The Implied Volatility of Forward Starting Options: ATM Short-Time Level, Skew and Curvature Elisa Alòs Antoine Jacquier Jorge A

The Implied Volatility of Forward Starting Options: ATM Short-Time Level, Skew and Curvature Elisa Alòs Antoine Jacquier Jorge A

The Implied Volatility of Forward Starting Options: ATM Short-Time Level, Skew and Curvature Elisa Alòs Antoine Jacquier Jorge A. León May 2017 Barcelona GSE Working Paper Series Working Paper nº 988 The implied volatility of forward starting options: ATM short-time level, skew and curvature Elisa Al`os∗ Antoine Jacquier Dpt. d'Economia i Empresa Department of Mathematics Universitat Pompeu Fabra Imperial College London and Barcelona GSE London SW7 2AZ, UK c/Ramon Trias Fargas, 25-27 08005 Barcelona, Spain Jorge A. Le´ony Control Autom´atico CINVESTAV-IPN Apartado Postal 14-740 07000 M´exico,D.F., Mexico Abstract For stochastic volatility models, we study the short-time behaviour of the at-the-money implied volatility level, skew and curvature for forward- starting options. Our analysis is based on Malliavin Calculus techniques. Keywords: Forward starting options, implied volatility, Malliavin cal- culus, stochastic volatility models JEL code: C02 1 Introduction Consider two moment times s > t. A forward-start call option with maturity T > s allows the holder to receive, at time s and with no additional cost, a call option expirying at T , with strike set equal to KSs, for some K > 0. So, the option life starts at s, but the holder pays at time t the price of the option. Some classical applications of forward starting options include employee stock options and cliquet options, among others (see for example Rubinstein (1991)). Under the Black-Scholes formula, a conditional expectation argument leads to show that the price of a forward starting option is the price of a plain vanilla ∗Supported by grants ECO2014-59885-P and MTM2016-76420-P (MINECO/FEDER, UE). yPartially supported by the CONACyT grant 220303 1 option with time to maturity T −t. In the stochastic volatility case, a change-of- measure links the price of the forward option with the price of a classical vanilla (see, for example, Rubinstein (1991), Musiela and Rutkowski (1997), Wilmott (1998), or Zhang (1998)). For stochastic volatility models, change of numeraire techniques can be applied to obtain a closed-form pricing formula in the context of the Heston model (see Kruse and N¨ogel(2005)). The implied volatility surface for forward starting options exhibits substan- tial differences to the classical vanilla case (see for example Jacquier and Roome (2015)). This paper is devoted to the study of the at-the-money (ATM) short time limit of the implied volatility for forward starting options. More precisely, we will use Malliavin Calculus techniques to compute the ATM short-time limit of the implied volatility level, skew and curvature. In particular, we will see that -contrary to the classical vanilla case- the ATM short-time level depends on the correlation parameter (see Lemma 6 and Theorem 7). We will also prove that the skew depends of the Malliavin derivative of the volatility process in a similar way as for vanilla options, while the curvature (see Theorems 14 and 15) is of order O(T − s). The paper is organized as follows. Section 2 is devoted to introduce forward starting options and the main notation used troghout the paper. In Section 3 we obtain a decomposition of the option price that will allow us to compute, in Sections 4, 5 and 6, the limits for the ATM implied volatility level, skew and curvature. 2 Forward start option We will consider the Heston model for stock price on a time interval [0;T ] under a risk neutral probability P ∗ : ∗ p 2 ∗ dSt =rS ^ tdt + σtSt ρdWt + 1 − ρ Bt ; t 2 [0;T ]; (1) wherer ^ is the instantaneous interest rate (supposed to be constant), W ∗ and B∗ are independent standard Brownian motions defined on a probability space (Ω; F;P ∗) and σ is a positive and square-integrable process adapted to the ∗ ∗ filtration generated by W ∗. In the following we will denote by F W and F B the filtrations generated by W ∗ and B∗, respectively. Moreover we define F := ∗ ∗ F W _F B : It will be convenient in the following sections to make the change of variable Xt = log (St) ; t 2 [0;T ]: Now, we consider a point s 2 [0;T ] : We want to evaluate the following option price: −r^(T −t) ∗ XT α Xs Vt = e Et e − e e + ; (2) ∗ where E denotes the conditional expectation given Ft and α is a real constant. Notice that if t ≥ s this is the payoff of a call option, while in the case t < s this defines a forward start option. We will make use of the following notation 2 • BS(t; x; K; σ) will represent the price of a European call option under the classical Black-Scholes model with constant volatility σ, current log stock price x, time to maturity T − t; strike price K and interest rater: ^ Remember that in this case x −r^(T −t) BS(t; x; K; σ) = e N(d+) − e KN(d−); where N is the cumulative probability function of the standard normal law and x − ln K +r ^(T − t) σ p d± := p ± T − t: σ T − t 2 •L BS (σ) will denote the Black-Scholes differential operator (in the log vari- able) with volatility σ : @ 1 @2 1 @ L (σ) = + σ2 + r^ − σ2 − r:^ BS @t 2 @x2 2 @x It is well known that LBS (σ) BS (·; ·; K; σ) = 0: 2 • G(t; x; K; σ) := (@xx − @x)BS(t; x; K; σ): 2 • H(t; x; K; σ) := @x(@xx − @x)BS(t; x; K; σ): • BS−1(a) := BS−1 (t; x; K; a) denotes the inverse of BS as a function of the volatility parameter. • α∗ :=r ^(T − s) We recall the following result, that can be proved following the same argu- ments as in that of Lemma 4.1 in Al`os,Le´onand Vives (2007) W Lemma 1 Let 0 ≤ t ≤ s; u < T and Gt := Ft _FT : Then for every n ≥ 0; there exists C = C(n; ρ) such that a) If u ≤ s − 1 (n+1) Z T ! 2 n Xu 2 jE (@x G (u; Xu;Mu; vu)j Gt)j ≤ CE e Gt σs ds : s b) If u ≥ s − 1 (n+1) Z T ! 2 n Xs 2 jE (@x G (u; Xu;Mu; vu)j Gt)j ≤ CE e Gt σs ds : s 3 3 A decomposition formula for forward option prices We will stand for ∗ α Xs • Mt := Et e e , t 2 [0;T ]. We observe that Z t α rs^ X0 α r^(s−u) Xu ∗ p 2 ∗ Mt = e e e + σu1[0;s](u)e e e ρdWu + 1 − ρ dBu 0 Z t^s α r^(s−u) Xu ∗ p 2 ∗ = M0 + e σue e ρdWu + 1 − ρ dBu : 0 1 2 Yt • vt := T −t ; with Z T Z T 2 2 Yt := σu1[s;T ](u)du = σudu t t_s p p Notice that, if t < s, vt T − t = vs T − s: We will need the following hypotheses: (H1) There exist two constants 0 < c < C such that c < σt < C, for all t2 [0;T ], with probability one. (H2) σ; σeX 2 L2;2 \ L1;4: Now we are in a position to prove the following theorem, that allows us to identify the impact of correlation in the forward option price Theorem 2 Consider the model (1) and assume that hypotheses (H1) and (H2) hold. Then, for all 0 ≤ t ≤ s ≤ T , ∗ α Vt = Et exp(Xt)BS (s; 0; e ; vs) Z T ρ −r^(u−t) W ∗ + e H(u; Xu;Mu; vu)σuΛu du 2 s s ρ Z ∗ α −r^(u−t) Xu W + G(s; 0; e ; vs) e e σuΛu du : 2 t W ∗ R T W ∗ 2 where Λu := u_s Du σθ dθ: ∗ α Xs Proof. Remember that Mt := Et e e and notice that −r^(T −t) ∗ XT α Xs −r^(T −t) ∗ XT Vt = e Et e − e e + = e Et e − MT + −r^(T −t) = e BS(T;XT ;MT ; νT ): 4 Therefore, the anticipating It^o'sformula for the Skorohod integral (see for ex- ample Nualart (2006)) allows us to write ∗ −rT^ Et e BS(T;XT ;MT ; vT ) " Z T ∗ −rt^ −ru^ = Et e BS(t; Xt;Mt; vt) − r^ e BS(u; Xu;Mu; vu)du t Z T −ru^ @BS + e (u; Xu;Mu; vu)du t @u Z T 2 −ru^ @BS σu + e (u; Xu;Mu; vu) r^ − du t @x 2 Z T 2 1 −ru^ @ BS 2 + e 2 (u; Xu;Mu; vu)σudu 2 t @x Z T 2 −ru^ @ BS + e (u; Xu;Mu; vu)d hX; Miu t @x@K Z T 2 1 −ru^ @ BS + e 2 (u; Xu;Mu; vu)d hM; Miu 2 t @K Z T 2 1 −ru^ @ @ 2 2 + e 2 − BS(u; Xu;Mu; vu) vu − σu1]s;T ](u) du 2 t @x @x Z T 2 1 −ru^ @ @ @ W ∗ + e 2 − BS(u; Xu;Mu; vu)σuρΛu du 2 t @x @x @x T 2 # 1 Z @ @ @ ∗ −ru^ α r^(s−u) Xu W + e 2 − BS(u; Xu;Mu; vu)σue e e ρΛu 1[0;s](u)du : 2 t @K @x @x 5 That is, using t ≤ s, ∗ Vt = Et BS(t; Xt;Mt; vt) Z T −r^(u−t) + e LBS (vu) BS(u; Xu;Mu; vu)du t Z T 2 1 −r^(u−t) @ @ 2 2 + e 2 − BS(u; Xu;Mu; vu) σu − vu du 2 t @x @x Z T 2 −r^(u−t) @ BS + e (u; Xu;Mu; vu)d hX; Miu t @x@K Z T 2 1 −r^(u−t) @ BS + e 2 (u; Xu;Mu; vu)d hM; Miu 2 t @K Z T 2 1 −r^(u−t) @ @ 2 2 + e 2 − BS(u; Xu;Mu; vu) vu − σu1]s;T ](u) du 2 t @x @x Z T 2 1 −r^(u−t) @ @ @ W ∗ + e 2 − BS(u; Xu;Mu; vu)σuρΛu du 2 t @x @x @x s 2 1 Z @ @ @ ∗ −r^(u−t) α r^(s−u) Xu W + e 2 − BS(u; Xu;Mu; vu)σue e e ρΛu du : 2 t @K @x @x Thus, we get, for t ≤ s, ∗ Vt = Et BS(t; Xt;Mt; vt) Z T −r^(u−t) + e LBS (vu) BS(u; Xu;Mu; vu)du t Z T 2 1 −r^(u−t) @ @ 2 2 + e 2 − BS(u; Xu;Mu; vu) σu − σu1]s;T ](u) du 2 t @x @x Z T 2 −r^(u−t) @ BS + e (u; Xu;Mu; vu)d hX; Miu t @x@K Z T 2 1 −r^(u−t) @ BS + e 2 (u; Xu;Mu; vu)d hM; Miu 2 t @K Z T 2 1 −r^(u−t) @ @ @ W ∗ + e 2 − BS(u; Xu;Mu; vu)σuρΛu du 2 t @x @x @x s 2 1 Z @ @ @ ∗ −r^(u−t) α r^(s−u) Xu W + e 2 − BS(u; Xu;Mu; vu)σue e e ρΛu du : 2 t @K @x @x Now, taking into account the facts that LBS (vu)(BS)(u; Xu;Mu; vu) = 0; 2 α r^(s−u) Xu d hM; Xiu = σue e e 1[0;s](u)du; 6 2 2α 2^r(s−u) 2Xu d hM; Miu = σue e e 1[0;s](u)du; @2BS 1 @2BS @BS (t; x; K; σ) = − − (t; x; K; σ) @x@K K @x2 @x and @2BS 1 @2BS @BS (t; x; K; σ) = − (t; x; K; σ) ; @K2 K2 @x2 @x it follows that ∗ Vt = Et BS(t; Xt;Mt; vt) Z T 2 1 −r^(u−t) @ @ @ W ∗ + e 2 − BS(u; Xu;Mu; vu)σuρΛu du 2 t @x @x @x s 2 1 Z @ @ @ ∗ −r^(u−t) α r^(s−u) Xu W + e 2 − BS(u; Xu;Mu; vu)σue e e ρΛu du 2 t @K @x @x ∗ = Et BS(t; Xt;Mt; vt) Z T 2 1 −r^(u−t) @ @ @ W ∗ + e 2 − BS(u; Xu;Mu; vu)σuρΛu du 2 s @x @x @x Z s 2 1 −r^(u−t) @ @ @ W ∗ + e 2 − BS(u; Xu;Mu; vu)σuρΛu du 2 t @x @x @x Z s 2 1 −r^(u−t) @ @ @ W ∗ + e 2 − BS(u; Xu;Mu; vu)σuMuρΛu du : 2 t @K @x @x Since we have @ @2BS @BS exN 0(d ) d − (t; x; k; σ) = p + 1 − p + @x @x2 @x σ T − t σ T − t and @ @2BS @BS exN 0(d ) d − (t; x; k; σ) = p + p + ; @K @x2 @x Kσ T − t σ T − t then it is easy to see that ∗ Vt = Et BS(t; Xt;Mt; vt) Z T ρ −r^(u−t) W ∗ + e H(u; Xu;Mu; vu)σuΛu du 2 s Z s ρ −r^(u−t) W ∗ + e G(u; Xu;Mu; vu)σuΛu du : 2 t 7 Hence, due to BS(t; Xt;Mt; vt) p −α +r ^(T − s) vs T − s = exp(Xt)N p + vs T − s 2 p α −r^(T −s) −α +r ^(T − s) vs T − s −e exp(Xt)e N p − vs T − s 2 α = exp(Xt)BS (s; 0; e ; vs) and, for all u < s, p −α+^r(T −s) eXu N 0 p + vs T −s vs T −s 2 G(u; Xu;Mu; vu) = p vs T − s Xu α = e G(s; 0; e ; vs); the proof is complete.

View Full Text

Details

  • File Type
    pdf
  • Upload Time
    -
  • Content Languages
    English
  • Upload User
    Anonymous/Not logged-in
  • File Pages
    21 Page
  • File Size
    -

Download

Channel Download Status
Express Download Enable

Copyright

We respect the copyrights and intellectual property rights of all users. All uploaded documents are either original works of the uploader or authorized works of the rightful owners.

  • Not to be reproduced or distributed without explicit permission.
  • Not used for commercial purposes outside of approved use cases.
  • Not used to infringe on the rights of the original creators.
  • If you believe any content infringes your copyright, please contact us immediately.

Support

For help with questions, suggestions, or problems, please contact us