On Structural Breaks and Nonstationary Fractional Intergration in Time Series

Economics Letters, 1994

We apply a new ARFIMA approach to distinguish between the trend and difference stationary models of long-run dynamics for a well-known representative macroeconomic dataset.

Journal of Policy Modeling, 2004

In this article we propose a new approach that permits us to simultaneously test unit and fractional roots at the long run and the seasonal frequencies. We examine the industrial production indexes in four Latin American countries (Brazil, Argentina, Colombia and Mexico), using new statistical tools based on seasonal and non-seasonal long memory processes. Results show that the root at the long run or zero frequency plays a much more important role than the seasonal one. Nevertheless, in the cases of Brazil and Argentina a component of long memory behaviour is also present at the seasonal structure, indicating that shocks modify the seasonal structure for a long period. Policy makers should thus pay attention to this result in choosing the optimal economic policy.

International Journal of Trade, Economics and Finance, 2013

The purpose of this study is to carry out a comprehensive examination of the unit root hypothesis and structural breaks in ASEAN macroeconomic time series from 1960 to 2010 using endogenous break ADF-type unit root tests. Once allowance is considered for structural breaks, the number of rejections of a unit root null is relatively higher than without breaks. The difference between ZA and LP models, is that ZA shows that US Dollar terms denomination series are more favorable of trend-stationary processes, whereas the series under local currency terms tend to reject the null hypothesis of a unit root in LP models. Moreover, the break points are closely associated with global economic events such as the first and second oil shocks in 1973-1975 and 1979-1980, respectively, the commodity crisis (1985-1986) and the Asian financial crisis (1997-1998). The policy authorities may use the historical information to forecast future movements in macroeconomic time series. Lastly, our findings also shed light on the importance of considering exchange rate fluctuations in the process of trend-stationary and unit root.

International Journal of Finance & Economics, 2020

This paper is a comparative analysis of Nigeria and Kenya, the largest economies in West and East Africa respectively, on the basis of the time series properties of their economic activities through the Gross Domestic Product (GDP) and growth rate series. It further analyses how differing policy and political economy processes contributed to the two countries' economic growth trajectories despite becoming independent republics at almost the same time. We study the two economies using a long-memory-fractionally integrated approach. The results show a high degree of persistence in both cases. When non-linearities are taken into account, evidence of mean reversion is found in the GDP series in the two countries. This is indicative of how the two countries in very distinct African contexts followed broadly different but, in some ways, similar paths toward economic growth since independence.

International Journal of Business Marketing and Management, 2020

This paper deals with the analysis of the GDP growth in the CEMAC countries, which are Chad, the Central African Republic, Congo, Gabon and Cameroon and Equatorial Guinea. We use time series techniques based on the concept of fractional integration and cointegration. The univariate analysis based on fractional integration aims to determine whether the series are I(1) or alternatively I(d) with d<1, which would imply mean reversion. We examine the nature of the shocks in the real GDP series in these countries along with the hypothesis of convergence in relation with the US. Starting with the univariate results, the results indicate strong evidence of persistence, with the orders of integration being around the I(1) case. Testing the hypothesis of convergence, by means of looking at the order of integration of the difference between the log real GDP (and real GDP per capita) or each country minus the one corresponding to the US, we obtain further evidence to support the unit roots, rejecting thus any possibility of long run relationships between these African countries and the US.

2014

This paper examined the long memory features of GDP per capita data before the global financial crisis, using a sample of 26 African countries. The study employed fractional integration and tested the stability of the differencing parameter across the sample period for each country. The results indicated that most of the countries’ GDP series were I(1) or higher. Evidence of mean reversion was observed in 10 countries where the disturbances were autocorrelated. There was strong evidence against mean reversion in the remaining 16 countries. The results also indicated that the fractional differencing parameter was stable in 17 countries, while the presence of structural breaks was investigated in the remaining 9 countries.

2005

Testing for unit roots has special significance in terms of both economic theory and the interpretation of estimation results. as there are several methods available, researchers face method selection problem while conducting the unit root test on time series data in the presence of structural break. this paper proposes a sequential search procedure to determine the best test method for

African Development Review, 2014

This study examines the time series behaviour of several Angolan macroeconomic variables, using monthly data from August 1996 to June 2011. The series are the inflation rate, M1, M2, the exchange rate at the beginning and the end of the period, and the monthly average exchange rate. In the first stage univariate fractional integration models are estimated in order to determine whether shocks to the variables have transitory or permanent effects. In the second stage fractional cointegration techniques are applied to test for the existence of long-run equilibrium relationships between the variables of interest. The results suggest a high degree of persistence in the individual series (that are not mean-reverting) and the existence of bivariate long-run cointegrating relationships between prices and money, and prices and nominal exchange rates.

Empirical Economics, 2009

This paper identifies structural breaks in the post-World War II joint dynamics of U.S. inflation, unemployment and the short-term interest rate. We derive a structural breakdate procedure which allows for long-memory behavior in all three series and perform the analysis for alternative data frequencies. Both long-memory and short-run coefficients are relevant for characterizing the changing patterns of U.S. macroeconomic dynamics. We provide an economic interpretation of those changes by examining the link between macroeconomic events and structural breaks.

Statistics and Computing, 2012

This paper considers a general model which allows for both deterministic and stochastic forms of seasonality, including fractional (stationary and nonstationary) orders of integration, and also incorporating endogenously determined structural breaks. Monte Carlo analysis shows that the suggested procedure performs well even in small samples, accurately capturing the seasonal properties of the series, and correctly detecting the break date. As an illustration, the model is estimated for four different US series (output, consumption, imports and exports). The results suggest that the seasonal patterns of these variables have changed over time: specifically, in the second subsample the systematic component of seasonality becomes insignificant, whilst the degree of persistence increases.

The research was conducted on time series data with the objective of trying to find out if there are structural breaks in the Ordinary Exchange Rate (OER) in Nigeria. But more specifically, the researchers applied various methods in testing the structural breaks such as the F Statistic in comparing statistical models fitted to the data set to find out if the model fits the population. The robustness of this test was further validated with Quant Andrew and Bai Perron test. And unit root tests were employed to test the stationarity of values using Bartlett kernel and Kwiatkowski-Phillips-Schmidt-Shin. With evidence of stationarity an equation was introduced to capture seasonality given that the data is times series. Further correlogram tests was used to test if the error term is stationary and the results indicate that the level of Autocorrelation and the Partial Autocorrelation were very insignificance. A major finding was that in using the ARIMA model it was evident that AR is stationary and MA is invertible. However, the MA has roots close to 1, implying that there may be evidence of over differencing of the series. In conclusion, the identified break dates of 1992Q2, 1995Q3 and 2005Q3 coincides with a period of persistent excess liquidity exacerbated by the monetization of excess crude receipts and the distribution of enhanced statutory allocation to the three tiers of government. The effect of the identified structural break was accommodated in our modeling approach to ensure that the estimated parameters are unbiased. The preliminary analysis shows that the exchange rate was more robust than other rates in explaining developments in the foreign exchange market. Policy recommendations include that policy makers give feasible proposals on diversification away from oil. Other measures include a review on regulatory, fiscal and monetary policy to reduce the impact of inflation and to increase global competitiveness of exports so as to attain the adequate level of exchange rate. Furthermore, the uncertainties inherent in the timing of the breaks and the different models used in capturing them in theory makes a case for more research on structural breaks in modeling and forecasting the volatility in exchange rate.

2011

Anecdotal evidence shows that country-specific inflation has remained largely persistent and heterogeneous across the West African Monetary Zone (WAMZ). Uncertainty about the nature of inflation persistence often undermines regional convergence due to the asymmetric responses from monetary authorities. Consequently, the objective of this paper is to identify whether or not structural breaks exist in the price level of member countries. The paper uses the method of Lee and Strazicich (2003) with multiple breaks to identify if there were spurious rejections in the ADF tests of selected macroeconomic variables of WAMZ countries. The result fails to reject the existence of the null hypothesis of unit root for fifteen (15) variables and a second root for twenty (20) variables. With the exception of Ghana, the CPI for other member countries has a second root, indicating that inflation is explosive with hysteresis effect. This feature of inflation is associated with a similar structural dy...

2004

This paper discusses the role of deterministic components in the DGP and in the auxiliary regression model which underlies the implementation of the Fractional DickeyFuller (FDF) test for I(1) against F I(d) processes with d 2 [0; 1): Invariant tests to the presence of a drift under the null of I(1) are derived. In common with the standard DF approach in the I(1) vs: I(0) framework, we also examine the consequences of including a constant and /or a linear trend in the regression model when there is a drift under the null. A simple testing strategy entailing only asymptotically normally-distributed tests is proposed. Finally, an empirical application is provided where the FDF test allowing for deterministic components is used to test for long-memory in the per capita GDP of several OCDE countries.

Economic Journal of Emerging Markets, 2006

This paper examines the robustness of the ADF (Augmented Dickey-Fuller) unit root test to the presence of one structural break. The ADF test results show one variables out of six to be stationary. To check their robustness, two separate additive outlier (AO) models are employed: one allowing for one endogenously-determined break in the intercept and the other in the trend. These two tests can not reject the unit root null hypothesis for all the vari-ables. However, when an innovational outlier (IO) model, that allows for one endogenously-determined break is estimated, the null hypothesis can be rejected for 3 more series. The estimated break dates mostly correspond to the 1998 financial crisis in Indonesia. Keywords: unit root; stationarity; structural break, additive & innovational outlier JEL classification: C1; C22

Journal of econometrics, 2002