Monthly Seasonal Factors of Selected Economic Time Series*

Analysis of seasonal behaviour of different economic time series is considered important from the policy perspective, as it enables the policy maker to differentiate between the seasonal changes and long-run changes in the variable and thereby design appropriate policy responses. This article presents the monthly seasonal factors of selected 74 economic/financial time series. Over the years, the intra-year seasonal variations of Aggregate Deposits, Bank Credit and Investments of Scheduled Commercial Banks exhibited almost similar movements. Within the WPI-All Commodities group, WPI-Primary Articles indicated higher seasonal fluctuations as compared to WPI-Manufactured Products group. CPI for ‘Agricultural Labourers’ and ‘Rural Labourers’ exhibited similar seasonal variations during the reference period (i.e., April 1994 to March 2009). Moreover, seasonal variations of CPI-Industrial Workers were higher than those of WPI-All Commodities. Among the use-based classification of industries, ‘Intermediate Goods’ industries experienced the lowest seasonal variations, throughout the reference period, whereas, IIP-Capital Goods exhibited the highest seasonal variations in all the years, during the reference period, except in 1999-00 and 2000-01. Within the Consumer Goods industries, ‘Consumer Non-Durable Goods’ exhibited greater seasonal fluctuations than ‘Consumer Durable Goods’.

1. Introduction

Economic time series, in general, exhibit regular, intra-year seasonal fluctuations around their annual trend path. These fluctuations are caused by the shifts in preferences, production cycle characteristics, seasonal nature of economic activity, and other causes like weather, festivals, vacation, etc. The study of direction and magnitude of seasonal fluctuations helps in providing suitable treatment to seasonality, which in turn is useful in interpretation of movements in economic variables. Moreover, the seasonal adjustment estimates remove seasonal effects from a time series, and thus provide better information on certain non-seasonal features. From the policy perspective, the information on seasonal factors of an economic variable is useful as it enables the policy maker to differentiate between the seasonal changes and long-run changes in the variable and thereby design appropriate policy responses. In the recent period, with the availability and usage of high frequency data, the analysis of seasonal factors has, in fact, become essential for policy design, especially of short duration of less than a year. In this backdrop, monthly seasonal factors for selected economic and financial time series of the Indian economy are being regularly published in the Reserve Bank of India Bulletin from 1980 onwards¹.

2.Data Coverage and Methodology

The present article covers monthly data from April 1994 to March 2009 (reference period) with respect to 74 economic/financial series, including eight new series, for the estimation of seasonal factors (Annex). The seasonal factors of these 74 time series are classified into six major groups, namely,

A. Monetary and Banking Indicators (21 series);
B. Wholesale Price Index (WPI) (18 series);
C. Consumer Price Index (CPI) (4 series);
D. Industrial Production (26 series);
E. Services Sector Indicators (2 series);
F. External Trade (3 series).

The seasonal factors for external trade variables are generated using data in US dollar terms. The data related to the series under study, except those of ‘Cement Production’, ‘Steel Production’, ‘Production of Commercial Motor Vehicles’ and Services Sector indicators are taken from the Reserve Bank of India Monthly Bulletin. The definitions of the series are the same as those in the ‘Notes on Tables’ of the Reserve Bank of India Monthly Bulletin. The data related to ‘Cement Production’, ‘Steel Production’, ‘Production of Commercial Motor Vehicles’ and Services Sector indicators are taken from Central Statistical Organisation (CSO) Monthly Capsule Report (on Infrastructure Sector Performance).

The seasonal factors have been estimated on the basis of X-12 ARIMA² methodology, developed by the US Bureau of Census. The monthly seasonal factors for the year 2008-09, the average monthly seasonal factors for the period 1994-95 to 2008-09 and the forecasts of monthly seasonal factors for 2009-10 are presented in Table 1. The peak and trough values of the seasonal factors along with their respective months of occurrence are also presented in Table 1. Table 2 presents the dispersion of the seasonal factors, during the last 14 years, in terms of ranges (difference between peak and trough values) of seasonal factors for all the 74 variables. Monthly seasonal factors, for the last 14 years, for some of the important series are presented in the Table 3. The salient features of variations in seasonal factors of some selected series during the last fourteen years in general, and for 2008-09 in particular, are briefly discussed in the subsequent paragraphs.

3. Salient Features

3.1 Monetary and Banking Aggregates

As a part of Monetary and Banking aggregates, 21 important indicators were selected to analyse the intra-year seasonal variations over the reference period.

3.1.1 Seasonal Factors

The seasonal factors of Reserve Money (M₀) and Narrow Money (M₁), during 2008-09, had their peaks in the month of March, whereas, the troughs occurred in the months of August and November, respectively. In the case of Broad Money (M₃), the peak and trough occurred in the months of April and December, respectively (Table 1 and Chart 1).

From the sources side of Broad Money, the Net Bank Credit to Government exhibited seasonal peak and trough in July and December, respectively during 2008-09, whereas the Bank Credit to Commercial Sector had peak and trough in the months of March and July, respectively (Table 1 and Chart 2).

During 2008-09, the seasonal factors of ‘Currency in Circulation’ and ‘Notes in Circulation’ observed similar movements having seasonal peaks and troughs in April and September, respectively. Although, the seasonal trough of ‘Currency with the Public’ occurred during September, the seasonal peak was observed in May (Table 1 and Chart 3).

During 2008-09, the peaks of Aggregate Deposits (SCBs) and Bank Credit (SCBs) occurred during the months of April and March, respectively, whereas, the troughs occurred during the months of December and August (Table 1). The seasonal variations in Aggregate Deposits (SCBs) appear to be relatively much less, as compared to those in Bank Credit (SCBs) (Chart 4).

3.1.2 Range of Seasonal Factors

Over the years, the intra-year seasonal variations, measured in terms of a range (i.e., difference between peak and trough values) of seasonal factors, of Aggregate Deposits (SCBs), Bank Credit (SCBs) and Investments (SCBs), exhibited almost similar movements; having lower seasonal variations during the period 2000-01 to 2002-03. During the last four years (viz., 2005-06 to 2008-09), the seasonal variations of Bank credit (SCBs) and Investments (SCBs) showed upward movement, whereas, the variation in Aggregate Deposits (SCBs) moved within the narrow band of 2.9 to 3.1 (Chart 5 and Table 2).

During the period 1995-96 to 2008-09, Demand Deposits (SCBs) have shown greater seasonal fluctuations than that of Time Deposits (SCBs) (Chart 6). The seasonal fluctuations in Demand Deposits (SCBs) gradually declined upto 2001-02, and thereafter, started rising. During the last three years, viz. 2006-07 to 2008-09, the seasonal variation of Demand Deposits (SCBs) moved within the narrow band of 14.4 to 14.6.

3.2. Price Indices

In order to investigate the seasonal movements of the price indices over the reference period, both Wholesale Price Indices (WPI) with base year 1993-94 and Consumer Price Indices (CPI) were taken into account. From the WPI group, 18 important series (commodities) were selected including food articles like cereals, wheat, rice and pulses. Seasonal factors for major groups like Primary Articles and Manufactured Products were also worked out. From, the CPI group, all 4 series based on occupations, viz., CPI-Agricultural Labourers (CPI-AL), CPI-Rural Labourers (CPI-RL), CPI-Industrial Workers (CPI-IW) and CPI-Urban Non-Manual Employees (CPI-UNME) were selected.

3.2.1 Seasonal Factors

During 2008-09, the seasonal factors of WPI All Commodities registered peak in September and trough in February (Table 1). In terms of average, it had peak in September and trough in March. The two main groups of All Commodities, viz., ‘Manufactured Products’ and ‘Primary Articles’ had the peaks and troughs, which occurred in the same months of September and February, respectively, during 2008-09 (Table 1 and Chart 7). The seasonal variations of ‘Manufactured Products’ prices were relatively much lower, as compared to ‘Primary Articles’.

During 2008-09, the seasonal peaks of the prices for Food Articles and Non-Food Articles occurred during the months of September and July, respectively, whereas the troughs occurred in the months of March and December (Chart 8). The seasonal variation of WPI-Food Articles was lower than that of WPI-Non-Food Articles, during this year.

The Consumer Price Index (CPI) for both Agricultural and Rural Labourers, observed the seasonal peaks and troughs in October and March, respectively, during the year 2008-09 (Table 1 and Chart 9). Over the reference period, the peaks of these two series varied over October and November, while the troughs varied over April and March, respectively.

During 2008-09, the seasonal movement for other two Consumer Price series, viz., CPI-IW and CPI-UNME followed the same patterns as were observed throughout the reference period. The seasonal peaks for both CPI-IW and CPI-UNME occurred in the month of October, whereas, the troughs occurred in February and March, respectively (Table 1 and Chart 10).

3.2.2 Range of Seasonal Factors

The range (i.e., difference between peak and trough values) of seasonal factors of WPI-All Commodities declined from 1.8 in 1995-96 to 1.2 in 2001-02, and thereafter, a steady increase was observed from 1.3 in 2002-03 to 1.8 in 2008-09 (Table 2).

Between the two broad groups of WPI-All Commodities, viz., WPI-Primary Articles and WPI-Manufactured Products, the range of seasonal factors for Primary Articles prices indicated higher seasonal fluctuations. The intra-year seasonal variation in these two price series had narrowed down from 2001-02 onwards (Table 2 and Chart 11).

Within the Primary Articles group, the prices of Food Articles observed substantial reduction in the intra-year seasonal variation throughout the reference period. In case of WPI-Non-Food Articles, a steady increase in seasonal fluctuation was observed from 2005-06 onwards. Although the intra-year seasonal variation in WPI of ‘Food Articles’ and ‘Non-Food Articles’ had narrowed down during 1998-99 to 2007-08; the divergence was observed in 2008-09 due to increase in seasonal variation in Non­Food Articles (Table 2 and Chart 12).

Within the Food Articles, the Pulses group displayed more seasonal variation, in terms of range of seasonal factors, than the Cereals group. The seasonal fluctuations of WPI-Pulses showed mixed patterns of movements during the reference period. From 2000-01 onwards, the range of seasonal factors for WPI-Pulses had decreased considerably (Table 2 and Chart 13). In the case of WPI-Cereals, the range of seasonal factors moved within the narrow band of 1.3 and 1.6, during this period. Among the Cereals group, seasonal fluctuations of WPI-Rice had steadily decreased, over the reference period (Table 2).

In the case of CPI for Agricultural and Rural labourers, similar seasonal variation had been observed during the reference period. During the period 1999-00 to 2002-03, both the series showed a sharp decline in the range of seasonal factors, and thereafter, from 2004-05 onwards, a moderate increase in seasonal variation was exhibited for both the series (Table 2 and Chart 14).

During the period 1999-00 to 2008-09, CPI-IW exhibited more seasonal variation than CPI-UNME. From 2004-05 onwards, the range of seasonal factors for both the series was steadily increasing, and the divergence in seasonal fluctuation between the two series was also noticed (Table 2 and Chart 15).

The range of seasonal factors of CPI-IW has decreased steadily from 2000-01 to 2004-05, and then increased in the last few years. Over the reference period, the seasonal variations of CPI-IW continued to be higher than that of WPI-All Commodities (Table 2 and Chart 16). Although the degree of divergence in the seasonal variations of the two series had narrowed during the period 2003-04 to 2005-06, some indications of marginal widening of the gap was visible in the recent years.

3.3. Index of Industrial Production (IIP)

In the case of Index of Industrial Production (IIP), both sectoral and use-based classifications were considered for analysing the intra-year seasonal variation. Among the 17 manufacturing industry groups of sectoral classification, 13 industry groups, which exhibited seasonality, were selected for analysis purpose. Apart from these industries, ‘Production of Steel’, ‘Production of Cement’ and ‘Production of Commercial Motor Vehicles’ were also selected for the study.

3.3.1 Seasonal Factors

During 2008-09, IIP-General Index had observed its peak in March and trough in April (Table 1). Although the peak of IIP-General Index in March had remained same over the years; the trough varied in the months of April, June and September (Table 3). During the reference period, the seasonal peaks of most of the industrial production series converged in the month of March. However, the seasonal troughs were varied in the months of February, April and October. The major driver of industrial sector, namely, Manufacturing exhibited its seasonal peak in the month of March, during 2008-09, and trough in April (Table 1 and Chart 17). The other two sectors, viz., Mining and Electricity, exhibited peaks in March and May, respectively, and troughs in September and February. Over the years, the peak of these three series has occurred in the month of March.

In respect of use-based classification, IIP of Basic Goods, Capital Goods and Intermediate Goods industries exhibited peaks in March and troughs in the months of September, April, and November respectively, during 2008-09 (Table 1 and Chart 18). Among these three industrial goods, the seasonal fluctuations were the highest in respect of Capital Goods.

During 2008-09, the peaks of IIP Consumer Durable and Consumer Non­Durable Goods occurred in the months of October and January, respectively, whereas,troughs occurred in December and October (Table 1 and Chart 19). Over the reference period, the Consumer Non-Durable Goods registered peaks in January, whereas, those of Consumer Durable Goods varied during the months of March and October.

During 2008-09, both the Cement and Steel Productions exhibited high degree of seasonal fluctuations. The peaks for ‘Cement’ and ‘Steel’ Productions occurred in the same month of March, whereas, the troughs occurred in the months of August and June, respectively (Table 1 and Chart 20).

3.3.2 Range of Seasonal Factors

The range of seasonal factors of IIP-General Index had declined during the period 1995-96 to 2003-04, and thereafter, it showed upward movement (Table 2). Among the three major sectors of IIP, viz., Mining, Manufacturing and Electricity, IIP-Mining exhibited the highest seasonal variation, whereas, IIP-Electricity exhibited the lowest seasonal variation (Table 2 and Chart 21). During the period 2002-03 to 2006-07, the seasonal variation of IIP-Mining increased substantially as compared to that of IIP-Manufacturing. However, IIP-Electricity witnessed a decline in seasonal variation during this period. During the last two years, the three industries showed moderate seasonal fluctuation.

Among the selected sub-groups of the Manufacturing sector, ‘Food Products’ had shown much higher seasonal variations than the other sectors, whereas, ‘Rubber, Plastic, Petroleum and Coal Products’ had the lowest seasonal variation. The four industries, viz., ‘Food Products’ (weight: 9.1 per cent), ‘Basic Chemicals & Chemical Products (except products of Petroleum & Coal)’ weight: 14.0 per cent), ‘Machinery and Equipment other than Transport equipment’ (weight: 9.6 per cent) and ‘Basic Metal and Alloy Industries’ (weight: 7.5 per cent), cover 50 per cent of the Manufacturing industry. Among these industries, ‘Basic Metal and Alloy Industries’ exhibited decline in seasonal variation during the last two years, whereas, the other three industries exhibited upward movement in the seasonal variation (Table 2 and Chart 22).

Among the use-based classification of industries, ‘Intermediate Goods’ industries experienced the lowest seasonal variations, throughout the reference period. However, IIP-Capital Goods exhibited the highest seasonal variations in all the years, except in 1999-00 and 2000-01. It was also observed from the range of seasonal factors that the magnitude of seasonal fluctuation of ‘Consumer Goods’ industry was mainly influenced by that of Consumer Non­Durable Goods. Among all these industries, ‘Capital Goods’ and ‘Consumer Durable Goods’ industries exhibited increase in seasonal variation during 2004-05 to 2008-09 (Table 2 and Chart 23).

The ranges of seasonal variations of productions of ‘Cement’ and ‘Steel’ exhibited opposite movements, over the reference period. For the last three years, the degree of seasonality of these two series had narrowed down (Table 2 and Chart 24).

During the last three years, the variation in Cement Production was almost same, whereas, Steel Production showed moderate increase in variation. The seasonal variations in ‘Production of Commercial Motor Vehicles’, which increased during 1995-96 to 1999-00, declined from 2000-01 to 2005-06. However, some moderate increase in seasonal variations of ‘Production of Commercial Motor Vehicles’ was observed in the last three years (Table 2).

3.4. Services Sector Indicators

3.4.1 Seasonal Factors

Seasonal factors in respect of select Services Sector indicators, viz., ‘Cargo Handled at Major Ports’ and ‘Revenue on Railway Freight Traffic’ exhibited peaks in March during 2008-09, whereas, the troughs occurred in June and September, respectively (Table 1 and Chart 25). Over the reference period, the peak of these indicators occurred during March.

3.4.2 Range of Seasonal Factors

The seasonal variation, in terms of range of seasonal factors, for ‘Revenue on Railway Freight Traffic’ was more as compared to that of ‘Cargo Handled at Major Ports’, over the reference period (Table 2 and Chart 26).

Although, from 2001-02 to 2006-07, ‘Cargo Handled at Major Ports’ showed sharp rise in range, a moderate decline in range was observed, during the last two years. However, from 2004-05 onwards, the ‘Revenue on Railway Freight Traffic’ showed steady increase the range of seasonal factors.

3.5. External Trade

3.5.1 Seasonal Factors

During 2008-09, the peaks of Exports and Non-Oil Imports occurred in the months of March and October, respectively, whereas, the troughs occurred in the months of November and February, respectively (Table 1 and Chart 27). Over the reference period under study, the peaks and troughs of Exports occurred in the month of March and November, respectively. In the case of Non-Oil Imports, the seasonal patterns varied over the reference period (Table 1).

3.5.2 Range of Seasonal Factors

Over the reference period, Exports exhibited higher seasonal variations than Non-Oil Imports (Table 2 and Chart 28).

Although, the ranges of the seasonal variations for Exports and Non-Oil Imports had narrowed down from 1995-96 to 1999-00, the degree of divergence increased to some extent, during 2000-01 and 2004-05. However, the decrease in the degree of divergence was evident once more during the last four years.

3.6. Inter-Sectoral Comparison of Seasonal Factors

Apart from the evaluation of the salient features (in terms of magnitude and direction) of seasonal factors for the selected series, a comparative study was done for throwing some light on the inter-sectoral performance among monetary and banking sectors, prices, industrial and external sectors. Some of the salient features, in terms of the movement of range of seasonal factors throughout the reference period, are discussed in this section.

The ranges of seasonal factors of Broad Money (M₃) and WPI-All Commodities revealed similar movements from 1995-96 to 2002-03 (Table 2 and Chart 29). From 2003-04 onwards, the increase in degree of seasonality of Broad Money was more than that in WPI-All Commodities. During the last three years, the seasonal variation in each of the two series remained at the same level.

The ranges of seasonal factors of Bank Credit (SCBs) and Index of Industrial Production exhibited similar movements over the reference period (Table 2 and Chart 30). Further, Index of Industrial Production had higher seasonal variation than that of Bank Credit (SCBs). During the last three years, both of the series exhibited stable seasonal variation.

During 1995-96 to 1998-99, the ranges of seasonal factors for Exports, Imports and Index of Industrial Production showed declining movement. Although, from 1999-00 onwards, the stability in seasonal variation of Index of Industrial Production was noticed, Imports exhibited steady increase in seasonal variation. In the case of Exports, after a steady increase during 2000-01 to 2004-05, the range of seasonal factors declined moderately during 2005-06 to 2008-09 (Table 2 and Chart 31).