NSCB - Statistically Speaking... What LEIS ahead?..

What LEIS ahead?..²

The NSCB generates the composite leading economic indicator (LEI) as a one quarter-ahead indicator of the overall direction of the economic activity for the country. The LEI makes use of 11 leading economic indicators (their individual business cycles to be exact) that tend to anticipate the changes in the direction of the overall economic activity.

But which economic variable do we use as a reference indicator to measure aggregate economic activity? I’m guessing you guessed GDP. You are correct…except that, in our LEI system, we drop the agriculture component because its cycle follows a different pattern from that of GDP ³. But the cycle of the non-agriculture component of GDP, i.e., the cycles of industry and services GVA (Gross Value Added) combined, does. So we use it as the reference series that depicts the cycle of the overall economic activity, the direction of which, is what the composite LEI tries to predict one quarter in advance.

Now how exactly do we combine the 11 leading indicators to come up with the composite LEI? Here’s the math…in words - the composite LEI is just the linear combination of the business cycle of the 11 indicators using correlation coefficients of the indicators with the non-agriculture GVA as weights. You can think of the indicators as 11 “tingtings” tied together to make up a stable, more reliable “walis tingting” that is the composite LEI, except that we use it as a telescope of sorts into the future rather than to sweep our backyards with.

But what’s a business cycle?. Economists think that the overall economy moves in a repeating pattern that includes periods of growth (or expansion) followed by periods of decline (or contraction). This wave-like pattern, characterized by peaks and troughs, is known as the business cycle and it does not have any fixed, predictable length (Figure 1). Primarily, the composite LEI’s purpose is to help business and economic planners see where the economy is headed in the cycle, and, in this sense, it is a “forecast” of what lies ahead.

Figure 2 shows a segment of the composite LEI (continuous line) and the business cycle for non-agriculture GVA (broken line) as computed for the second quarter 2007 LEI. Notice that the composite LEI extends one quarter ahead of the non-agriculture GVA’s cycle. Aha!…that extra line segment shows the forecast for the direction of the overall economic activity for the second quarter 2007…an upward rise from the first quarter. Even better days to come, one might say!

Here are some uses of the LEI ⁴. The results of the quarterly computation of LEI are inputs to the GDP forecasting work of NEDA and BSP. The National Planning and Policy Staff (NPPS) of NEDA uses the LEIS, along with a consensus forecast, to benchmark forecast GDP from its internal forecasting model called the Quarterly Growth Indicators System (QGIS). The BSP’s Department of Economic Statistics (DES), meanwhile, uses the values and quarter-on-quarter slopes (Table 1) of the composite LEI to generate a one-quarter ahead forecast of GDP that is used, among others, by the Monetary Board in inflation targeting. Needless to say, the LEI is a helpful tool used by our economic planners to track future movements of the economy.

In the same light, guidance on economic trends is invaluable for businesses making major strategic decisions. A multinational company’s decision to set up a plant in the Philippines may hinge on whether the economic climate in the Philippines is favorable or unfavorable in its initial years of operation. For Juan de la Cruz, meanwhile, he may use the LEI in making investment decisions and ponder how prevailing and emerging economic trends will affect him. Investing in stocks, for example, may be good or bad, depending on the current point in the business cycle.

Let’s talk forecasting methods a bit. The robustness of the LEI as an input to economic forecasting has its roots in its simple approach. Macroeconomic models, which are usually regression-based, assume an underlying parametric structure that may be limited when the model is not correctly specified and the assumed functional relationships cease to hold. In comparison, leading economic indicator systems need not assume a parametric model in order to make a forecast of the future direction of the economy ⁵. Hmmmnnn…that only means that with the LEI approach, we have fewer constraints in including indicators that may be used as explanatory variables to “predict” economic trends. Needless to say, both the LEI approach and macroeconomic models have their pros and cons and the LEI may be used as inputs to macroeconomic models.

Wait, we’ve been talking about leading economic indicators but which ones are they?. The 11 leading economic indicators ⁶, composed of both economic and financial variables, are stock price index, exchange rate, money supply, consumer price index, merchandise imports, tourist arrivals, terms of trade new businesses, hotel occupancy, electric energy consumption, and wholesale price index. These indicators were drafted based on five economic rationales ⁷ - Production time, ease of adaptation, market expectation, prime movers and change vs. movers. You might want to grab an economics book at this point, but it suffices to say that these criteria were developed based on existing theories as well as empirical studies on leading indicators and business cycles.

The cycles of each of the leading indicators lead the direction of the reference cycle, i.e., that of the non-agriculture GVA. This “lead” is measured in number of quarters called “lead periods” which are identified as the number of quarter(s) when the cycles of an indicator show the highest significant correlation with the cycle of the reference series (Table 2).

As we said earlier, the composite LEI is just the a linear combination of the business cycles of these 11 indicators using correlation coefficients of the indicators with the non-agriculture GVA as weights. The contribution of each indicator is computed as the slope (or the quarter-on-quarter change in business cycle values) multiplied by its weight (the correlation coefficient), and taking its linear relationship ⁸ with the reference series into account, either pushes the LEI up (positive contribution) or pulls the LEI down (negative contribution) (Table 3).

Now, that’s a mouthful.

Let’s apply that to the stock price index, which was the largest positive contributor to the 2nd quarter 2007 composite LEI (Table 2). It’s contribution is computed to be 0.844 (weight or correlation coefficient of 0.422 x slope of 2.001) The positive coefficient value indicates that the stock price index has a direct relationship with the non-agriculture GVA, meaning that it contributes positively to the composite LEI and to the positive outlook for the economy.

In the second quarter 2007 LEIS (latest LEI report), the positive contributors outweighed the negative contributors. The eight positive contributors – beginning with the largest positive contributor – were stock price index, exchange rate, money supply, consumer price index, merchandise imports, tourist arrivals, terms of trade and new businesses. The three negative contributors - beginning with the largest negative contributor – were, hotel occupancy, electric energy consumption, and wholesale price index.

The computed composite LEI for the second quarter 2007 increased to 0.294 from the 0.151 in the first quarter - a quarter-on-quarter slope of 0.143 representing a rise in LEI (Figure 2). Now we know (I hope) that the upward and downward movements of the composite LEI signal the direction of the economy one quarter ahead. Thus, we can say that the increase in the composite LEI for the second quarter of 2007 signals an optimistic outlook for the period.

Backtracking to Q1 2007, the composite LEI also showed a steep rise to 0.151 for first quarter 2007 from 0.035 in the fourth quarter of 2006. This behavior is reflected in the outstanding GDP growth of 6.9 percent registered for the first quarter of 2007 and supports the observation that there are signs of better economic times ahead,..at least in the short term.

The Philippine Leading Economic Indicators System (LEIS) was developed jointly by the National Statistical Coordination Board (NSCB) and the National Economic and Development Authority (NEDA) in 1995 to serve as a basis for short-term forecasting of the macroeconomic activity in the country. Two years after, in 1997, the NSCB started compiling data for the leading economic indicators and officially generated the composite LEI on a quarterly basis.

We in the NSCB have been continually reviewing the LEIS in order to improve it. Users of the LEIS need to be mindful, of course, that leading economic indicator systems, just like other statistical tools, have their limitations.

For one, in our desire to release the LEIS report two months before the end of the reference quarter to enhance its usefulness, we forecast data to impute for the missing months or quarters of the 11 indicators that must be available based on the individual indicator lead period. As such, the values of the leading indicators are subject to revisions that may change the forecasted direction of the economy.

The pool of leading indicators likewise needs reviewing to be able to capture structural changes that may have occurred in the economy, and to address data issues such as timeliness and availability. The NSCB SANA-LEIS ⁹ Team is tasked to come up with an improved LEIS in time for the release of the first quarter 2008 LEI report.

Moreover, there is also a need to map better the peaks and troughs (or turning points) of the Philippine business cycle that define the expansion and contraction of economic activities in the Philippines.

Fortunately, our rank of allies to improve the system includes time series and business cycle experts like Dr. Lisa Grace Bersales ¹⁰, with whom the NSCB worked in a project to improve the LEIS, and who continues to actively and tirelessly help us in our work in the seasonal adjustment of the national accounts and on the leading economic indicators system.

Table 1. Composite Leading Economic Indicator (LEI): Q1 2005 to Q2 2007

Table 2. Correlation and Lead Periods of the 11 Leading Indicators
with the Non- Agriculture GVA: Q2 2007 LEIS

Table 3. Contributions of the leading economic indicators: Q2 2007 LEIS

* Inverse relationship with Non-agriculture GVA
1/ Contribution = slope x correlation coefficient
2/ Total contribution = summation of the absolute values of contribution.
3/ Share to total contribution = percentage share of the contribution of each indicator to total contribution.
4/ Share to total contribution = percentage share of contributors by type of contribution.
5/ Rank = rank of the indicators in contribution, 1 being the highest.

Figure 1. The Cycle

Figure 2. Composite LEI vs Non-Agri GVA cycle: Q1 2000 to Q2 2007

Reactions and views are welcome thru email to the author at cis.bacani@nscb.gov.ph.

______________________

¹ Secretary General of the National Statistical Coordination Board (NSCB) and Chairman of the Statistical Research and Training Center (SRTC). He holds a Ph. D. in Statistics from the University of Michigan in Ann Arbor, USA and has taught mathematics and statistics at the University of the Philippines. He is also a past president of the Philippine Statistical Association.

² This article was written by Christopher Ivo S. Bacani (cis.bacani@nscb.gov.ph), Statistical Coordination Officer III of the Economic Indicators and Satellite Accounts Division (EISAD) of the NSCB and Group leader of the LEIS Group of the NSCB SANA-LEIS Team.

³ Agriculture GVA cycle was excluded to eliminate the irregularities it may contribute to the GDP cycle.

⁴Based on discussions with Bien Ganapin of the NEDA-NPPS & Terry Deveza of BSP-DES.

⁵See Final project report, “Leading Economic Indicator System of the Philippines”, October 10, 1996, pp. 2-4.

⁶ See LEI technical notes at http://www.nscb.gov.ph/technotes/lei/lei_tech2q07.asp for a brief definition of the indicators.

⁷ See Final project report, “Leading Economic Indicator System of the Philippines”, October 10, 1996, pp. 12-44. See also, F. de Leeuw, 1991. “Towards a Theory of leading indicators,.” In K. Lahiri and G. D. Moore eds., Leading Economic Indicators: New Approaches and Forecasting Records. Cambridge: Cambridge University Press.

⁸ Also indicated by the correlation coefficient. A positive coefficient indicates a direct relationship with the Non-agri GVA while a negative value indicates an indirect or inverse relationship with the Non-agri GVA.

⁹The NSCB Seasonal Adjustment of the National Accounts and Leading Economic Indicators System Team was created in 2006 to conduct further analyses and studies to improve the SANA and LEIS.

¹⁰Dean of the School of Statistics, University of the Philippines and Chair of the NSCB Technical Committee on Seasonal Adjustment of the Philippine Time Series (TCSAPTS).

Posted 09 July 2007.


E-mail the webmaster	Terms of Use	Home • Top of Page
	1997-2008, National Statistical Coordination Board Makati City, Philippines