Abstract
As mentioned, the transition of a continuous time model into a discrete time model is not an easy issue. We discuss here various discretization procedures to turn continuous time into discrete time models. There are many methods to convert continuous time models into discrete time variants. The main discretization methods are the Euler method, the Milstein method and a new local linearization method. All those will be illustrated here to obtain discrete-time approximate models.
Access this chapter
Tax calculation will be finalised at checkout
Purchases are for personal use only
Notes
- 1.
The stochastic integration with respect to dW t is the Itô integration, see Karatzas and Shreve (1991).
- 2.
If the process deviates from \(\frac{c} {\beta }\) (the mean), for example, \(X_{t}> \frac{c} {\beta }\), then the process is drifting down and it is pulled up when \(X_{t} <\frac{c} {\beta }\).
- 3.
The application of the Milstein method for approximating diffusion processes is independently developed by the authors. In the appendix of this chapter we present our application and show that it is equivalent to that of Elerian (1998).
- 4.
The white noise in the discrete-time models is represented by Brownian increments Δ W t .
- 5.
See Lo (1988).
- 6.
See Sect. 7.3.
- 7.
We suppose this is the main reason for Shoji and Ozaki applying the NLL method for the nonlinear drift case in Shoji and Ozaki (1998).
- 8.
See Sect. 7.6.
- 9.
See Kloeden and Platen (1992, 345).
- 10.
See Kloeden and Platen (1992, Chap. 10).
- 11.
The data source is “OECD—Main Economic Indicators/immediate rates”.
- 12.
See Gallant and White (1990) Definition 3.13, p. 27 with \(Z_{ni} = U_{i}U_{i-k}\). One can see v m = 0 when m ≥ k.
- 13.
See Gallant and White (1990), Theorem 5.3, p. 76. The conditions of the theorem are satisfied because under the null U i is independent and \(v_{n} = n - k\).
- 14.
See Breiman (1973, p. 189).
- 15.
Because the variance is normalized to 1, the concentration of the distribution around 0 let the variance smaller. In order to keep the variance as 1, there must be more weight in the tail.
- 16.
We undertake simulation with an interval 0. 01 and then pick up the simulated series with an interval 1.
- 17.
See Box et al. (1994).
- 18.
See Engle (1982).
- 19.
There is a thick tail effect if the kurtosis, defined as \(\frac{E(\epsilon ^{4})} {(E(\epsilon ^{2}))^{2}}\), is greater than 3—the kurtosis of normal distribution.
- 20.
See Brenner et al. (1996, p. 95) “The Ljung-Box Q(ε t ∕σ t ) statistics indicate that both models have significant serial correlation in the residuals.”
- 21.
See Bollerslev (1986).
- 22.
This value has been proposed by the CIR model, see Cox et al. (1985b).
- 23.
See for example, Diebold et al. (2006).
Bibliography
Ait-Sahalia, Y. (1996). Testing continuous-time models of the spot interest rate. The Review of Financial Studies, 9(2), 385–426.
Ait-Sahalia, Y. (1999). Transition densities for interest rate and other nonlinear diffusions. Journal of Finance, 54(4), 1361–1395.
Ait-Sahalia, Y. (2002). Maximum likelihood estimation of discretely sampled diffusions: A closed-form approximation approach. Econometrica, 70, 223–262.
Andersen, T., & Lund, J. (1997). Estimating continuous-time stochastic volatility models of the short-term interest rate. Journal of Econometrics, 77, 343–377.
Bollerslev, T. (1986). Generalized autoregressive conditional heteroskedasticity. Journal of Econometrics, 31, 307–327.
Box, G., Jenkins, G., & Reinsel, G. (1994). Time series analysis (3rd ed.). Prentice Hall.
Box, G., & Pierce, D. (1970). Distribution of residual autocorrelations in autoregressive-integrated moving average time series models. Journal of American Statistical Association, 65, 1509–1526.
Breiman, L. (1973). Statistics. Boston, MA: Houghton Mifflin.
Brenner, R., Harjes, R., & Kroner, K. (1996). Another look at models of the short-term interest rate. Journal of Financial and Quantitative Analysis, 31(1), 85–107.
Chan, K., Karolyi, G., Longstaff, F., & Sanders, A. (1992). An empirical comparison of alternative models of the short-term interest rate. Journal of Finance, 47(3), 1209–1227.
Cox, J., Ingersoll, J., & Ross, S. (1985a). An intertemporal general equilibrium model of asset prices. Econometrica, 53(2), 363–384.
Cox, J., Ingersoll, J., & Ross, S. (1985b). A theory of the term structure of interest rates. Econometrica, 53(2), 385–407.
Dacunha-Castelle, D., & Florens-Zmirou, D. (1986). Estimation of the coefficients of a diffusion from discrete observations. Stochastics, 19, 263–284.
Diebold, F.X., Rudebusch, G.D., & Auroba, S.B. (2006). The macroeconomy and the yield curve: a dynamic latent factor approach. Journal of Econometrics, 131, 309–338.
Duffie, D., & Glynn, P. (2004). Estimation of continuous time Markov processes sampled at random time intervals. Econometrica, 72, 1773–1808.
Durbin, J. (1970). Testing for serial correlation in least-squares regression when some of the regressors are lagged dependent variables. Biometrika, 38, 410–421.
Elerian, O. (1998). A note on the existence of a closed form conditional transition density for the Milstein scheme. http://www.nuff.ox.ac.uk/users/elerian/milpaper.pdf
Engle, R. F. (1982). Autoregressive conditional heteroscedasticity with estimates of the variance of United Kingdom inflation. Econometrica, 50(4), 987–1007.
Gallant, A., & Tauchen, G. (1996). Which moments to match. Econometric Theory, 12, 657–681.
Gallant, A., & White, H. (1990). A unified theory of estimation and inference for nonlinear dynamic system. Journal of Economic Literature, 28(1), 90–92.
Gourieroux, C., Monfort, A., & Renault, E. (1993). Indirect inference. Journal of Applied Econometrics, 8, 85–118.
Hansen, L., & Scheinkman, J. (1995). Back to the future: Generating moment implications for continuous-time Markov processes. Econometrica, 63(4), 767–804.
Karatzas, I., & Shreve, S. (1991). Brownian motion and stochastic calculus. New York: Springer.
Kloeden, P. E., & Platen, E. (1992). Numerical solution of stochastic differential equations. Berlin: Springer.
Koedijk, K., Nissen, F., Schotman, P., & Wolff, C. (1997). The dynamics of short-term interest rate volatility. European Finance Review, 1, 105–130.
Lo, A. (1988). Maximum likelihood estimation of generalized Itô processes with discretely sampled data. Econometric Theory, 4, 231–247.
Perdersen, A. (1995). A new approach to maximum likelihood estimation for stochastic differential equations based on discrete observations. Scandinavian Journal of Statistics, 22, 55–71.
Shoji, I., & Ozaki, T. (1997). Comparative study of estimation methods for continuous time stochastic processes. Journal of Time Series Analysis, 18(5), 485–506.
Shoji, I., & Ozaki, T. (1998). Estimation for nonlinear stochastic differential equations by local linearization method. Stochastic Analysis and Applications, 16(4), 733–752.
Thompson, S. (2002). Evaluating the Goodness of Fit of Conditional Distributions, with an Application to Affine Term Structure Models. Working Paper, Harvard University.
Vasicek, O. (1977). An equilibrium characterization of the term structure. Journal of Financial Economics, 5, 177–188.
Author information
Authors and Affiliations
Appendix
Appendix
7.1.1 Tables: Estimation Results
7.1.2 The Likelihood Function of the Milstein Approximation
Here we show the derivation of the likelihood function when using the Milstein method. Following (7.3), the dynamic of the SDE (7.1) is approximated by
where \(\varDelta t_{i} = t_{i+1} - t_{i}\), \(\varDelta W_{t_{i}} = W_{t_{i+1}} - W_{t_{i}}\). Let
Then (7.14) becomes
Let \(x_{i} \in \mathbb{R}\) still be the realizations of \(X_{t_{i}}\) and y i be the realizations of \(Y _{t_{i}}\) for i = 0, …, N correspondingly. We solve Eq. (7.15) to obtain the realizations of \(\varDelta W_{t_{i}} = u_{i+1}^{+},u_{i+1}^{-}\), where
Then the conditional density is given by
as \(1 + \frac{2\gamma y_{i+1}} {x_{i}}> 0\). If \(1 + \frac{2\gamma y_{i+1}} {x_{i}} <0\), then the density above is infinity. If \(1 + \frac{2\gamma y_{i+1}} {x_{i}} <0\), which means there is no real solution of \(\varDelta W_{t_{i}}\) in (7.15) for such y i+1, therefore the density is equal to zero
Comparing the density function (7.16) and the density function in (2.5) p. 7 in Elerian (1998), it is not difficult to show the identity of these two functions by some calculation. By numerical operations of the ML estimations we must modify the density function, because when \(1 + \frac{2\gamma y_{i+1}} {x_{i}} = 0\), the value of the density function is infinity. Therefore we apply the following density function for the ML estimations:
Rights and permissions
Copyright information
© 2016 Springer-Verlag Berlin Heidelberg
About this chapter
Cite this chapter
Chiarella, C., Semmler, W., Hsiao, CY., Mateane, L. (2016). Continuous and Discrete Time Modeling. In: Sustainable Asset Accumulation and Dynamic Portfolio Decisions. Dynamic Modeling and Econometrics in Economics and Finance, vol 18. Springer, Berlin, Heidelberg. https://doi.org/10.1007/978-3-662-49229-1_7
Download citation
DOI: https://doi.org/10.1007/978-3-662-49229-1_7
Published:
Publisher Name: Springer, Berlin, Heidelberg
Print ISBN: 978-3-662-49228-4
Online ISBN: 978-3-662-49229-1
eBook Packages: Economics and FinanceEconomics and Finance (R0)