 SVJJProcess - Maple Help

Finance

 SVJJProcess
 create new SVJJ process Calling Sequence SVJJProcess(${S}_{0}$, ${V}_{0}$, r, theta, kappa, sigma, rho, lambda, alpha, beta, delta, t) Parameters

 ${S}_{0}$ - real constant; initial value of the return process ${V}_{0}$ - non-negative constant; initial value of the variance r - real constant; risk-neutral drift theta - non-negative constant, algebraic expression or procedure; long-run mean of the volatility kappa - positive constant; speed of mean reversion sigma - real constant; volatility of the variance process rho - non-negative constant; instantaneous correlation between the return process and the variance process lambda - non-negative constant; jump intensity alpha - non-negative constant; mean relative jump size beta - real constant; standard deviation of the relative jump size delta - real constant; jump size of the variance process t - name; time variable Description

 • The SVJJProcess command creates a new stochastic volatility process with jumps (SVJJ). This is a process governed by the stochastic differential equation (SDE)

$\frac{\mathrm{dS}\left(t\right)}{S\left(t\right)}=\left(-\mathrm{\lambda }\mathrm{\mu }+r\right)\mathrm{dt}+\sqrt{V\left(t\right)}{\mathrm{dW}}_{1}\left(t\right)+\left(J-1\right)\mathrm{dN}\left(t\right)$

$\mathrm{dV}\left(t\right)=\mathrm{\kappa }\left(\mathrm{\theta }-V\left(t\right)\right)\mathrm{dt}+\mathrm{\sigma }\sqrt{V\left(t\right)}{\mathrm{dW}}_{2}\left(t\right)+\mathrm{\delta }\mathrm{dN}\left(t\right)$

 where
 – $r$ is the risk-neutral drift,
 – $\mathrm{\theta }$ is the long-run mean of the variance process,
 – $\mathrm{\kappa }$ is the speed of mean reversion of the variance process,
 – $\mathrm{\sigma }$ is the volatility of the variance process,
 – $\mathrm{\delta }$ is the volatility jump size,
 and
 – $W\left(t\right)$ is the two-dimensional Wiener process with instantaneous correlation $\mathrm{\rho }$,
 – $N\left(t\right)$ is a Poisson process, independent of $W\left(t\right)$, with constant intensity $\mathrm{\lambda }$,
 – $J$ is a lognormal random variable with mean $\mathrm{\alpha }$ and variance ${\mathrm{\beta }}^{2}$.
 • The parameters $\mathrm{\mu }$, $\mathrm{\alpha }$, and $\mathrm{\beta }$ are related by the following equation

$\mathrm{ln}\left(1+\mathrm{\mu }\right)=\mathrm{\alpha }+\frac{{\mathrm{\beta }}^{2}}{2}$

 • This process was introduced by A. Matytsin. Special cases of this process include

 Bates SVJ process $\mathrm{\delta }=0$ Heston SV process $\mathrm{\lambda }=0$ Examples

 > $\mathrm{with}\left(\mathrm{Finance}\right):$

First construct an SVJJ process with variable parameters. You will assign numeric values to these parameters later.

 > $Y≔\mathrm{SVJJProcess}\left(100,0.008836,r,\mathrm{\theta },\mathrm{\kappa },\mathrm{\sigma },\mathrm{\rho },\mathrm{\lambda },\mathrm{\alpha },\mathrm{\beta },\mathrm{\delta },t\right):$
 > $\mathrm{\kappa }≔3.99$
 ${\mathrm{\kappa }}{≔}{3.99}$ (1)
 > $\mathrm{\theta }≔0.014$
 ${\mathrm{\theta }}{≔}{0.014}$ (2)
 > $\mathrm{\sigma }≔0.27$
 ${\mathrm{\sigma }}{≔}{0.27}$ (3)
 > $\mathrm{\rho }≔-0.79$
 ${\mathrm{\rho }}{≔}{-0.79}$ (4)
 > $r≔0.0319$
 ${r}{≔}{0.0319}$ (5)
 > $\mathrm{\alpha }≔0.1$
 ${\mathrm{\alpha }}{≔}{0.1}$ (6)
 > $\mathrm{\beta }≔0.15$
 ${\mathrm{\beta }}{≔}{0.15}$ (7)
 > $\mathrm{\lambda }≔0.11$
 ${\mathrm{\lambda }}{≔}{0.11}$ (8)
 > $T≔5.0$
 ${T}{≔}{5.0}$ (9)
 > $K≔100$
 ${K}{≔}{100}$ (10)
 > $\mathrm{\delta }≔0.1$
 ${\mathrm{\delta }}{≔}{0.1}$ (11)
 > $M≔100;$$N≔{10}^{4}$
 ${M}{≔}{100}$
 ${N}{≔}{10000}$ (12)

Generate 10 replications of the sample path and plot sample paths for the state variable and the variance process.

 > $A≔\mathrm{SamplePath}\left(Y\left(t\right),t=0..T,\mathrm{timesteps}=30,\mathrm{replications}=10\right)$
 ${A}{≔}\begin{array}{c}\left[\begin{array}{cc}{100.}& {97.6121587078836}\\ {103.614146543789}& {98.3718062669329}\\ {104.208032435127}& {97.2231274901719}\\ {101.542983778051}& {99.8913190983984}\\ {108.572564213206}& {100.550915178494}\\ {99.0499056108087}& {101.676712784682}\\ {101.756087731456}& {104.519571337341}\\ {97.8019929951040}& {102.097337518808}\\ {97.4320763045216}& {99.8232007432970}\\ {98.4877523541519}& {101.028527200751}\end{array}\right]\\ \hfill {\text{slice of 10 × 2 × 31 Array}}\end{array}$ (13)
 > $\mathrm{PathPlot}\left(A,1,\mathrm{thickness}=3,\mathrm{color}=\mathrm{red}..\mathrm{blue},\mathrm{axes}=\mathrm{BOXED},\mathrm{gridlines}=\mathrm{true}\right)$ > $\mathrm{PathPlot}\left(A,2,\mathrm{thickness}=3,\mathrm{color}=\mathrm{red}..\mathrm{blue},\mathrm{axes}=\mathrm{BOXED},\mathrm{gridlines}=\mathrm{true}\right)$ > $\mathrm{exp}\left(-rT\right)\mathrm{ExpectedValue}\left(\mathrm{max}\left(Y\left(T\right)\left[1\right]-K,0\right),\mathrm{timesteps}=M,\mathrm{replications}=N,\mathrm{output}=\mathrm{value}\right)$
 ${19.96717514}$ (14)

Consider different parameters.

 > $\mathrm{\kappa }≔0.$
 ${\mathrm{\kappa }}{≔}{0.}$ (15)
 > $\mathrm{\theta }≔0.$
 ${\mathrm{\theta }}{≔}{0.}$ (16)
 > $\mathrm{\lambda }≔1.0$
 ${\mathrm{\lambda }}{≔}{1.0}$ (17)
 > $\mathrm{\sigma }≔0.$
 ${\mathrm{\sigma }}{≔}{0.}$ (18)

Generate 10 replications of the sample path of the new process and plot sample paths for the state variable and the variance process.

 > $A≔\mathrm{SamplePath}\left(Y\left(t\right),t=0..T,\mathrm{timesteps}=30,\mathrm{replications}=10\right)$
 ${A}{≔}\begin{array}{c}\left[\begin{array}{cc}{100.}& {94.8349949005841}\\ {92.0096388274855}& {134.574895263471}\\ {91.7809841586387}& {159.155707455031}\\ {90.8677193827435}& {46.6931634950871}\\ {81.1739938276965}& {48.8135013308782}\\ {72.5772503707256}& {39.4491818726571}\\ {81.9657722173052}& {38.5398562531824}\\ {80.1673331494771}& {37.0013320960280}\\ {85.2130952991886}& {37.3946383621617}\\ {90.9891954127822}& {52.6943048829217}\end{array}\right]\\ \hfill {\text{slice of 10 × 2 × 31 Array}}\end{array}$ (19)
 > $\mathrm{PathPlot}\left(A,1,\mathrm{thickness}=3,\mathrm{color}=\mathrm{red}..\mathrm{blue},\mathrm{axes}=\mathrm{BOXED},\mathrm{gridlines}=\mathrm{true}\right)$ > $\mathrm{PathPlot}\left(A,2,\mathrm{thickness}=3,\mathrm{color}=\mathrm{red}..\mathrm{blue},\mathrm{axes}=\mathrm{BOXED},\mathrm{gridlines}=\mathrm{true}\right)$  References

 Bates, D., Jumps and stochastic volatility: the exchange rate processes implicit in Deutsche Mark options, Review of Financial Studies, Volume 9, 69-107, 1996.
 Duffie, D., Pan, J., and Singleton, K.J. Transform analysis and asset pricing for affine jump-diffusions. Econometrica, Volume 68, 1343-1376, 2000.
 Glasserman, P., Monte Carlo Methods in Financial Engineering. New York: Springer-Verlag, 2004.
 Matytsin, A. Modelling volatility and volatility derivatives, Columbia Practitioners Conference on the Mathematics of Finance, 1999. Compatibility

 • The Finance[SVJJProcess] command was introduced in Maple 15.