`Journal of Probability and StatisticsVolume 2011 (2011), Article ID 691058, 11 pageshttp://dx.doi.org/10.1155/2011/691058`
Review Article

## The CSS and The Two-Staged Methods for Parameter Estimation in SARFIMA Models

1Department of Statistics, Ondokuz Mayis University, 55139 Samsun, Turkey
2Department of Statistics, Hacettepe University, Beytepe, 06800 Ankara, Turkey

Received 19 May 2011; Accepted 1 July 2011

Copyright © 2011 Erol Egrioglu et al. This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

#### Abstract

Seasonal Autoregressive Fractionally Integrated Moving Average (SARFIMA) models are used in the analysis of seasonal long memory-dependent time series. Two methods, which are conditional sum of squares (CSS) and two-staged methods introduced by Hosking (1984), are proposed to estimate the parameters of SARFIMA models. However, no simulation study has been conducted in the literature. Therefore, it is not known how these methods behave under different parameter settings and sample sizes in SARFIMA models. The aim of this study is to show the behavior of these methods by a simulation study. According to results of the simulation, advantages and disadvantages of both methods under different parameter settings and sample sizes are discussed by comparing the root mean square error (RMSE) obtained by the CSS and two-staged methods. As a result of the comparison, it is seen that CSS method produces better results than those obtained from the two-staged method.

#### 1. Introduction

In the recent years, there have been a lot of studies about Autoregressive Fractionally Integrated Moving Average (ARFIMA) models in the literature. However, most of time series in real life may have seasonality, in addition to long-term structure. Therefore, SARFIMA models have been introduced to model such time series. Generally, SARFIMA process is given in the following form: where is a time series, is the back shift operator, such as , is the seasonal lag, and represent the nonseasonal and seasonal fractionally differences; respectively, is a white noise process and has normal distribution , and , , , and are given by where and are the orders of the nonseasonal and seasonal parameters, respectively.

Baillie [1] and Hassler and Wolters [2] examined the basic characteristics of ARFIMA models, while some significant contributions to the SARFIMA models were presented by Giraitis and Leipus [3], Arteche and Robinson [4], Chung [5], Velasco and Robinson [6], Giraitis et al. [7], and Haye [8]. When all parameters are different from zero in (1.1) and when some parameters such as are equal to zero, different parameter estimation methods are compared by performing simulation studies in the literature [911].

Seasonal long-term structure exists in time series in various study fields such as the cumulative money series in Porter-Hudak [12], the IBM input series in Ray [13], and the Nile River data in Montanari et al. [14]. Candelon and Gil-Alana [15] forecasted the industrial production index of countries in South America by employing the SARFIMA models. Gil-Alana [16] found that the GDP series in Germany, Italy, and Denmark had a structure which was suitable to use SARFIMA models.

Brietzke et al. [17] utilized Durbin-Levinson algorithm for the model. Ray [13] modified the method proposed by Hosking [18] and used this modified method for a special SARFIMA process having two different seasonal difference parameters. Darné et al. [19] adapted the method, proposed for ARFIMA by Chung and Baillie [20], to SARFIMA models. However, the properties of the CSS method employed in Darné et al. [19] have not been examined by a simulation study yet.

Arteche and Robinson [4] introduced a semiparametric method based on spectral density functions while estimating parameters for SARFIMA model in the case of . GPH method used in ARFIMA is extended to be used in SARFIMA models for by Porter-Hudak [12], and GPH estimator has been modified by Ooms and Hassler [21]. Also, a simulation study for different values of , and sample size has been conducted using GPH, Whittle and Exact Maximum likelihood (EML) by Reisen et al. [9, 10] and Palma and Chan [11]. In addition to these studies, many methods for determining seasonal long-term structure have been proposed by Hassler and Wolters [22], Gil-Alaña and Robinson [23], Arteche [24], and Gil-Alana [25, 26].

We examine the properties of the CSS and two staged estimation methods by a simulation study in which both methods are compared based on various parameter settings and sample sizes. In the simulation study, a specific form of the model given in (1.1) in which , , and are equal to zero is examined by using the both CSS and two staged estimation methods. This model can also be expressed as SARFIMA  . After simulation study was conducted, the results obtained from the CSS and two staged estimation methods are compared, and it is observed that better results are obtained when the CSS method is employed.

The outline of this study is as follows. Section 2 contains brief information related to SARFIMA models. The CSS method and two staged methods are explained in Sections 3 and 4, respectively. The outline of the simulation study and the results are given in Section 5. Finally, the results obtained from the simulation study are summarized in the last section.

#### 2. SARFIMA Models

When , , , , and are set to zero in model (1.1), this model is called as Seasonal Fractionally Integrated (SFI) model. The SFI model was firstly introduced by Arteche and Robinson [4], and basic information about the model can be found in Baillie [1]. SFI model can be given by Infinite moving average presentation of the model (2.1) is as follows: where , (, for ).

Infinite autoregressive presentation of the model (2.1) is as follows: where , (, for ).

For model (2.1), autocovariance and autocorrelation functions can be, respectively, written as follows: when For model (2.1), spectral density function is as follows: Note that the spectral density function is infinite at the frequencies , .

When , , , , , and are different from zero in model (1.1), closed form for autocovariances cannot be determined. However, some methods, such as the splitting method presented by Bertelli and Caporin [27], employed to calculate autocovariances of ARFIMA models, can also be used for those in SARFIMA models.

Let denote the autocovariance function of SARFIMA models. Autocovariances are calculated in terms of splitting method as follows: and are autocovariances functions for SARFIMA and SARFIMA models, respectively. is calculated using splitting method given in a following expression: and are autocovariances functions for SARFIMA and SARFIMA models, respectively. The closed form for is given in (2.4). The autocovariances of are autocovariances of fractionally integrated process and the closed form is given by [28] as follows: To generate series, which are appropriate for SARFIMA models, the following algorithm is applied.

Step 1. Generate random variable vector with standard normal distribution.

Step 2. Obtain the matrix , by utilizing the expression (2.4).

Step 3. Split the covariance matrix as follows: where, is a lower triangular matrix.

This splitting is called Cholesky. It is possible to obtain Cholesky decomposition of positive definite and symmetric matrices. Note that matrix is positive definite and symmetric.

Step 4. Obtain series by using formula. has a suitable structure for SARFIMA model.

Step 5. Generate series according to SARMA   model by taking as error series. By this way, the new generated series have the structure of SARFIMA . This algorithm is easily extended to SARFIMA model.

#### 3. The Two-Staged Method

The two-staged method can be used to estimate the parameters of SARFIMA model. In the first phase of this method, it is assumed that the time series has a suitable structure to use the SARFIMA model, and seasonal fractionally difference parameter is estimated. In the second phase, estimation of the parameter, EML method given below, can be employed.

Theoretical autocovariance and autocorrelation functions for SARFIMA model are shown in (2.4) and (2.5) respectively. Let time series have observations , and let represent the autocorrelation matrix of . Therefore, the likelihood function of is as follows: Cholesky decomposition is used for the matrix as multiplication of lower and upper triangular matrices in calculation of the likelihood function. Instead of calculating the inverse of matrix (), inverses of lower and upper triangular matrices are calculated by using the decomposition. Thus, the decomposition decreases computational difficulty and calculation time. Cholesky decomposition of the matrix is written as follows: Let , and it can be written Thus, (3.1) can be rewritten as The likelihood function given in (3.4) is maximized in terms of seasonal fractionally difference parameter by using an optimization algorithm. After seasonal fractionally difference parameter is estimated by using EML, the rest of the parameters of SARMA   model are estimated in the second phase by using the classic method. In the second phase, the order of the seasonal model can be determined by using the Box-Jenkins approach. Therefore, the two-staged method can be summarized as follows.

Phase 1. Estimate the parameter by assuming the time series suitable for SARFIMA .

Phase 2. Estimate seasonal autoregressive and moving average parameters by using the Box-Jenkins methodology.

#### 4. The CSS Method

Chung and Baillie [20] proposed a method based on minimization of conditional sum of square. This method can be used for SARFIMA   models. Conditional sum of square method for SARFIMA model is as follows: In the CSS method, firstly, seasonal fractionally difference procedure is executed for . Secondly, fractionally difference procedure is executed for . Thirdly, SARMA filtering is applied to . By calculating sum of squares of this obtained series , conditional sum of square is calculated for a fixed value of and . Chung and Baillie [20] also emphasize that the estimations of parameters obtained by the CSS method have less bias when the mean value of the series is known. It is easy to use the CSS method because it does not need to calculate autocovariances. In the literature, the CSS method for the SARFIMA   model has been used only by Darné et al. [19].

#### 5. Simulation Study

In this section, the parameters of SARFIMA model are estimated by using the CSS and the two-staged methods separately under different parameter settings and sample sizes. Also, the advantages and the disadvantages of both methods are discussed.

The algorithm, whose steps are given in Section 2, is used to generate various SARFIMA models. SARFIMA and SARFIMA models are emphasized in the simulation study. For SARFIMA model, 36 different cases are examined such as seasonal fractionally difference , seasonal autoregressive parameter , sample sizes , and period . Similarly, the same parameters are also used for SARFIMA model by taking . For each case, 1000 time series are generated, so totally we generate 72000 time series. The parameters of the generated time series are estimated by using both the CSS and two-staged methods whose results are summarized in Tables 1 and 2. For each 1000 time series, the mean, standard deviation, and root mean square error (RMSE) values of estimated parameters are exhibited in these tables. RMSE values are computed by where and denote the real and estimated values of parameter, respectively.

Table 1: Simulation results of the CSS and two-stage method in SARFIMA .
Table 2: Simulation results of CSS and two-stage method in SARFIMA .

In Table 1, for SARFIMA model, the simulation results for different values of and sample size are shown when the CSS and the two-stage methods are executed. From this table, for CSS method, we observe that RMSE values have sharply decreased for the estimated parameters of seasonal fractional difference and seasonal autoregressive, when the sample size increases. It is also seen that the values of RMSE do not change much whether the sign of parameter of seasonal autoregressive is positive or not. In the case of having larger value of seasonal autoregressive parameter in absolute, RMSE values of seasonal autoregressive () parameters get smaller. When and are compared, the values of in are smaller than those in , whereas the values of in , , and are close with each other. Note that the values of in are larger than those in .

According to Table 1, when the two-staged method is executed, it is observed that the sample size does not affect significantly the values of RMSE, especially for when . However, when the absolute value estimated of seasonal autoregressive parameter increases, the values of increase dramatically in and . The values of are not affected by both the sign and magnitude of seasonal autoregressive parameter, especially in . It is worth to point out that the values of are quite larger for the negative values of seasonal autoregressive parameters in both and . It can be inferred from the comparison between and that for the negative values of seasonal autoregressive parameter, both the values of and increase gradually while is increasing. Especially, the values of in get the biggest values when the seasonal autoregressive parameter is negative. Therefore, for the negative values of seasonal autoregressive parameters, we can say that the estimation error gets bigger while the order of seasonal fractional difference is increasing.

In Table 2, for the SARFIMA model, the simulation results for different values of parameter and sample size are shown for the CSS and two-staged methods. From this table, we observe that RMSE values have decreased for the estimated parameters of seasonal fractional difference and seasonal moving average, when the sample size increases, the CSS method is executed. It is also seen that the values of RMSE do not change much whether the sign of parameter of seasonal moving average is positive or not. In the case of having larger value of seasonal moving average parameter for the negative values, RMSE values for seasonal fractionally difference () are smaller. When we compare with , the values of in are smaller than those in , whereas the values of among , , and are close with each other. Note that the values of in are larger than those in .

When Table 2 is examined, it is observed that the values of decrease when sample size increases for two-staged method. However, there is no positive or negative relations between the value of seasonal moving average parameter and the values of and when two-stage method is executed. We would like to remark that has the smallest value in each sample size for and that values of are quite big for the negative values of seasonal moving average parameter with respect to its positive values when , and 0.3 in Table 2.

#### 6. Discussions

In the literature, the two-staged method is a widely used method to estimate parameters of SARFIMA models. Although there is another method called CSS, this method has not been employed to estimate the parameters of SARFIMA model. In this study, the CSS and the two-staged methods are employed to estimate parameters of the SARFIMA models by conducting a simulation study, and by this way the properties of these two methods are examined under different parameter settings and sample sizes.

From the results of the simulation, we deduce that when the sample size increases, the CSS method gives more accurate estimates. Besides, we can infer that when seasonal autoregressive parameter in SARFIMA model gets close to 1 or −1, the parameter estimates of the CSS method have less error. The CSS method produces quite good estimates for when the seasonal autoregressive parameter in SARFIMA model and the seasonal moving average parameter in SARFIMA model are positive.

When the CSS method is compared with the two-staged method, the CSS method has lower RMSE values than the two-staged method under different parameter settings and sample sizes, especially in autoregressive models. Two-staged method generates misleading results when is chosen near −1 . However, this is not the case for the CSS method. Based on the obtained results and simplicity of the method, for forthcoming studies it can be easily suggested that the CSS method should be preferred rather than the two-staged method in the parameter estimation for SARFIMA models.

#### References

1. R. T. Baillie, “Long memory processes and fractional integration in econometrics,” Journal of Econometrics, vol. 73, no. 1, pp. 5–59, 1996.
2. U. Hassler and J. Wolters, “On the power of unit root tests against fractional alternatives,” Economics Letters, vol. 45, no. 1, pp. 1–5, 1994.
3. L. Giraitis and R. Leipus, “A generalized fractionally differencing approach in long-memory modeling,” Lithuanian Mathematical Journal, vol. 35, no. 1, pp. 53–65, 1995.
4. J. Arteche and P. M. Robinson, “Semiparametric inference in seasonal and cyclical long memory processes,” Journal of Time Series Analysis, vol. 21, no. 1, pp. 1–25, 2000.
5. C.-F. Chung, “A generalized fractionally integrated autoregressive moving-average process,” Journal of Time Series Analysis, vol. 17, no. 2, pp. 111–140, 1996.
6. C. Velasco and P. M. Robinson, “Whittle pseudo-maximum likelihood estimation for nonstationary time series,” Journal of the American Statistical Association, vol. 95, no. 452, pp. 1229–1243, 2000.
7. L. Giraitis, J. Hidalgo, and P. M. Robinson, “Gaussian estimation of parametric spectral density with unknown pole,” The Annals of Statistics, vol. 29, no. 4, pp. 987–1023, 2001.
8. M. O. Haye, “Asymptotic behavior of the empirical process for Gaussian data presenting seasonal long-memory,” European Series in Applied and Industrial Mathematics, vol. 6, pp. 293–309, 2002.
9. V. A. Reisen, A. L. Rodrigues, and W. Palma, “Estimating seasonal long-memory processes: a Monte Carlo study,” Journal of Statistical Computation and Simulation, vol. 76, no. 4, pp. 305–316, 2006.
10. V. A. Reisen, A. L. Rodrigues, and W. Palma, “Estimation of seasonal fractionally integrated processes,” Computational Statistics & Data Analysis, vol. 50, no. 2, pp. 568–582, 2006.
11. W. Palma and N. H. Chan, “Efficient estimation of seasonal long-range-dependent processes,” Journal of Time Series Analysis, vol. 26, no. 6, pp. 863–892, 2005.
12. S. Porter-Hudak, “An application of the seasonal fractionally differenced model to the monetary aggregates,” Journal of American Statistical Association, vol. 85, no. 410, pp. 338–344, 1990.
13. B. K. Ray, “Long-range forecasting of IBM product revenues using a seasonal fractionally differenced ARMA model,” International Journal of Forecasting, vol. 9, no. 2, pp. 255–269, 1993.
14. A. Montanari, R. Rosso, and M. S. Taqqu, “A seasonal fractional ARIMA model applied to the Nile River monthly flows at Aswan,” Water Resources Research, vol. 36, no. 5, pp. 1249–1259, 2000.
15. B. Candelon and L. A. Gil-Alana, “Seasonal and long-run fractional integration in the Industrial Production Indexes of some Latin American countries,” Journal of Policy Modeling, vol. 26, no. 3, pp. 301–313, 2004.
16. L. A. Gil-Alana, “Seasonal long memory in the aggregate output,” Economics Letters, vol. 74, no. 3, pp. 333–337, 2002.
17. E. H. M. Brietzke, S. R. C. Lopes, and C. Bisognin, “A closed formula for the Durbin-Levinson's algorithm in seasonal fractionally integrated processes,” Mathematical and Computer Modelling, vol. 42, no. 11-12, pp. 1191–1206, 2005.
18. J. R.M. Hosking, “Modeling persistence in hydrological time series using fractional differencing,” Water Resources Research, vol. 20, no. 12, pp. 1898–1908, 1984.
19. O. Darné, V. Guiraud, and M. Terraza, “Forecasts of the Seasonal Fractional Integrated Series,” Journal of Forecasting, vol. 23, no. 1, pp. 1–17, 2004.
20. C.-F. Chung and R. T. Baillie, “Small sample bias in conditional sum-of-squares estimators of fractionally integrated ARMA models,” Empirical Economics, vol. 18, no. 4, pp. 791–806, 1993.
21. M. Ooms and U. Hassler, “On the effect of seasonal adjustment on the log-periodogram regression,” Economics Letters, vol. 56, no. 2, pp. 135–141, 1997.
22. U. Hassler and J. Wolters, “Long memory in inflation rates: international evidence,” Journal of Business & Economic Statistics, vol. 13, no. 1, pp. 37–45, 1995.
23. L. A. Gil-Alaña and P. M. Robinson, “Testing of seasonal fractional integration in UK and Japanese consumption and income,” Journal of Applied Econometrics, vol. 16, no. 2, pp. 95–114, 2001.
24. J. Arteche, “Semiparametric robust tests on seasonal or cyclical long memory time series,” Journal of Time Series Analysis, vol. 23, no. 3, pp. 251–285, 2002.
25. L. A. Gil-Alana, “Seasonal misspecification in the context of fractionally integrated univariate time series,” Computational Economics, vol. 22, no. 1, pp. 65–74, 2003.
26. L. Gil-Alana, “Modelling seasonality with fractionally integrated processes,” Tech. Rep. number, 2003.
27. S. Bertelli and M. Caporin, “A note on calculating autocovariances of long-memory processes,” Journal of Time Series Analysis, vol. 23, no. 5, pp. 503–508, 2002.
28. J. Beran, Statistics for Long-Memory Processes, vol. 61, Chapman & Hall/CRC, New York, NY, USA, 1994.