Computations in the Hull-White Model

Similar documents
Tentamen i 5B1575 Finansiella Derivat. Måndag 27 augusti 2007 kl Answers and suggestions for solutions.

Hull-White one factor model Version

7 pages 1. Hull and White Generalized model. Ismail Laachir. March 1, Model Presentation 1

Matematisk statistik Tentamen: kl FMS170/MASM19 Prissättning av Derivattillgångar, 9 hp Lunds tekniska högskola. Solution.

Black-Scholes Model and Risk Neutral Pricing

The Mathematics Of Stock Option Valuation - Part Four Deriving The Black-Scholes Model Via Partial Differential Equations

Tentamen i 5B1575 Finansiella Derivat. Torsdag 25 augusti 2005 kl

Heath Jarrow Morton Framework

MAFS Quantitative Modeling of Derivative Securities

Extended One-Factor Short-Rate Models

Research Article A General Gaussian Interest Rate Model Consistent with the Current Term Structure

INSTITUTE OF ACTUARIES OF INDIA

where r() = r(s)e a( s) + α() α(s)e a( s) + σ e a( u) dw(u) s α() = f M (0, ) + σ a (1 e a ) Therefore, r() condiional on F s is normally disribued wi

Once we know he probabiliy densiy funcion (pdf) φ(s ) of S, a European call wih srike is priced a C() = E [e r d(s ) + ] = e r d { (S )φ(s ) ds } = e

Models of Default Risk

Alexander L. Baranovski, Carsten von Lieres and André Wilch 18. May 2009/Eurobanking 2009

Market Models. Practitioner Course: Interest Rate Models. John Dodson. March 29, 2009

Brownian motion. Since σ is not random, we can conclude from Example sheet 3, Problem 1, that

INTEREST RATES AND FX MODELS

Introduction to Black-Scholes Model

MA Advanced Macro, 2016 (Karl Whelan) 1

Option Valuation of Oil & Gas E&P Projects by Futures Term Structure Approach. Hidetaka (Hugh) Nakaoka

Interest Rate Products

May 2007 Exam MFE Solutions 1. Answer = (B)

Lecture Notes to Finansiella Derivat (5B1575) VT Note 1: No Arbitrage Pricing

CHAPTER 3 How to Calculate Present Values. Answers to Practice Questions

Bond Prices and Interest Rates

Option pricing and hedging in jump diffusion models

1. (S09T3) John must pay Kristen 10,000 at the end of 1 year. He also must pay Ahmad 30,000 at the end of year 2.

FIXED INCOME MICHAEL MONOYIOS

Term Structure Models: IEOR E4710 Spring 2005 c 2005 by Martin Haugh. Market Models. 1 LIBOR, Swap Rates and Black s Formulae for Caps and Swaptions

An Analytical Implementation of the Hull and White Model

Financial Econometrics (FinMetrics02) Returns, Yields, Compounding, and Horizon

On multicurve models for the term structure.

Change of measure and Girsanov theorem

1. (S09T3) John must pay Kristen 10,000 at the end of 1 year. He also must pay Ahmad 30,000 at the end of year 2.

IJRSS Volume 2, Issue 2 ISSN:

UCLA Department of Economics Fall PhD. Qualifying Exam in Macroeconomic Theory

Pricing FX Target Redemption Forward under. Regime Switching Model

Elton, Gruber, Brown, and Goetzmann. Modern Portfolio Theory and Investment Analysis, 7th Edition. Solutions to Text Problems: Chapter 21

HULL-WHITE ONE FACTOR MODEL: RESULTS AND IMPLEMENTATION

Exotic FX Swap. Analytics. ver 1.0. Exotics Pricing Methodology Trading Credit Risk Pricing

Pricing formula for power quanto options with each type of payoffs at maturity

LIBOR MARKET MODEL AND GAUSSIAN HJM EXPLICIT APPROACHES TO OPTION ON COMPOSITION

Forwards and Futures

VALUATION OF THE AMERICAN-STYLE OF ASIAN OPTION BY A SOLUTION TO AN INTEGRAL EQUATION

Optimal Consumption and Investment with Habit Formation and Hyperbolic discounting. Mihail Zervos Department of Mathematics London School of Economics

AN EASY METHOD TO PRICE QUANTO FORWARD CONTRACTS IN THE HJM MODEL WITH STOCHASTIC INTEREST RATES

STOCHASTIC METHODS IN CREDIT RISK MODELLING, VALUATION AND HEDGING

Math 373 Fall 2016 Test 3 November 15, 2016

FAIR VALUATION OF INSURANCE LIABILITIES. Pierre DEVOLDER Université Catholique de Louvain 03/ 09/2004

MATH 373 Test 4 Spring 2017 May 5, 2017

Economic Growth Continued: From Solow to Ramsey

Basic Economic Scenario Generator: Technical Specications. Jean-Charles CROIX ISFA - Université Lyon 1

VaR and Low Interest Rates

Foreign Exchange, ADR s and Quanto-Securities

Hull & White Convexity Adjustments for Credit Riskless Interest Rate Swaps Under CSA

Equivalent Martingale Measure in Asian Geometric Average Option Pricing

PRICING OF SHIBOR-BASED FLOATING-RATE BONDS IN CHINA

Pricing corporate bonds, CDS and options on CDS with the BMC model

Macroeconomics II A dynamic approach to short run economic fluctuations. The DAD/DAS model.

Interest rate models enhanced with local volatility

You should turn in (at least) FOUR bluebooks, one (or more, if needed) bluebook(s) for each question.

MORNING SESSION. Date: Wednesday, October 30, 2013 Time: 8:30 a.m. 11:45 a.m. INSTRUCTIONS TO CANDIDATES

Drift conditions on a HJM model with stochastic basis spreads. Teresa Martínez Quantitative Product Group. Santander Quantitative Product Group

Brownian Moving Averages and Applications Towards Interst Rate Modelling

Systemic Risk Illustrated

LIDSTONE IN THE CONTINUOUS CASE by. Ragnar Norberg

Research Paper Series. No. 64. Yield Spread Options under the DLG Model. July, 2009

DEBT INSTRUMENTS AND MARKETS

MORNING SESSION. Date: Wednesday, April 26, 2017 Time: 8:30 a.m. 11:45 a.m. INSTRUCTIONS TO CANDIDATES

Proceedings of the 48th European Study Group Mathematics with Industry 1

AMS Q03 Financial Derivatives I

Modeling of Tradeable Securities with Dividends

Estimation of Future Initial Margins

Robust Replication of Default Contingent Claims

where lnp(, ) f(, ) = P(, ) = exp { f(, u)du} = exp{q(, )} Q(, ) = f(, u)du Heah, Jarrow, and Moron (1992) claimed ha under risk-neural measure, he dr

A Note on Forward Price and Forward Measure

Dynamic Programming Applications. Capacity Expansion

INSTITUTE OF ACTUARIES OF INDIA

Econ 546 Lecture 4. The Basic New Keynesian Model Michael Devereux January 2011

a. If Y is 1,000, M is 100, and the growth rate of nominal money is 1 percent, what must i and P be?

Available online at ScienceDirect

AMS Computational Finance

Martingale Methods in Financial Modelling

Erratic Price, Smooth Dividend. Variance Bounds. Present Value. Ex Post Rational Price. Standard and Poor s Composite Stock-Price Index

ECON Lecture 5 (OB), Sept. 21, 2010

University of Cape Town

Applications of Interest Rate Models

On Monte Carlo Simulation for the HJM Model Based on Jump

Bruno Dupire. Banque Paribas Swaps and Options Research Team 33 Wigmore Street London W1H 0BN United Kingdom

Modeling of Tradeable Securities with Dividends

On the multiplicity of option prices under CEV with positive elasticity of variance

Pricing Inflation-Indexed Derivatives Using the Extended Vasicek Model of Hull and White

Continuous-time term structure models: Forward measure approach

Black-Scholes and the Volatility Surface

Mean Field Games and Systemic Risk

db t = r t B t dt (no Itô-correction term, as B has finite variation (FV), so ordinary Newton- Leibniz calculus applies). Now (again as B is FV)

HEDGING VOLATILITY RISK

Heath Jarrow Morton (HJM) Methodology

Transcription:

Compuaions in he Hull-Whie Model Niels Rom-Poulsen Ocober 8, 5 Danske Bank Quaniaive Research and Copenhagen Business School, E-mail: nrp@danskebank.dk

Specificaions In he Hull-Whie model, he Q dynamics of he spo rae is given by he following sochasic differenial equaion SDE) also know as he Ohrnsein-Uhlenbeck process dr) = Θ) r)) d + σdw) ) where Θ) is he long erm level o which he spo rae, r), is moving, is he rae a which he spo rae rae is pushed owards he long erm level and W) is a Brownian moion under Q. σ is he consan volailiy of changes in he spo rae. σ is assumed consan in his noe. Using he spo rae defined by ), we consruc a money marke accoun by A) = e Ê rudu ) da) = r A)d 3) The specificaion of he spo rae, means ha he Hull-Whie model belongs o he affine class of ineres rae models and hus prices of zero coupon bonds a ime for he ime T mauriy have he following form The T-yield a ime, y, T) is defined as P, T) = e α,t)+β,t)r 4) y, T) = lnp, T) T We generally wan he model o be calibraed o he marke oday meaning ha model prices of zero coupon bonds oday, P, T) T, is equal o he prices observed in he marke. This can be achieved by choosing Θ) in equaion ) so ha he iniial yield curve is mached. Closed orm Soluion for Prices of Zero Coupon Bonds We will now find explici formulas for he funcions α, T) and β, T) in 4) and hus closed form soluions for zero coupon bonds in he Hull-Whie model. irs, however, we derive he fundamenal parial differenial equaion for zero coupon prices in he Hull- Whie model. Sar by finding he dynamics of zero coupon prices by employing Io s lemma. dp, T) = P P d + r dr) + P dr)) r Insering he spo rae dynamics ) yields ) dp, T) α, T) β, T) = + r) d + β, T)dr) + P, T) β, T)dr)) α, T) β, T) = + r) + β, T)Θ) β)r + ) β, T)σ d +β, T)σdW)

We now use he dynamics of he money marke accoun given by 3) and he dynamics of he zero coupon bonds in 5) o find he dynamics of deflaed zero coupon bond prices {}}{ da)dp, T) = A)dP, T) + P, T)dA) + da)dp, T) Again insering wha is know we ge = da)dp, T) A)P, T) = α, T) + β, T) r) + β, T)Θ) β)r + β, T)σ r ) d + β, T)σdW) Under he equivalen maringale measure, Q, deflaed prices are maringales. According o he maringale represenaion heorem we mus hus have ha he d-erm mus be equal o zero, and his holds for all and r. Thus α, T) + β, T)Θ) + β, T)σ = β, T) αt, T) = β, T) = βt, T) = 5) 6) 7) 8) We solve he wo ordinary differenial equaions by firs posulaing a soluion for β, T) β, T) = ) e T ) 9) I is easy o check ha he soluion in 9) in fac solves he ODE in 7) subjec o he boundary condiion in 8). Since he derivaive of α, T) only depends on β, T) simple inegraion of αu,t) u beween and T is a soluion o 5). Recall ha and hus we have α, T) du = αu, T)] T = αt, T) α, T) ) α, T) = βu, T)Θu)du + β u, T)σ du ) As menioned above we wan he model o mach curren zero coupon prices. This is done by choosing Θu) in ) so ha he iniial yield curve is mached. Insead of calibraing he model o zero coupon yields direcly, we calibrae he model o he erm srucure of forward raes. orward raes are defined as rom 9) we ge f M lnp, T), T) = α, T) β, T)r β, T) = e T ) )

Using Leibniz s rule for differeniaing inegrals we have from ), 6) and 8) T α, T) = βt, T)ΘT) + β T, T)σ + σ βu, T)Θu)du βu, T) βu, T)du Insering 9) and ) yields T α, T) = = Puing hings ogeher we ge e T u) Θu)du σ e T u) Θu)du σ ) e T u) e T u) du e T ) e T)] f M, T) = e T u) Θu)du 3) + σ e T ) e T)] + e T r 4) To isolae ΘT) we differeniae wih respec o T f M, T) Using 3) his can be wrien as And hus f M, T) = ΘT) e T u) Θu)du + σ e T e T) e T r = ΘT) f M, T) + σ e T)) e T))] +e T r + σ e T e T) e T r = ΘT) f M, T) σ ΘT) = fm, T) e T ) + f M, T) + σ e T ) 5) Now ha we know ΘT) we can plug i ino ) o find an expression for α, T). We compue firs he inegral σ = σ = σ β u, T)du e T u) ) du e T )) + T ) e T ))] 6) 3

Nex we compue he inegral Insering 5) yields βu, T)Θu)du = T + σ = f e T u) M, u) u ) e T u) Θu)du e T u) Θu)du Compuing he las inegral yields σ e T u) ) e u ) du = We now have σ 3 ) + f M, u) du Θu)du f M, u) f M, u)du u e T u) ) e u ) du 7) e T ) + e T e T+) + ] e σ T ) + σ 3 σ βu, T)Θu)du = e T u)fm, u) T du + e T u) f M, u)du u f M, u) ] T f M, u)du e T ) + e T e T+) + ] e T ) 8) Now use he inegraion by pars formula on he firs erm on he righ hand side in equaion 8) T f e T u) M ), u) du = ] T e T u) f M, u) u e T u) f M, u)du o ge + βu, T)Θu)du = ] T e T u) f M, u) + σ 3 e T u) f M, u)du e T u) f M, u)du f M, T) f M, ) ) f M, u)du e T ) + e T e T+) + ] e σ T ) 9) 4

Now simplifying gives + σ 3 βu, T)Θu)du = f M, )β, T) f M, u)du e T ) + e T e T+) + ] e σ T ) ) Combining he wo inegrals 6) and ) we ge α, T) = Simplifying yields βu, T)Θu)du + σ β, T)du = f M, )β, T) f M, u)du + σ 3 e T ) + e T e T+) + ] e σ + σ e T )) + T ) T e ))] T ) We also have ha α, T) = f M, )β, T) f M, u)du + σ 4 β, T) e ) ) P, T) = f,u)du which leads o and hus and we have ln lnp, T) = f, u)du ) P, T) T = f, u)du P, ) α, T) = f M, )β, T) + ln ) P, T) P, ) + σ 4 β, T) e ) ) 3 Solving he Sochasic Differenial Equaion The soluion o he SDE in equaion ) can be found by employing Io s lemma o find he dynamics of e r) and hen inegraing up. This yields he soluion o ) r T = e T ) r + e T u) Θu)du + σ e T u) dw u 5

Since E Q σ e T u) dw u ] =, we have Noice ha E Q r T] = e T ) r + e T u) Θu)du Var Q r T] = E Q e T u) Θu)du = Liebniz s rule for differeniaing inegral gives ) ] r E Q r T] = σ e T u) du βu, T)Θu)du = βt, T)ΘT) + βu, T)Θu)du βu, T)Θu)du where he firs erm on he righ hand side is equal o zero according o 8). Thus we have e T u) Θu)du = Differeniaing ) wih respec o T we ge βu, T)Θu)du e T u) Θu)du = f M, ) β, T) + f M, u)du σ ) e T ) e T + e T+) = f M, )e T ) + f M, T) fm, u) }{{} σ ) e T ) e T + e T+) Now add and subrac σ e kt ) and simplify o ge where e T u) T ) Θu)du = γt) γ)e γ) = f M, ) + σ e k) Thus he condiional expecaion of he fuure spo rae can now be wrien as And he condiional varians is given by E Q r T] = e T ) r + γt) γ)e T ) 3) Var Q = σ e T )) 4) = 6

4 loaers and Caples In his secion he Hull-Whie model is used o value a fuure sochasic rae oday. We wan o compue he price of a paymen received a ime T A. This paymen covers ineres over he period from ime T A o T A. The paymen is no known unil ime T where i is fixed as he simple rae over he period T o T. T T A T A T The following noaion is used T : Sar dae of fixing period T : End dae of fixing period T A : Sar dae of accrual period T A : End dae of accrual period : Accrual period in years T A T A ) : ixing period in years T T ) A ime T he spo LIBOR rae over he fixing period is given by LT,T ) = ) PT,T ) 5) Since he accrual period is he paymen received a ime T A is equal o LT,T ). The value a ime of he unknown paymen is given by V ) = E Q e Ê ] T A r u du LT,T ) 6) = E Q = E Q A T r u du LT,T )E Q T ] r u du PT,T A ) LT,T ) ]] r u du Changing from he risk-neural measure o he forward measure wih he zero-coupon bond mauring a ime T as numeraire yields V ) = P, T )E T 7) 8) PT, T A ) LT,T ) ] 9) which is he same as V ) = P, T ) E T PT, T A )] ) PT,T ) PT,T A ) { = P, T )E T PT, T A )] } ) PT,T P, T A ) ) In he Hull-Whie model, he raion PT,T A )/PT,T ) is equal o PT,T A ) PT,T ) = eαt,ta ) αt,t )+βt,ta ) βt,t ))r T 3) 7

Which is used o ge { V ) = P, T )e αt,ta ) αt,t ) E T The only unknown objec in Equaion 3) is E T ] } e βt,ta ) βt,t ))r T P, T A ) 3) ] e βt,ta ) βt,t ))r T, bu his expecaion can easily be compued since r T is normally disribued. )] E T exp βt,t A ) βt,t ))r T = exp βt,t A ) βt,t ))E T r T ] + ) βt,t A ) βt,t )) Var T r T ] which can be insered ino 3) o ge a closed form soluion for he value of a single floa paymen. ixing a an Average Rae V ) = E Q = M A r u du M i= E Q M A ] M LT i, T i) i= r u du LT i, T i) The las expression has exacly he same form as he expression found in he previous secion and i s value is hus know in closed form. Caple Pricing wih Unnaural Time Lag We will now find he value of a caple ha caps he simple compounded ineres rae given in Equaion 5). The rae is fixed over he period from T o T bu i is paid ou a ime T A. cpl, T,T,T A ) = E Q = P, T A )E TA = P, T A )E TA = P, T A )E TA A r u du LT, T ) K ) ] + = P, T A )e αt,t ) E TA LT, T ) K ) + ] ) ] + PT,T ) + K) ) ] + e αt,t ) βt,t )r T + K) ] 3) ) ] + e βt,t )r T e αt,t ) + K) Since r T ΦM r, V r ), where M r and V r is he mean and variance of r T respecively and Φ is he sandard cumulaive normal disribuion funcion, we have βt,t )r T Φ βt,t )M r, βt,t ) Vr ). rom Brigo and Mercurio ) we have he following for a lognormally disribued sochasic variable X wih mean M and variance V E X K) +] ) ) = e M+ V M lnk) + V M lnk) Φ KΦ 33) V V 8

The caple value can now be compued wih M = βt,t )M r and V = βt,t )V r which yields cpl, T,T,T A ) = P, T A )e αt,t ) e βt,t )Mr+ βt,t ) V r Φd ) ) e αt,t ) + K) Φ d βt,t ) ) )V r 34) where d = βt,t )M r lne αt,t ) + K)) + βt,t ) Vr βt,t 35) )V r 9

References Brigo, D. and. Mercurio ): Springer, firs ediion. Ineres Rae Models Theory and Pracice,