By Chan Huan Chiang and Ong Wooi Leng

Introduction

We take keen interest on economic news only because we realise that recessions have a way of creeping up on us and messing up whatever big plans we have for the future. Timing is everything, the way Solomon’s wisdom teaches us in the Book of Ecclesiastes – a time to plant, a time to build, a time to keep and a time to throw away. Still, some three millenniums will pass before formal studies and public announcements are made of cyclical changes occurring in the economy.

In the U.S., although the business cycle dating committee¹ was only established in 1978, the National Bureau of Economic Research (NBER) has been publishing business cycle dates since 1929. Constructing retrospectively from historical records, analysis goes back to when the economy was at the bottom in 1854.² Formal business cycle research was pioneered by Burns and Mitchell.³

Today, there is a wide variety of methods used by different agencies to detect and announce dates when the economy makes a regime change from the peak and going into recession, and from the trough to begin another period of expansion. For instance, ECFIN Economic Brief is published by the European Commission’s Directorate General for Economic and Financial Affairs.⁴ Leading indicators are also published monthly by the Organisation for Economic Cooperation and Development (OECD).⁵ In France, the Centre d’Observation Economique releases the numbers for economic indicators monthly on its website⁶ and in Australia, the Bureau of Statistics announces seasonal and trend estimates to show the direction of the underlying economy.⁷

Business cycle research in Malaysia dates back to the late 1980s through collaborative efforts between Universiti Malaya and the Institute of Developing Economies (IDE), Tokyo.⁸ The Malaysian Institute of Economic Research (MIER) subsequently began making public announcements of composite leading, coincidental and lagging indicators in the early 1990s. Today these indicators are published monthly by Malaysia’s Department of Statistics.⁹

Publication of such economic indicators is meant to provide timely signals of changing economic conditions. Policy makers could plan ahead for countercyclical policy either to stem inflation or cushion a fall, while the business sector can adjust sales and investments strategies accordingly.

Turning points

Economic indicators are not the same as the gross domestic product or GDP, which is the traditional measure for economic performance in any particular country. Reports of GDP are made retrospectively. For instance, the 2010 GDP numbers for Malaysia would only be officially announced towards the end of March 2011 when the Bank Negara Report is released. Even so, the figures will even then remain preliminary which means that it is subject to further revisions.

The challenge in analysing economic indicators is the detection of turning points, which are defined as beginning a downturn from the peak or an upturn from the trough. The problem is the numbers of an economic indicator can be quite volatile, jumping up and down frequently, making it hard to tell true peaks and true troughs from seemingly random shifts in the data. True turning points are generally accepted as those that satisfy the three-D conditions: depth, duration and diffusion, or three-P conditions: pronounced, persistent and pervasive.¹⁰ When the economy makes a turn, the change must be significant yet lasting and widespread.

Different methods adopted for analysing business cycles thus make use of different sets of rules before a peak or a trough is accepted as such when the required conditions are met. They then go on to verify the number of times false peaks and troughs were identified by those rules, suggesting that even the best methods remain prone to error.

Index of industrial production

The index of industrial production for the country, published by the Department of Statistics, is a well referred set of numbers representative of economic performance because of their timeliness. Unlike the GDP which is one number, oft en revised, for the whole year, the industrial production index provides fairly up-to-date information of monthly ups and downs. There is only about two months of lag.¹¹

The 3/12 and 12/12 pressure curves, shown in Figure 1, are signals calculated on the basis of three- and 12- month running totals of the industrial production index divided by the value of the totals 12 months before.¹² For example, the May 2008 3/12 signal is the sum from March 2008 through May 2008 divided by the May 2007 sum (obtained by summing from March 2007 through May 2007). The May 2008 12/12 signal is the sum from June 2007 through May 2008 divided by the May 2007 sum (obtained by summing from June 2006 through May 2007).

Both the 3/12 and 12/12 signals attempt to measure changes of the economic circumstance from 12 month before and thus an annual rise will be indicated by a value above 100 and an annual fall by a value below 100. The 3/12 is a short-term (three months) change signal while the 12/12 is a longer term signal.

Figure 1: 3/12 and 12/12 pressure signals of Malaysia’s industrial production index, 2004–2010

Thus a sharp rise followed by an equally sharp fall occurring within a 12 month period would not be detected in the 12/12 signal. Such within one year shocks will, however, be picked up as a 3/12 signal as can be seen on Figure 1, where the more volatile 3/12 signal meanders about the smoother 12/12 signal.

These two signals can be used as simple leading and coincidental indicators of the economy in which the shorter term “mood swings” offer a possible early warning as to what might likely happen, “structurally”, over the longer term year-by-year assessment of economic performance. When the economy is on an expansion, the 3/12 signal is higher than the 12/12, signalling that the current short-term (three-month) circumstance is more favourable than the past year taken as a whole. This is the “high” signal that can be seen in Figure 1 before the end of 2003.

At the top of the expansion at the beginning of 2004, the 3/12 signal made a turn downwards before the 12/12, signalling that the overheating economy was peaking as short-term economic circumstance were showing less favourable signs when compared over the past year. Given that the 3/12 turns downwards ahead of the 12/12 signal, there came a point, around July 2004 when the 3/12 dipped below the 12/12. This is the signalling point of confirmation that the “peak” had indeed occurred and the economy was on its downward path. The 3/12 signal below the 12/12 says that short term falls have become more acute than the overall year-by-year fall.

The retraction of the growth cycle, observed was a prolonged 40 months that lasted till the ending months of 2007. Between the “low” detected in July 2007 and the trough there were about four months of early warning. So long as the signal values are above 100, economic circumstance remained better than the previous year. Growth rates have come down but remained positive. This was merely a retraction of the growth cycle but not a retraction of the actual business cycle.

The respite was, however, only very brief, because the 3/12 signal turned in February 2008 and by midyear the peak was confirmed to begin another cycle of retraction. When, indeed the signal values fell below 100, the real “low” signal of business cycle contraction of the economy is detected.

Growth was now negative. This period of recession lasted 16 months ending at the trough when the economy bottomed out in August 2009 indicated by the 3/12 signal overtaking the 12/12. There was six months of early warning given by the 3/12 signal turning upwards in February 2009. The next round of expansion also appeared brief, since the “high” signal was detected in April 2010 and the peak confirmed in August 2010.

Figure 2: 3/12 and 12/12 pressure signals of Penang export figures, 2004–2010

Figure 3: 3/12 and 12/12 pressure signals of total employment in Penang, 2004–2010

The potential for ground-up monitoring of Penang’s local economy

Although it is accepted that economic indicators bear little resemblance to the GDP, they offer great potential for ground-up monitoring of the local economy in Penang. GDP numbers for the nation are officially announced far too late for any form of business decisions. Gross regional product of GRP numbers (the state equivalent of the national GDP) are announced even less frequently (as infrequent as once in five years, in the Malaysian five-year plans). It makes sense, therefore, to resort to tracking, economic indicators from the ground up at the state level in Penang.

Doing this is much more challenging. There are eight indicators that make up the leading index tracked by the Department of Statistics. Six more indicators make up the coincidental index and another five, the lagging index, in other words, 19 separate data series in all.¹³ Many of these (e.g. index of industrial production, the Kuala Lumpur Composite Index of stock prices, the consumer price index, M1 money supply) are only available at the national level and though important economic indicators, they are not helpful at all in pointing specifically at the economic performance in Penang.

Some indicators specific to Penang are, however, available on a timely basis. The 3/12 and 12/12 signals for export figures through the port and airport in Penang is shown in Figure 2. The “high” signal (3/12 turning down) was detected around June 2004 and confirmed by the “peak” signal (3/12 dipped below 12/12) around October 2004. The growth cycle began its decline lasting four and a half years with the last one and a half being also a business cycle recession (when the signal values fell below 100).

The “low” signal was finally detected around May 2009 and the “trough” confirmed four months later in September 2009. During the preceding one year or so, there were a lot of flip-flops in the 3/12 signals that shifted directions two to three times making it difficult to call whether they were genuine. As mentioned above, the 3/12 are mood swings that may or may not show up structurally as a 12/12 signal depending on whether they reverse direction within a short space of time to cancel out one another.

It is a popular view that recessions are called based on two quarters of continuous fall in the business cycle and thus one is safe to ignore short-term flip-flops in the data. The 3/12 signal turned downwards (high signal) again in February 2010 and crossed the 12/12 around June (peak signal) 2010, lasting only 12 or so month between trough and peak. These signals pass the two quarter test but the subsequent months of data will tell whether they will turn up as false.

Noise in the data

Employment data are included as part of the coincidental index tracked by the Department of Statistics for the national economy. Attempts at calculating the 3/12 and 12/12 pressure signals were also made using employment data for Penang as shown in Figure 3.

The flip-flops of the 3/12 signal are very apparent here and as a result it becomes impossible to tell where the crossing points were between the 3/12 and 12/12 signals, that would allow us to detect peaks and troughs. Penang has had, one might say, a volatile job market. Although Penang’s economy typically follows business cycle changes, the full employment and leaning on labour scarcity situation leads to job–hopping, resulting in quickly changing numbers of total employed.

Employment numbers have grown around three per cent annually over the past five years (12/12 signals above 100) despite the clearly identified 2008–2009 recession that show up among other economic indicators. There appears to be, in other words, some form of decoupling between job placements and economic circumstance, suggesting that managing the human resource in the state becomes a critically important policy issue.

Human resource input is a major factor of economic production such that the share of labor input to achievable output is typically a fairly fixed number. Noise of this kind found in Penang’s employment figures not oft en found elsewhere warrants joint-intervention by all parties concerned: industry, business, the labour union, and policy makers.

Conclusions

Malaysia has long been managed centrally as a single economy in which regional disparity is but the only concern when different parts of the country are taken into account. This view is today quickly changing, because spatial heterogeneity is beginning to be seen as offering varying potential for economic growth which can and should be leveraged on by adopting different development strategies.

Different locations, across the country, doing things differently in pursuit of common growth for the nation, is the new growth model. Accordingly, how well each location performs needs to be independently monitored so that policy prescriptions and investments can be made and adjusted.

Notes:

1 www.nber.org/cycles/members.html

2 For example, see www.nber.org/cycles/sept2010.html; see also Oscar Jorda (2010) “Diagnosing Recessions”, FRBSF Economic Letter 2010-05, Federal Reserve Bank of San Francisco.

3 A F Burns and W C Mitchell (1946) Measuring Business Cycles, National Bureau Of Economic Research, Columbia University Press.

4 Available for download at http://ec.europa.eu/economy_finance/publications/economic_briefs/index_en.htm 

5 M J Artis, R C Bladder-Howell and W Zhang (1995) “Turning points in the international business cycles: an ex-post analysis of the OECD leading indicator series for the G-7 countries” OECD Economic Studies 24(II) 125-65

6 Jacques Anas and Laurent Ferrara (2002) “Detecting cyclical turning points: an ABCD approach and two probabilistic indicators” 26th CIRET Conference Taipei, October.

7 Australia Bureau of Statistics (2006) “Some aspects of turning point detection in seasonal adjusted estimates.” Research Paper for Methodology Advisory Committee Catalogue No.1352.0.55.079.

8 Mokhtar Tamin (1991) “Business cycle indicators for Malaysia” in Hiroshi Osada and Daisuke Hiratsuka (eds.) Business Cycles in Asia IDE Occasional Paper Series No.26, Tokyo.

9 Available online at www.statistics.gov.my/portal/index.php?option=com_content&view=article&id=385%3Afree-download&catid=61&Itemid=53〈=en

10 A Banerji (1995) “The three P’s: simple tools for economic cycles” Business Economics 34(3):72-76

11 Also available on Bank Negara Malaysia’s homepage at www.bnm.gov.my/files/publication/msb/2010/11/xls/3.5.1.xls

12 These signals were developed by the Institute of Trend Research and although simplistic they have been shown to compare favourably to much more sophisticated method such as the one used by NBER. See Michael P Neimira and Philip A Klein (1994) Forecasting Financial and Economic Cycles John Wiley and Sons, N.Y. p.195.

13 Norzalelawati Ahmad (2003) “Malaysia’s economic indicators – leading, coincidental and lagging indicators”, Workshop on Composite Indicators and Business Tendency Surveys, Bangkok, February 24–26.

** Republished with permission. This article first appeared in the February 2011 issue of the Penang Economic Monthly. Chan Huan Chiang is a senior research fellow at the Socio-Economic and Environment Research Institute. Ong Wooi Leng is a research analyst at the Socio-Economic and Environmental Research Institute (SERI).

Comments (0)

Show/Hide comments

Show/Hide comment form