Initial Term Structure Research Articles

The VIX Future in Bergomi Models: Fast Approximation Formulas and Joint Calibration with S&P 500 Skew

Using the exponential generating function of Hermite polynomials, we expand the prices of VIX power payoffs (including VIX futures) in various Bergomi models at any order in the volatility-of-volatility. We introduce the notion of volatility of the VIX squared implied by a VIX power payoff, which we call “implied VIX$^2$ volatility” and also expand at any order. We cover the one-factor and (skewed) two-factor Bergomi models and allow for maturity-dependent and/or time-dependent parameters. When the initial term-structure of variance swaps is flat, we provide the expansions up to order 8 in closed form; otherwise, they simply involve one-dimensional integrals, which are extremely fast to compute. Extensive numerical experiments show that the implied volatility expansion converges much faster than the price expansion and is extremely accurate for a wide range of model parameters, including typical market calibrating parameters with very large volatilities-of-volatility. It leads to new, simple approximation formulas for the price of a VIX power payoff that shed light on how those prices depend on model parameters. We combine the new expansion and the Bergomi--Guyon expansion of the vanilla smile [L. Bergomi and J. Guyon, Risk, May (2012), pp. 60--66] to calibrate the two-factor Bergomi model jointly to the term-structures of S&P 500 at-the-money skew and VIX futures. Very interestingly, the joint fit selects (1) much larger values of volatility-of-volatility and mean reversion than those previously reported in [L. Bergomi, Risk, October (2005), pp. 67--73] and [L. Bergomi, Stochastic Volatility Modeling, CRC Press, 2016], and (2) fully correlated Brownian motions, thus producing a (Markovian) pure path-dependent volatility model with rough-like paths.

The VIX Future in Bergomi Models: Analytic Expansions and Joint Calibration with S&P 500 Skew

We derive the formal expansion of the price of a VIX future (and more generally a VIX power payoff) in various Bergomi models at any order in powers of the volatility-of-volatility. We introduce the notion of volatility of the VIX squared implied by the VIX future, which we call implied volatility, and also expand this quantity at any order. We cover the one-factor, two-factor, and skewed two-factor Bergomi models and allow for maturity-dependent and/or time-dependent parameters. When the initial term-structure of variance swaps is flat, the expansions are in closed form; otherwise, they involve one-dimensional integrals which are extremely fast to compute. Our extensive numerical experiments show that both the price and implied volatility expansions provide very accurate approximations for a wide range of model parameters which covers the values typically used in the equity and FX derivatives markets and even well beyond---in particular, the expansions are accurate even for very large values of the volatility-of-volatility. They also show that the implied volatility expansion converges much faster than the price expansion. Its truncation at order 1 is already so accurate that it leads to a simple formula for the price of the VIX future that can virtually be considered exact. The derivation of the expansions naturally involves the (classical or dual bivariate) Hermite polynomials and exploits their orthogonality properties; it is interesting in itself. The expansions allow us to precisely pinpoint the roles of all the model parameters (volatility-of-volatility, mean reversions, correlations, mixing fraction) in the formation of the prices of VIX futures and VIX power payoffs in Bergomi models. We also use them together with the Bergomi-Guyon expansion of the S&P 500 smile to (instantaneously) calibrate the two-factor Bergomi model jointly to the term-structures of S&P 500 at-the-money skew and VIX2 implied volatility. Our tests and the new expansions shed more light on the inability of traditional stochastic volatility models to jointly fit S&P 500 and VIX market data. The (imperfect but decent) joint fit requires much larger values of volatility-of-volatility and fast mean reversion than the ones previously reported in Bergomi (2005, 2016).

Pure jump models for pricing and hedging VIX derivatives

Recent non-parametric statistical analysis of high-frequency VIX data (Todorov and Tauchen, 2011) reveals that VIX dynamics is a pure jump semimartingale with infinite jump activity and infinite variation. To our best knowledge, existing models in the literature for pricing and hedging VIX derivatives do not have these features. This paper fills this gap by developing a novel class of parsimonious pure jump models with such features for VIX based on the additive time change technique proposed in Li et al. (2016a, 2016b). We time change the 3/2 diffusion by a class of additive subordinators with infinite activity, yielding pure jump Markov semimartingales with infinite activity and infinite variation. These processes have time and state dependent jumps that are mean reverting and are able to capture stylized features of VIX. Our models take the initial term structure of VIX futures as input and are analytically tractable for pricing VIX futures and European options via eigenfunction expansions. Through calibration exercises, we show that our model is able to achieve excellent fit for the VIX implied volatility surface which typically exhibits very steep skews. Comparison to two other models in terms of calibration reveals that our model performs better both in-sample and out-of-sample. We explain the ability of our model to fit the volatility surface by evaluating the matching of moments implied from market VIX option prices. To hedge VIX options, we develop a dynamic strategy which minimizes instantaneous jump risk at each rebalancing time while controlling transaction cost. Its effectiveness is demonstrated through a simulation study on hedging Bermudan style VIX options.

JOINING THE HESTON AND A THREE-FACTOR SHORT RATE MODEL: A CLOSED-FORM APPROACH

In this paper, we present an innovative hybrid model for the valuation of equity options. Our approach includes stochastic volatility according to Heston (1993) [Review of Financial Studies 6 (2), 327–343] and features a stochastic interest rate that follows a three-factor short rate model based on Hull and White (1994) [Journal of Derivatives 2 (2), 37–48]. Our model is of affine structure, allows for correlations between the stock, the short rate and the volatility processes and can be fitted perfectly to the initial term structure. We determine the zero bond price formula and derive the analytic solution for European type options in terms of characteristic functions needed for fast calibration. We highlight the flexibility of our approach, by comparing the price and implied volatility surfaces of our model with the Heston model, where we in particular focus on the correlation structure. Our approach forms a comprehensive market model with an intuitive correlation structure that connects both the equity and interest market to the market volatility.

HEAT KERNEL MODELS FOR ASSET PRICING

A heat kernel approach is proposed for the development of a novel method for asset pricing over a finite time horizon. We work in an incomplete market setting and assume the existence of a pricing kernel that determines the prices of financial instruments. The pricing kernel is modeled by a weighted heat kernel driven by a multivariate Markov process. The heat kernel is chosen so as to provide enough freedom to ensure that the resulting model can be calibrated to appropriate data, e.g. to the initial term structure of bond prices. A class of models is presented for which the prices of bonds, caplets, and swaptions can be computed in closed form. The dynamical equations for the price processes are derived, and explicit formulae are obtained for the short rate of interest, the risk premium, and for the stochastic volatility of prices. Several of the closed-form models presented are driven by combinations of Markovian jump processes with different probability laws. Such models provide a basis for consistent applications in various market sectors, including equity markets, fixed-income markets, commodity markets, and insurance. The flexible multidimensional and multivariate structure on which the resulting price models are based lends itself well to the modeling of dependence across asset classes. As an illustration, the impact of spiraling debt, a typical feature of a financial crisis, is modeled explicitly, and the contagion effects can be readily observed in the dynamics of the associated asset returns.

A generalized procedure for building trees for the short rate and its application to determining market implied volatility functions

One-factor no-arbitrage models of the short rate are important tools for valuing interest rate derivatives. Trees are often used to implement the models and fit them to the initial term structure. This paper generalizes existing tree building procedures so that a very wide range of interest rate models can be accommodated. It shows how a piecewise linear volatility function can be calibrated to market data and, using market data from days during the period 2004–2013, finds a best fit to cap prices.

Joining the Heston and a Three-Factor Hull-White Model: A Closed-Form Approach

In this article, we present an innovative hybrid model for the valuation of equity options. Our approach includes stochastic volatility according to Heston (1993) and features a stochastic interest rate that follows a three-factor short rate model based on Hull-White (1994). Our model is of affine structure, allows for correlations between the stock, the short rate and the volatility processes and can be fitted perfectly to the initial term structure. We determine the zero bond price formula and derive the analytic solution for European type options in terms of characteristic functions needed for fast calibration. We highlight the flexibility of our approach, by comparing the price and implied volatility surfaces of our model with the Heston model, where we in particular focus on the correlation structure. Our approach forms a comprehensive market model with an intuitive correlation structure that connects both the equity and interest market to the market volatility.

A Flexible Spot Multiple-Curve Model

We propose a model for the instantaneous risk-free spot rate and for the spot LIBOR, driven by a time-homogeneous Markovian process. We introduce deterministic time-shifts in order to match any initial term-structure. By doing so, the model automatically becomes an exogenous term-structure model, in the spirit of Brigo and Mercurio (2001) who proposed this approach in the single curve case. A calibration exercise based on real data illustrates the flexibility of our approach for some typical speci fications used in the literature and in the bank industry.

A Generalized Procedure for Building Trees for the Short Rate

One-factor no-arbitrage models of the short rate are important tools for valuing interest rate derivatives. Trees are often used to implement the models and fit them to the initial term structure. This paper generalizes existing tree building procedures so that a very wide range of interest rate models can be accommodated. It shows how a piecewise linear volatility function can be calibrated to market data and, using market data from days during the period 2004 to 2013, finds that the best fit to cap prices is provided by a function remarkably similar to that estimated by Deguillaume et al (2013) from historical data.

Calibrating Drift of Short-Rate Models with Jump-Diffusion

The drift or the mean-reversion level of short-rate models under jump-diffusion is derived to fit the initial term-structure of zero-coupon bond. In particular, the drift is obtained for Hull-White and Cox-Ingersoll-Ross short-rate models. The purpose of obtaining the drift is for the Monte-Carlo simulations for pricing exotic equity options under stochastic interest rates and exotic interest rate derivatives. The theoretical expressions for the drift are verified by Monte-Carlo simulations of the short-rate and bond paths, and show that the expectation of the discounted entities reproduce the initial zero-coupon bond term-structure. For the Cox-Ingersoll-Ross model the drift is obtained by a numerical computation of a second-order ordinary differential equation using a perturbation method. The correctness of this method is verified by the Monte-Carlo results. Further, the bond partial differential equation (PDE) for the Cox-Ingersoll-Ross jump-diffusion short-rate model is also solved numerically and the results reproduce the initial zero-coupon bond term-structure. This bond PDE computation verifies the self-consistency of the drift computation, and also validates the numerical boundary conditions used in the finite-difference computations of this final-value problem. Similar results are presented from the finite-difference computation of the Hull-White PDE also.

On the Reconciliation of the Extended Nelson-Siegel and the Extended Vasicek Models (with a View Towards Swap and Swaption Valuation)

Extended Nelson-Siegel models are widely used by e.g. practitioners and central banks to estimate current term structures of riskless zero-coupon interest rates, whereas other models such as the extended Vasicek model (a.k.a. the Hull-White model) are popular for pricing interest rate derivatives. This paper establishes theoretical consistency between these two types of models by showing how to specify the extended Vasicek model such that its implied initial term structure curve precisely matches a given extended Nelson-Siegel specification. That is, we show how to reconcile the two classes of models at the initial time point (and here only). A case study and some illustrative numerical examples involving swaps and swaption pricing are provided at the end of the paper.

Clean Valuation Framework for the USD Silo - An Implication for the Forthcoming Standard Credit Support Annex (SCSA)

In the forthcoming ISDA Standard Credit Support Annex (SCSA), the trades denominated in non-G5 currencies as well as those include multiple currencies are expected to be allocated to the USD silo, where the contracts are collateralized by USD cash, or a different currency with an appropriate interest rate overlay to achieve the same economic effects. In this paper, we have presented a simple generic valuation framework for the clean price under the USD silo with the the detailed procedures for the initial term structure construction. We have also shown that Cross Currency Swap (CCS) basis spread can be expressed as a difference between two swap rates.

Pricing Asian-Style Interest Rate Swaps within a Multi-factor Gaussian HJM Framework

Considering the Asian-style interest rate swap is a stream of interest payments between two parties, this study employs the multi-factor Gaussian Heath-Jarrow-Morton term structure model for developing an analytic evaluation model applied to Asian-style interest rate swaps. The numerical investigation illustrates that the pricing properties of Asian-style interest rate swaps coincide with conventional financial theorems such as the initial term structure of interest rate and, the length of reset periods. But the volatilities of interest rates do not generate significant influence. Finally, Asian-style interest rate swaps are not necessarily less costly than conventional interest rate swaps.

Semi-Markov regime switching interest rate models and minimal entropy measure

In this paper, we present a discrete time regime switching binomial-like model of the term structure where the regime switches are governed by a discrete time semi-Markov process. We model the evolution of the prices of zero-coupon when given an initial term structure as in the model by Ho and Lee that we aim to extend. We discuss and derive conditions for the model to be arbitrage free and relate this to the notion of martingale measure. We explicitly show that due to the extra source of uncertainty coming from the underlying semi-Markov process, there are an infinite number of equivalent martingale measures. The notion of path independence is also studied in some detail, especially in the presence of regime switches. We deal with the market incompleteness by giving an explicit characterization of the minimal entropy martingale measure. We give an application to the pricing of a European bond option both in a Markov and semi-Markov framework. Finally, we draw some conclusions.

An artificial boundary method for the Hull–White model of American interest rate derivatives

The Hull–White (HW) model is a widely used one-factor interest rate model because of its analytical tractability on liquidly traded derivatives, super-calibration ability to the initial term structure and elegant tree-building procedure. As an explicit finite difference scheme, lattice method is subject to some stability criteria, which may deteriorate the computational efficiency for early exercisable derivatives. This paper proposes an artificial boundary method based on the partial differential equations (PDEs) to price interest rate derivatives with early exercise (American) feature under the HW model. We construct conversion factors to extract the market information from the zero-coupon curve and then reduce the infinite computational domain into a finite one by using an artificial boundary on which an exact boundary condition is derived. We then develop an implicit θ-scheme with unconditional stability to solve the PDE in the reduced bounded domain. With a finite computational domain, the optimal exercise strategy can be determined efficiently. Our numerical examples show that the proposed scheme is accurate, robust to the truncation size, and more efficient than the popular lattice method for accurate derivative prices. In addition, the singularity-separating technique is incorporated into the artificial boundary method to enhance accuracy and flexibility of the numerical scheme.

A New Taxonomy of the Dynamic Term Structure Models

This paper gives a new taxonomy of dynamic term structure models that classifies all existing TSMs as either fundamental models or preference-free singleplus, double-plus, and triple-plus models. We exemplify the new taxonomy by considering preference-free versions of some well-known fundamental short rate models. Single-plus extensions of the fundamental models are shown to be both timehomogeneous and preference-free - two characteristics which do not simultaneously hold under any existing class of TSMs. Though the analytical apparatus for pricing fixed income securities is identical under fundamental models and single-plus models, the latter models are consistent with general non-linear forms of MPRs which may also depend upon an arbitrary set of state variables, leading to better estimates of riskneutral parameters. The preference-free double-plus and triple-plus extensions of the fundamental models are similar to the Heath, Jarrow, and Morton [1992] models, in that time-inhomogeneous drifts and volatilities are used as “smoothing variables” to fit the initial bond prices and initial term structure of volatilities, respectively.

Pricing caps and floors with the extended CIR model

AbstractWe use the extended CIR model to value interest rate caps and floors. The extension allows arbitrary initial term structure, in the spirit of Hull and White, a crucial feature since we show that the pricing of the considered contingent claims improves dramatically after taking into account the big stake of market information contained in the yield curve. We compute model prices of at the money caps using only yield curve data, and we compare prices with those obtained when other well‐known short‐rate models are used, including the extended Vasicek model. With respect to natural benchmarks, we find a significant decrease in the pricing error for longer maturity caps when our model is used. These results witness that when a better specification of the dynamics of spot rate is provided, a satisfactory pricing of caps and floors is possible using only the information available in the bond market. Copyright © 2008 John Wiley & Sons, Ltd.

A multi-factor jump-diffusion model for commodities†

In this paper we develop an arbitrage-free model for the pricing of commodity derivatives. The model generates futures (or forward) commodity prices consistent with any initial term structure. The model is consistent with mean reversion in commodity prices and also generates stochastic convenience yields. Our model is a multi-factor jump-diffusion model, one specification of which allows the prices of long-dated futures contracts to jump by smaller magnitudes than short-dated futures contracts, which, to our knowledge, is a feature that has not previously appeared in the literature, in spite of it being in line with stylised empirical observations (especially for energy-related commodities). Our model also allows for stochastic interest rates. The model produces semi-analytic solutions for standard European options, which enable option prices to be evaluated in typically about 1/50th of a second (depending upon parameter values and the required accuracy). This opens the possibility to calibrate the model parameters by deriving implied parameters from the market prices of options. We perform such a calibration on crude oil options and show that, allowing long-dated futures contracts to jump by smaller magnitudes than short-dated contracts, gives a greatly enhanced fit. †This is a condensed and enhanced version of a previous working paper (Crosby 2005) entitled “Commodities: a simple multi-factor jump-diffusion model”.

TERM STRUCTURE OF VANILLA OPTIONS

Every maturity-dependent derivative contract entails a term structure. For example, when the value of the portfolio consisting of a long position in a stock and a short position in a vanilla option is expressed in units of its instantaneous exercise value, the resulting quantity defines a discount function. Thus, the derivative of the discount function with respect to the time left until maturity defines a term structure density function, and the "hazard rate" associated with the discount function determines the forward rates for the vanilla option portfolio. The dynamics associated with these quantities are obtained in the complete market setting. In particular, one can model vanilla options based on the associated forward rates. The formulation based on forward rates for options extends the approach based on modeling the implied volatility process. As an illustrative example, the initial term structure of the Black–Scholes model is considered. It is shown in this example that the implied volatility smile has the effect of making the option forward rates homogeneous across different strikes.

Arbitrary Initial Term Structure within the CIR Model: A Perturbative Solution

Single factor interest rate models with constant coefficients are not consistent with arbitrary initial term structures. An extension which allows both arbitrary initial term structure and analytical tractability has been provided only in the Gaussian case. In this paper, within the context of the HJM methodology, we provide an extension of the CIR model which admits arbitrary initial term structure. We show how to calculate bond prices via a perturbative approach, and we provide closed formulas at every order. Since the parameter we select for the expansion is typically estimated to be small, the perturbative approach turns out to be adequate to our purpose. Using results on affine models, we estimate the extended CIR model via maximum likelihood on a time series of daily interest rate yields. Our results show that the CIR model has to be rejected with respect to the proposed extension, and point out that the extended CIR model provides a more flexible characterization of the link between risk neutral and natural probability.