No-Arbitrage Conditions for a Finite Options System

Similar documents
Notes: This is a closed book and closed notes exam. The maximal score on this exam is 100 points. Time: 75 minutes

Non replication of options

AN ANALYTICALLY TRACTABLE UNCERTAIN VOLATILITY MODEL

On an optimization problem related to static superreplicating

Necessary and Sufficient Conditions for No Static Arbitrage among European Calls

( 0) ,...,S N ,S 2 ( 0)... S N S 2. N and a portfolio is created that way, the value of the portfolio at time 0 is: (0) N S N ( 1, ) +...

Applying the Principles of Quantitative Finance to the Construction of Model-Free Volatility Indices

Options. An Undergraduate Introduction to Financial Mathematics. J. Robert Buchanan. J. Robert Buchanan Options

Option Properties Liuren Wu

INTRODUCTION TO ARBITRAGE PRICING OF FINANCIAL DERIVATIVES

Analysis of the sensitivity to discrete dividends : A new approach for pricing vanillas

The Forward PDE for American Puts in the Dupire Model

American options and early exercise

3.2 No-arbitrage theory and risk neutral probability measure

4: SINGLE-PERIOD MARKET MODELS

HIGHER ORDER BINARY OPTIONS AND MULTIPLE-EXPIRY EXOTICS

Option Pricing. Chapter Discrete Time

MATH 425 EXERCISES G. BERKOLAIKO

How do Variance Swaps Shape the Smile?

Financial Derivatives Section 3

Chapter 2 Questions Sample Comparing Options

Econ Financial Markets Spring 2011 Professor Robert Shiller. Problem Set 6

Lecture 8: Asset pricing

No-arbitrage Pricing Approach and Fundamental Theorem of Asset Pricing

MATH3075/3975 FINANCIAL MATHEMATICS TUTORIAL PROBLEMS

Finance: Lecture 4 - No Arbitrage Pricing Chapters of DD Chapter 1 of Ross (2005)

Pricing with a Smile. Bruno Dupire. Bloomberg

MASSACHUSETTS INSTITUTE OF TECHNOLOGY 6.265/15.070J Fall 2013 Lecture 19 11/20/2013. Applications of Ito calculus to finance

A No-Arbitrage Theorem for Uncertain Stock Model

RECOVERING RISK-NEUTRAL PROBABILITY DENSITY FUNCTIONS FROM OPTIONS PRICES USING CUBIC SPLINES

MATH 5510 Mathematical Models of Financial Derivatives. Topic 1 Risk neutral pricing principles under single-period securities models

1.12 Exercises EXERCISES Use integration by parts to compute. ln(x) dx. 2. Compute 1 x ln(x) dx. Hint: Use the substitution u = ln(x).

From Discrete Time to Continuous Time Modeling

SMILE EXTRAPOLATION OPENGAMMA QUANTITATIVE RESEARCH

University of Colorado at Boulder Leeds School of Business MBAX-6270 MBAX Introduction to Derivatives Part II Options Valuation

Volatility of Asset Returns

Lecture 16: Delta Hedging

Advanced Stochastic Processes.

TEACHING NOTE 98-04: EXCHANGE OPTION PRICING

3 Arbitrage pricing theory in discrete time.

Lecture 4. Finite difference and finite element methods

Option Pricing. 1 Introduction. Mrinal K. Ghosh

Name: T/F 2.13 M.C. Σ

Forwards and Futures. Chapter Basics of forwards and futures Forwards

Martingale Pricing Theory in Discrete-Time and Discrete-Space Models

LECTURE 2: MULTIPERIOD MODELS AND TREES

Department of Mathematics. Mathematics of Financial Derivatives

University of Texas at Austin. HW Assignment 5. Exchange options. Bull/Bear spreads. Properties of European call/put prices.

PRMIA Exam 8002 PRM Certification - Exam II: Mathematical Foundations of Risk Measurement Version: 6.0 [ Total Questions: 132 ]

Equity Warrant Difinitin and Pricing Guide

Multivariate Binomial Approximations 1

Unbiased Expectations Theory

Arbitrage-Free Pricing of XVA for Options in Discrete Time

American Option Pricing: A Simulated Approach

A Harmonic Analysis Solution to the Basket Arbitrage Problem

Follow links for Class Use and other Permissions. For more information send to:

Mathematics in Finance

ECON4510 Finance Theory Lecture 10

A Preference Foundation for Fehr and Schmidt s Model. of Inequity Aversion 1

Chapter 15: Jump Processes and Incomplete Markets. 1 Jumps as One Explanation of Incomplete Markets

Homework Assignments

Online Shopping Intermediaries: The Strategic Design of Search Environments

Pricing Implied Volatility

We consider three zero-coupon bonds (strips) with the following features: Bond Maturity (years) Price Bond Bond Bond

PAULI MURTO, ANDREY ZHUKOV. If any mistakes or typos are spotted, kindly communicate them to

d St+ t u. With numbers e q = The price of the option in three months is

FINANCIAL OPTION ANALYSIS HANDOUTS

- Introduction to Mathematical Finance -

The Binomial Model for Stock Options

Mean-Variance Optimal Portfolios in the Presence of a Benchmark with Applications to Fraud Detection

Final Exam. Please answer all four questions. Each question carries 25% of the total grade.

sample-bookchapter 2015/7/7 9:44 page 1 #1 THE BINOMIAL MODEL

OPTION PRICE WHEN THE STOCK IS A SEMIMARTINGALE

Hedging Basket Credit Derivatives with CDS

S 0 C (30, 0.5) + P (30, 0.5) e rt 30 = PV (dividends) PV (dividends) = = $0.944.

Bounds on some contingent claims with non-convex payoff based on multiple assets

Option Pricing Models for European Options

Help Session 2. David Sovich. Washington University in St. Louis

Implied Volatilities

Binomial Option Pricing and the Conditions for Early Exercise: An Example using Foreign Exchange Options

COMP331/557. Chapter 6: Optimisation in Finance: Cash-Flow. (Cornuejols & Tütüncü, Chapter 3)

One Period Binomial Model: The risk-neutral probability measure assumption and the state price deflator approach

Final Projects Introduction to Numerical Analysis Professor: Paul J. Atzberger

M5MF6. Advanced Methods in Derivatives Pricing

THE USE OF NUMERAIRES IN MULTI-DIMENSIONAL BLACK- SCHOLES PARTIAL DIFFERENTIAL EQUATIONS. Hyong-chol O *, Yong-hwa Ro **, Ning Wan*** 1.

Lecture 8: Introduction to asset pricing

Mathematics of Finance Final Preparation December 19. To be thoroughly prepared for the final exam, you should

Pricing Options with Mathematical Models

Lecture 12. Stock Option boundary conditions. Agenda:

ECON FINANCIAL ECONOMICS

Computing Bounds on Risk-Neutral Measures from the Observed Prices of Call Options

A note on sufficient conditions for no arbitrage

RMSC 2001 Introduction to Risk Management

Gallery of equations. 1. Introduction

18. Diffusion processes for stocks and interest rates. MA6622, Ernesto Mordecki, CityU, HK, References for this Lecture:

Robust Trading of Implied Skew

Lecture 16. Options and option pricing. Lecture 16 1 / 22

DERIVATIVE SECURITIES IMBA Fudan University The University of Hong Kong Second Semester 2003/2004

SINCE THE CHICAGO BOARD OPTIONS EXCHANGE INTRODUCED THE FIRST INDEX OPTION CON-

FX Smile Modelling. 9 September September 9, 2008

Transcription:

No-Arbitrage Conditions for a Finite Options System Fabio Mercurio Financial Models, Banca IMI Abstract In this document we derive necessary and sufficient conditions for a finite system of option prices (with the same maturity) to be arbitrage free. We start by considering the particular case of three market quotes (European calls), which is typical of the FX options market. We then derive stricter conditions by allowing also the trading in forward contracts and in the underlying asset. We finally hint at the general case of an arbitrary, but finite, number of given option prices. Notation and Conventions S t : the underlying asset price at time t. r: the (constant) domestic instantaneous risk-free rate. q: the (constant) foreign instantaneous risk-free rate. T : a fixed maturity. C(X): the market price of a European call option with maturity T and strike X. P (X): the market price of a European put option with maturity T and strike X. We call positive the real numbers strictly larger than zero. Any number larger than or equal to zero is then called nonnegative. 1.1 The Case of Three European Call Prices In a given options market, we consider a maturity T and three strikes K 1, K 2 and K 3, K 1 < K 2 < K 3. We denote by C(K 1 ), C(K 2 ) and C(K 3 ) the related option prices. In the following, we derive necessary and sufficient conditions for the three prices to form an arbitrage-free system. 1 Proposition 1.1. The three option prices C(K 1 ), C(K 2 ) and C(K 3 ) form an arbitrage-free system if and only if C(K 3 ) > 0 C(K 2 ) > C(K 3 ) (1) ( )C(K 1 ) (K 3 K 1 )C(K 2 ) + (K 2 K 1 )C(K 3 ) > 0 1 These conditions can also be found in Cox and Rubinstein (1985). 1

Proof. The option prices C(K 1 ), C(K 2 ) and C(K 3 ) form an arbitrage-free system if and only if every linear combination of the related calls yielding a nonnegative payoff that is positive with positive probability, has a positive price. Since, for any (real) combinators α 1, α 2 and α 3, 0 x K 1 α i (x K i ) + α 1 (x K 1 ) K 1 < x K 2 = α 1 (x K 1 ) + α 2 (x K 2 ) K 2 < x K 3 (α 1 + α 2 + α 3 )x α 1 K 1 α 2 K 2 α 3 K 3 x > K 3 we then have α 1 0 α i (S T K i ) + 0 α 1 (K 3 K 1 ) + α 2 ( ) 0 α 1 + α 2 + α 3 0 Setting ᾱ := (α 1, α 2, α 3 ) and A := {ᾱ (0, 0, 0) : α 1 0, α 1 (K 3 K 1 ) + α 2 ( ) 0, α 1 + α 2 + α 3 0} we have therefore an arbitrage-free system if and only if α i C(K i ) > 0 for each ᾱ A (2) Defining the new variables vector β = (β 1, β 2, β 3 ) as β 1 = α 1 β 2 = α 1 (K 3 K 1 ) + α 2 ( ) β 3 = α 1 + α 2 + α 3 condition (2) is equivalent to β 1 ( C(K 1 ) K 3 K 1 C(K 2 ) + K 2 K 1 C(K 3 ) ) + β 2 C(K 2 ) C(K 3 ) + β 3 C(K 3 ) > 0 (3) for each β B := { β (0, 0, 0) : β 1 0, β 2 0, β 3 0}. Now, since β i can take any arbitrary nonnegative value and β (0, 0, 0), (3) is true if and only if the coefficients of β 1, β 2 and β 3 are all positive, namely if and only if (1) holds. The financial interpretation of the inequalities in (1) is as follows. The first states that the call price with the highest strike is positive. The second that the call spread on the last two strikes is also positive. The third that the butterfly built on the given strikes has itself a positive price. In fact, each call option and each call spread have positive prices, as we show in the following. 2

Corollary 1.2. The three option prices C(K 1 ), C(K 2 ) and C(K 3 ) are positive and (strictly) decreasing in the strike: C(K 1 ) > C(K 2 ) > C(K 3 ) (4) Proof. Assume that C(K 1 ) C(K 2 ). Then by the second inequality in (1), we would have ( )C(K 1 ) (K 3 K 1 )C(K 2 ) + (K 2 K 1 )C(K 3 ) ( )C(K 2 ) (K 3 K 1 )C(K 2 ) + (K 2 K 1 )C(K 3 ) = (K 2 K 1 )[C(K 3 ) C(K 2 )] < 0 which contradicts the third inequality in (1). Therefore, C(K 1 ) > C(K 2 ), which also implies the positivity of all prices, since C(K 3 ) > 0. 1.2 Trading Calls, Forward Contracts and the Underlying Asset We now assume we can also trade in forward contracts and in the underlying asset. By the put-call parity, this is equivalent to the possibility of building portfolios based both on European calls and European puts. Proposition 1.3. If we allow for trading also in forward contracts and in the underlying asset, the three option prices C(K 1 ), C(K 2 ) and C(K 3 ) form an arbitrage-free system if and only if C(K 3 ) > 0 C(K 2 ) > C(K 3 ) ( )C(K 1 ) (K 3 K 1 )C(K 2 ) + (K 2 K 1 )C(K 3 ) > 0 (5) K 1 [C(K 2 ) S 0 e qt ] K 2 [C(K 1 ) S 0 e qt ] > 0 C(K 1 ) S 0 e qt + K 1 e rt > 0 Proof. The option prices C(K 1 ), C(K 2 ) and C(K 3 ) form an arbitrage-free system if and only if every linear combination of the related calls with forward contracts and the underlying asset, yielding a nonnegative payoff that is positive with positive probability, has a positive price. Since, for any (real) combinators α 1, α 2, α 3, α 4 and α 5, α 1 +α 2 x+ α 1 + α 2 x x K 1 α i+2 (x K i ) + α 1 + α 2 x + α 3 (x K 1 ) K 1 < x K 2 = α 1 + α 2 x + α 3 (x K 1 ) + α 4 (x K 2 ) K 2 < x K 3 α 1 α 3 K 1 α 4 K 2 α 5 K 3 + (α 2 + α 3 + α 4 + α 5 )x x > K 3 3

we then have α 1 0 α 1 + α 2 K 1 0 α 1 + α 2 S T + α i+2 (S T K i ) + 0 α 1 + α 2 K 2 + α 3 (K 2 K 1 ) 0 α 1 + α 2 K 3 + α 3 (K 3 K 1 ) + α 4 ( ) 0 α 2 + α 3 + α 4 + α 5 0 Setting ᾱ := (α 1, α 2, α 3, α 4, α 5 ) and α 1 0 A := { α 1 + α 2 K 1 0 ᾱ (0, 0, 0, 0, 0) : α 1 + α 2 K 2 + α 3 (K 2 K 1 ) 0 α 1 + α 2 K 3 + α 3 (K 3 K 1 ) + α 4 ( ) 0 α 2 + α 3 + α 4 + α 5 0 } we have therefore an arbitrage-free system if and only if α 1 e rt + α 2 S 0 e qt + α i+2 C(K i ) > 0 for each ᾱ A (6) Defining the new variables vector β = (β 1, β 2, β 3, β 4, β 5 ) as β 1 = α 1 β 2 = α 1 + α 2 K 1 β 3 = α 1 + α 2 K 2 + α 3 (K 2 K 1 ) condition (6) is equivalent to β 4 = α 1 + α 2 K 3 + α 3 (K 3 K 1 ) + α 4 ( ) β 5 = α 2 + α 3 + α 4 + α 5 C(K 1 ) S 0 e qt + K 1 e rt K 1 [C(K 2 ) S 0 e qt ] K 2 [C(K 1 ) S 0 e qt ] β 1 + β 2 K 1 K 1 (K 2 K 1 ) ( )C(K 1 ) (K 3 K 1 )C(K 2 ) + (K 2 K 1 )C(K 3 ) + β 3 ( )(K 2 K 1 ) C(K 2 ) C(K 3 ) + β 4 + β 5 C(K 3 ) > 0 (7) for each β B := { β (0, 0, 0, 0, 0) : β i 0, i = 1,..., 5} Now, since β i can take any arbitrary nonnegative value and β (0, 0, 0, 0, 0), (7) is true if and only if the coefficients of β i, i = 1,..., 5, are all positive, namely if and only if (5) holds. 4

Remark 1.4. By the put-call parity, conditions (5) can be equivalently expressed in terms of put prices as follows: P (K 3 ) + S 0 e qt K 3 e rt > 0 P (K 2 ) K 2 e rt > P (K 3 ) K 3 e rt ( )P (K 1 ) (K 3 K 1 )P (K 2 ) + (K 2 K 1 )P (K 3 ) > 0 (8) K 1 P (K 2 ) K 2 P (K 1 ) > 0 P (K 1 ) > 0 Remark 1.5. From the proof of Proposition 1.3 and the put-call parity, we can infer that any portfolio of calls and puts (equivalently, any portfolio of calls, forward contracts and underlying asset) with nonnegative payoff can be expressed as a nonnegative linear combination of five basic derivatives with nonnegative payoff: 1. A call with strike K 3. Payoff: (S T K 3 ) + 2. A call spread with strikes K 2 and K 3. Payoff: (S T K 2 ) + (S T K 3 ) + 3. A butterfly with strikes K 1, K 2 and K 3. Payoff: ( )(S T K 1 ) + (K 3 K 1 )(S T K 2 ) + + (K 2 K 1 )(S T K 3 ) + 4. A butterfly put spread with strikes K 1 and K 2. Payoff: K 1 (K 2 S T ) + K 2 (K 1 S T ) + 5. A put with strike K 1. Payoff: (K 1 S T ) + We can now better understand why the positivity of the price of any nonnegative payoff (not identically equal to zero) is granted by the price positivity of these five basic derivatives. Corollary 1.6. The three call prices C(K 1 ), C(K 2 ) and C(K 3 ) are positive, (strictly) decreasing in the strike and satisfy the classical no-arbitrage relation ( S0 e qt K i e rt ) + < C(Ki ) < S 0 e qt i = 1, 2, 3 (9) Equivalently, the three put prices P (K 1 ), P (K 2 ) and P (K 3 ) are positive, (strictly) increasing in the strike, and satisfy ( Ki e rt S 0 e qt ) + < P (Ki ) < K i e rt i = 1, 2, 3 Proof. As to the left inequality in (9), by Corollary 1.2 and the last two inequalities in (5), we just have to show that C(K 3 ) S 0 e qt + K 3 e rt > 0. To this end we prove, by contradiction, that C(K 3 ) S 0 e qt + K 3 e rt > C(K 2 ) S 0 e qt + K 2 e rt > 0 5

Assume C(K 3 ) + K 3 e rt would have C(K 2 ) + K 2 e rt. Then by the forth inequality in (5), we ( )C(K 1 ) (K 3 K 1 )C(K 2 ) + (K 2 K 1 )C(K 3 ) ( )C(K 1 ) (K 3 K 1 )C(K 2 ) + (K 2 K 1 )[C(K 2 ) ( ) e rt ] = ( )[C(K 1 ) C(K 2 ) (K 2 K 1 ) e rt ] < 0 which contradicts the third inequality in (5). The right inequality in (9) is proved by considering the portfolio with weights ᾱ = (0, 1, 0, 0, 1). The corresponding β is β = (0, K 1, K 2, K 3, 0) so that we can write S 0 e qt C(K 3 ) =K 1 K 1 [C(K 2 ) S 0 e qt ] K 2 [C(K 1 ) S 0 e qt ] K 1 (K 2 K 1 ) + K 2 ( )C(K 1 ) (K 3 K 1 )C(K 2 ) + (K 2 K 1 )C(K 3 ) ( )(K 2 K 1 ) + K 3 C(K 2 ) C(K 3 ) > 0 The statements on the put prices can be proved in a similar fashion to those of the corresponding calls. 1.3 A general finite system We finally hint at the general case of an options market where an arbitrary, but finite, number N of calls is given. For N = 3, we obviously recover the previously considered case. The quoted strikes are denoted by K i, i = 1,..., N and the associated call prices by C(K i ), i = 1,..., N. Applying the same reasoning of the previous section, we have that the option prices C(K i ) form an arbitrage-free system if and only if every linear combination of the related calls with forward contracts and the underlying asset yielding a nonnegative payoff that is positive with positive probability, has a positive price. Since, for any (real) combinators α i, i = 1,..., N, and j = 1,..., N 1, N α 1 + α 2 x x K 1 α 1 + α 2 x + α i+2 (x K i ) + = α 1 + α 2 x + j α i+2(x K i ) K j < x K j+1 α 1 + α 2 x + N α i+2(x K i ) x > K N we then have α 1 + α 2 S T + α 1 0 N α i+2 (S T K i ) + α 1 + α 2 K 1 0 0 α 1 + α 2 K j+1 + j α i+2(k j+1 K i ) 0 N i=0 α i+2 0 6

Setting ᾱ := (α 1,..., α N ) and α 1 0 A := { α 1 + α 2 K 1 0 ᾱ (0,..., 0) : α 1 + α 2 K j+1 + j α i+2(k j+1 K i ) 0, j = 1,..., N 1 N i=0 α i+2 0 } we have an arbitrage-free system if and only if α 1 e rt + α 2 S 0 e qt + N α i+2 C(K i ) > 0 for each ᾱ A (10) We now define 1 0 0 1 K 1 0 0 1 K 2 K 2 K 1 0 0 A = 1 K 3 K 3 K 1 0 0....... 0 1 1 1 1 1 and the new variables vector β = (β 1,..., β N ) as β = Aᾱ so that β 1 = α 1 β 2 = α 1 + α 2 K 1. β j+2 = α 1 + α 2 K j+1 + j α i+2(k j+1 K i ). β N+2 = N i=0 α i+2 Noting that matrix A is invertible since we then have the following. N 1 det(a) = K 1 (K j+1 K j ) > 0 j=1 Proposition 1.7. The option prices C(K 1 ), C(K 2 ),..., C(K N ) form an arbitrage-free system if and only if va 1 > 0 (11) where v := ( e rt, S 0 e qt, C(K 1 ),..., C(K N ) ) 7

Proof. Condition (10) can be written as By definition of β this is equivalent to vᾱ > 0 for each ᾱ A va 1 β > 0 for each β B := { β (0,..., 0) : β i 0, i = 1,..., N} Since β i can take any arbitrary nonnegative value and β can not be the null vector, this is true if and only if the coefficients of β i, i = 1,..., N, are all positive, namely if and only if (11) holds. References [1] Black, F. and Scholes, M. (1973) The Pricing of Options and Corporate Liabilities. Journal of Political Economy 81, 637-659. [2] Cox, J. and Rubinstein, M. (1985) Options Markets, Prentice-Hall. 8