutros Replies to Lino SpiteriUploaded bysevee2081
- msds formUploaded byapi-262008253
- MTN GeneXpert Presentation Natasha 2016Uploaded byRajeev Pareek

## Much more than documents.

Discover everything Scribd has to offer, including books and audiobooks from major publishers.

Cancel anytime.