## _wp03170

## Source details

**Canonical URL:** [_wp03170](https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2003/_wp03170.pdf)

## Other formats

- [Markdown version](/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2003/_wp03170.pdf.md)
- [Structured JSON version](/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2003/_wp03170.pdf.json)

---

### I. Introduction and purpose
- Timely understanding of the direction of economic activity is essential for macroeconomic policy formulation.
- Purpose of paper:
  - Present estimation of coincident and leading indicators for Jordan.
  - Provide a road map for estimating these indicators in other emerging markets.
- Structure overview:
  - Section II: literature and implementations (NBER, OECD).
  - Section III: Jordan’s economic developments in the 1990s and available data.
  - Section IV: estimation of composite index of coincident indicators.
  - Section V: estimation of composite index of leading indicators.
  - Section VI: conclusions.

### II. Jordan’s economic performance (late 1980s–2002): key facts and statistics
- Country and macro context:
  - 2002 per capita GDP averaged about $1,750.
  - Over 70 percent of value-added concentrated in services.
  - Tradable sector (exports + imports of goods and nonfactor services) = 106 percent of GDP in 2002.
  - Since October 1995: Jordanian dinar effectively pegged to U.S. dollar.
  - 1993–1999: economic growth averaged 4 percent.
  - Inflation reduced to single digits by 1991 and averaged 3 percent thereafter.
  - Despite negative shocks (renewed Palestinian-Israeli conflict and September 11, 2001 events), economic growth accelerated to 4.2 percent in 2001 and to 4.9 percent in 2002.
- Export drivers for 4.9 percent in 2002:
  - Accession to the WTO in 2000.
  - Duty- and quota-free access to U.S. markets under the 1994 Peace Agreement with Israel (Qualified Industrial Zones require a minimum value added from both Israeli and Jordanian sources).
  - The 2001 Free Trade Agreement with the United States.
  - Major export destination: The United States.
- Selected economic indicators (1993–2002) — preserved exactly as reported:
  - Real GDP at market prices (annual percent changes): 4.5, 5.0, 6.2, 2.1, 3.3, 3.0, 3.1, 4.2, 4.2, 4.9
  - GDP deflator at market prices (annual percent changes): 2.8, 6.9, 1.9, 2.1, 1.2, 6.0, -0.2, -0.1, 0.0, 0.5
  - Nominal GDP at market prices (annual percent changes): 7.4, 12.2, 8.2, 4.2, 4.6, 9.2, 2.8, 4.1, 4.3, 5.3
  - Consumer price index (annual average): 3.3, 3.6, 2.3, 6.5, 3.0, 3.1, 0.6, 0.7, 1.8, 1.8
  - Consumer price index (end of period): 1.6, 5.1, 4.1, 2.6, 6.3, -0.8, 2.8, -1.9, 3.8, 0.5
  - Nominal GDP at market prices (in millions of JD): 3,925; 4,401; 4,774; 4,982; 5,138; 5,610; 5,767; 6,002; 6,260; 6,594
  - Gross domestic investment (percent of GDP): 36.6, 33.3, 33.0, 30.5, 25.7, 21.8, 21.6, 22.1, 22.4, 23.0
    - Government (percent of GDP): 6.3, 6.0, 7.0, 7.3, 5.6, 6.0, 5.8, 5.2, 5.9, 6.8
    - Other (percent of GDP): 30.4, 27.3, 25.9, 23.2, 20.1, 15.8, 15.9, 16.9, 16.5, 16.3
  - Gross national savings (percent of GDP): 30.4, 31.3, 33.5, 32.0, 30.4, 25.8, 29.8, 22.8, 22.4, 27.9
    - Government (percent of GDP): 5.7, 4.6, 3.1, 4.4, 3.1, 0.1, 2.2, 0.5, 2.2, 1.7
    - Other (percent of GDP): 24.7, 26.6, 30.4, 27.6, 27.3, 25.7, 27.7, 22.3, 20.2, 26.2
  - Savings/investment balance (percent of GDP): -11.5, -6.4, -3.8, -3.2, 0.4, 0.3, 5.0, 0.7, -0.1, 4.9
    - Government: -0.5, -1.4, -3.9, -2.8, -2.5, -6.0, -3.5, -4.7, -3.7, -5.0
    - Other: -10.9, -5.0, 0.1, -0.4, 2.9, 6.2, 8.5, 5.4, 3.6, 10.0
  - Revenue and grants (percent of GDP): 34.8, 32.6, 34.4, 33.6, 31.5, 30.4, 31.0, 30.0, 30.8, 30.5
    - Of which: grants (percent of GDP): 4.1, 4.0, 3.6, 4.5, 4.4, 3.7, 3.5, 4.2, 4.4, 5.2
  - Expenditure and net lending (incl. off-budget accounts) (percent of GDP): 36.4, 34.5, 36.0, 36.6, 34.5, 36.8, 34.9, 34.7, 34.5, 35.5
  - Overall fiscal balance (after grants) (percent of GDP): -0.5, -1.4, -3.9, -2.8, -2.5, -6.0, -3.5, -4.7, -3.7, -5.0
  - Government and government guaranteed debt (percent of GDP) 1/: ............, 94.3, 104.5, 111.3, 99.7, 97.8, 101.4
    - Of which: external: ............, 89.2, 89.3, 95.5, 84.0, 79.4, 81.2
  - Current account balance (after grants) (percent of GDP): -11.5, -6.4, -3.8, -3.2, 0.4, 0.3, 5.0, 0.7, -0.1, 4.9
  - Domestic merchandise exports (annual percent changes in U.S. dollar terms): 7.0, 13.8, 26.2, 2.3, 2.6, -1.9, 0.5, 2.8, 25.1, 13.7
  - Merchandise imports (annual percent changes in U.S. dollar terms): -6.6, -4.6, 9.5, 16.2, -4.4, -6.7, -3.3, 23.7, 5.6, 2.7
  - Net foreign assets (changes in percent of beginning of period broad money): -1.7, 2.8, 2.5, 0.4, 8.1, 1.9, 10.4, 12.6, 1.8, 5.4
  - Broad money (annual percent changes): 8.3, 8.1, 6.5, 0.3, 7.8, 7.6, 12.0, 10.2, 5.8, 7.0
  - Nominal per capita GDP (in U.S. dollars): 1,404.1; 1,506.7; 1,568.6; 1,559.0; 1,575.3; 1,663.8; 1,660.1; 1,679; 1,703; 1,746
  - In months of prospective imports of GNFS 4/1.: 1.9, 1.2, 1.1, 1.8, 4.6, 3.3, 4.8, 6.4, 6.1, 7.5
  - As percent of Jordanian dinar broad money: 10.6, 6.8, 6.6, 11.1, 25.0, 16.9, 25.6, 33.2, 30.1, 38.1
  - Net International Reserves (in millions of U.S. dollars) 2/: 612; 463; 534; 442; 1,508; 967; 1,463; 2,275; 2,111; 3,032
  - U.S. dollar per Jordanian dinar (period average): 1.44, 1.43, 1.43, 1.41, 1.41, 1.41, 1.41, 1.41, 1.41, 1.41

### III. Data sample, seasonal treatment, and stationarity
- Data sample for composite indicators:
  - Monthly observations of 40 variables from real, fiscal, monetary, and external sectors.
  - Sample period: January 1996 to December 2002 (84 observations).
- Seasonal adjustment and Muslim holidays:
  - Gregorian-calendar seasonal adjustment: U.S. Census Bureau X12.
  - Residual seasonality associated with Eid-ul-Fitr and Eid-ul-Adha addressed by dummy-variable regression (following Alper and Aruoba (2001)).
  - Eid-ul-Fitr dummy significant only for: industrial production, construction activity, reserve money, real imports, and the trade balance.
  - Eid-ul-Adha dummy insignificant for all variables.
- Stationarity and transformations:
  - Variables tested in logarithms using Phillips-Perron tests.
  - First differences (growth rates) found to be stationary; regressions expressed in first differences.
  - Industrial production index chosen as reference series for the state of the economy (validation: correlation with real GDP — 88 percent trended, 45 percent detrended).

### IV. Methodological approach and model specification
- Departure from ad hoc selection: econometric approach with statistical inference and parsimony.
- Variable selection: recursive elimination of variable with lowest T-ratio (as in Stock and Watson (1989)), with attention to multicollinearity.
- Reduced-form regressions estimated instead of dynamic factor models due to limited sample.
- Error specification: MA(1) error term; standard errors and covariance matrices estimated using Newey-West heteroskedastic-consistent procedure.

### V. Coincident Economic Indicator (CEI) — estimation results and interpretation
- Estimated reduced form:
  - ∆LRIIP_t = α + β' ∆LCI_t + u_t
  - u_t = ε_t + θ ε_{t-1} (MA(1))
- Selected coincident indicators (five variables): TOT, TB, IMPCG, EMPTAX, COPERM.
- Statistical results (Table 2; dependent variable: IIP) — preserved exactly:
  - Sample (adjusted): 1996:02 2002:12
  - Included observations: 83 after adjusting endpoints
  - Convergence achieved after 10 iterations
  - Newey-West HAC Standard Errors & Covariance (lag truncation=3)
  - Backcast: 1996:01
  - Coefficients and inference:
    - C: Coefficient 0.004075  Std. Error 0.001332  t-Statistic 3.058719  Prob. 0.0031
    - TOT: Coefficient 0.506743  Std. Error 0.138262  t-Statistic 3.665089  Prob. 0.0005
    - TB: Coefficient 0.561191  Std. Error 0.173907  t-Statistic 3.226965  Prob. 0.0018
    - IMPCG: Coefficient 0.067893  Std. Error 0.024072  t-Statistic 2.820390  Prob. 0.0061
    - EMPTAX: Coefficient 0.053902  Std. Error 0.018522  t-Statistic 2.910185  Prob. 0.0047
    - COPERM: Coefficient 0.158926  Std. Error 0.024182  t-Statistic 6.572115  Prob. 0.0000
    - MA(1): Coefficient -0.739768  Std. Error 0.062502  t-Statistic -11.835910  Prob. 0.0000
  - Goodness-of-fit and diagnostics:
    - R-squared 0.667357
    - Adjusted R-squared 0.641095
    - Mean dependent var 0.003213
    - S.D. dependent var 0.065715
    - S.E. of regression 0.039369
    - Akaike info criterion -3.551120
    - Schwarz criterion -3.347122
    - Log likelihood 154.3715
    - Sum squared resid 0.117792
    - F-statistic 25.41216
    - Prob(F-statistic) 0.000000
    - Durbin-Watson stat 1.790883
    - Inverted MA Roots .74
  - Note: while θ is close to unity in absolute value, the null hypothesis θ = 1 can be rejected at the 5 percent confidence interval.
- Economic interpretation:
  - External factors (terms of trade and trade balance) are important determinants of current activity.
  - IMPCG proxies investment (capital goods imports).
  - EMPTAX proxies fiscal-sector response via payroll deductions.
  - COPERM proxies construction activity (procyclical).
  - U.S. and Euro area global indicators not significant in CEI.

### VI. Leading Economic Indicator (LEI) — estimation results and interpretation
- Lead horizon: six months; specification:
  - ∆LRIIP_t+6 = α + β' ∆LLI_t + u_t
  - u_t = ε_t + θ ε_{t-1} (MA(1))
- Selected leading variables: CLAIMSPS, SPREAD3M, RESERVUSD, DEXP, ASTOCEXCH.
- Statistical results (Table 3; dependent variable: IIP(+6)) — preserved exactly:
  - Sample (adjusted): 1996:02 2002:06
  - Included observations: 77 after adjusting endpoints
  - Convergence achieved after 19 iterations
  - Newey-West HAC Standard Errors & Covariance (lag truncation=3)
  - Backcast: 1996:01
  - Coefficients and inference:
    - CLAIMSPS: Coefficient 0.967311  Std. Error 0.156717  t-Statistic 6.172354  Prob. 0.0000
    - SPREAD3M: Coefficient -0.001210  Std. Error 0.000348  t-Statistic -3.480252  Prob. 0.0009
    - RESERVUSD: Coefficient 0.113004  Std. Error 0.023433  t-Statistic 4.822525  Prob. 0.0000
    - DEXP: Coefficient 0.176229  Std. Error 0.055984  t-Statistic 3.147841  Prob. 0.0024
    - ASTOCEXCH: Coefficient 0.084349  Std. Error 0.056545  t-Statistic 1.491731  Prob. 0.1402
    - MA(1): Coefficient -0.980252  Std. Error 0.016004  t-Statistic -61.24882  Prob. 0.0000
  - Goodness-of-fit and diagnostics:
    - R-squared 0.489537
    - Adjusted R-squared 0.453589
    - Mean dependent var 0.003078
    - S.D. dependent var 0.067135
    - S.E. of regression 0.049626
    - Akaike info criterion -3.093884
    - Schwarz criterion -2.911250
    - Log likelihood 125.1145
    - Sum squared resid 0.174855
    - Durbin-Watson stat 1.933549
    - Inverted MA Roots .98
- Statistical significance:
  - All variables significant at the 5 percent confidence interval except ASTOCEXCH (significant at the 15 percent level).
- Economic interpretation:
  - Monetary indicators: CLAIMSPS (real credit to private sector) positive; reductions in SPREAD3M (interest spread) positive.
  - External indicators: RESERVUSD (net usable reserves) and DEXP (demand for domestic exports) positive.
  - Equity market: ASTOCEXCH rise indicates increased confidence; weaker statistical significance.
  - U.S. and Euro area leading indicators not significant for LEI.

### VII. Forecasting performance, caveats, and notable out-of-sample result
- LEI forecasting and out-of-sample exercise:
  - Regression estimates used to forecast growth rates of industrial production from January to June 2003.
  - Projections in the source suggest a slowdown in economic growth (seasonally adjusted) for Q1 2003.
- Actual outcomes noted in source:
  - Industrial production contracted by 14 percent in Q1 2003.
  - Real GDP growth slowed to 2.8 percent in Q1 2003, compared to 4.7 percent in Q4 2002.
  - Contraction in industrial production in February 2003 exacerbated by a snow storm that halted production for about a week.
- Caveats:
  - Confidence interval for forecasts is relatively wide and increases with forecast horizon (about 50 percent of variation in activity unexplained by the LEI regression).
  - Some value judgment required; geopolitical shocks (e.g., war in Iraq) and other qualitative information are difficult to capture quantitatively.

### VIII. Conclusions and operational guidance
- Composite CEI and LEI provide useful summary statistics for current and future economic direction.
- Econometric approach produced statistically significant and economically intuitive CEI and LEI for Jordan despite small sample.
- Limitations and recommendations:
  - A larger sample is likely necessary before indexes can be relied upon for more than qualitative assessment.
  - The proposed indexes should be considered experimental and used operationally only after a testing period confirming out-of-sample performance.
  - Forecasts must incorporate qualitative information, including geopolitical uncertainties and macroeconomic policy changes.
  - Regular updating of these indexes could provide a useful tool for the Jordanian authorities.
  - The econometric roadmap may be useful for other emerging markets.

*Source: _wp03170 - References (IMF staff document).*

### References..............................................................................................................

### References

### I. Introduction
- Timely understanding of the direction of economic activity is essential for macroeconomic policy formulation.
- Relevant statistics to judge economic direction are often available with considerable lag, especially in emerging markets where:
  - publication lags are longer than in advanced economies, and
  - alternative sources (e.g., confidence indicators) are usually lacking.
- Purpose of paper:
  - Present estimation of coincident and leading indicators for Jordan.
  - Provide a road map for estimating these indicators in other emerging markets.
- Structure:
  - Section II: literature and implementations (NBER, OECD).
  - Section III: Jordan’s economic developments in the 1990s and available data.
  - Section IV: estimation of composite index of coincident indicators.
  - Section V: estimation of composite index of leading indicators.
  - Section VI: conclusions.

### II. Literature on Coincident and Leading Indicators
- Historical background:
  - Burns and Mitchell (1946) pioneered identification of series that lead, coincide, or lag U.S. business cycle turning points.
  - Moore and Shiskin (1967) developed a formal scoring/weighting scheme based on economic significance, statistical adequacy, cyclical timing, and business cycle conformity; choice of variables and weights was subjective.
  - In the U.S., production of composite indexes moved from the Department of Commerce to the Conference Board in December 1995.
- OECD approach:
  - Began publishing leading indicators in 1987 for member countries and six aggregate zones.
  - OECD and NBER-Conference Board methods are similar (turning-point analysis) but differ on trend adjustments: NBER-CB does not rely on trend adjustments; OECD uses a modified Phase Average Trend (PAT) method.
- Criticisms of traditional methods:
  - Limited economic-theory guidance in indicator selection ("measurement without theory").
  - Subjectivity in methodology.
- Econometric advances:
  - Stock and Watson (1989):
    - Define composite coincident index as a single unobserved variable, “the state of the economy.”
    - Use dynamic factor analysis and maximum likelihood estimation.
    - Derive composite leading index by forecasting the coincident index with a vector autoregressive model, selecting seven leading indicators from over 250 candidates.
    - Despite sophistication, Stock and Watson indexes failed to predict the 1991 U.S. recession; the link between individual indicators and composite indexes is less transparent in dynamic factor models.
- Coincident and leading indicators in emerging markets:
  - Less developed literature; contributing factors include:
    - Predominance of agriculture and weather-driven cycles (Mall, 1999).
    - Limitations in data quality and frequency.
    - Proneness to sudden crises and macroeconomic volatility (Agenor, McDermott, and Prasad, 2000).
  - Findings from studies:
    - Business cycles in emerging markets tend to be shorter and faster from peak to trough and vice versa (Rand and Tarp, 2001).
  - Country-specific work noted:
    - Mall (1999) for India (leading indicators to forecast industrial production).
    - Simone (2000) for Argentina (coincident and leading indicators).
    - Dua and Banerji (1999, 2001) for India using NBER methodology.
  - For Jordan:
    - The Economic Cycle Research Institute has produced coincident and leading indicators for Jordan since 1998; results are not publicly available.

### III. Jordan’s Economic Performance (late 1980s–2002)
- Country context:
  - Low- to middle-income country with few natural resources.
  - 2002 per capita GDP averaged about $1,750.
  - Over 70 percent of value-added concentrated in services (trade, tourism, banking).
  - Manufacturing: phosphates, potash, petroleum refining, cement; light manufacturing (textiles, apparel) growing.
  - Tradable sector (exports + imports of goods and nonfactor services) = 106 percent of GDP in 2002.
  - Large remittances from expatriates in neighboring Arab countries.
- Financial crisis and adjustments:
  - Late 1980s severe financial crisis and Gulf war led to devaluation and debt rescheduling.
  - 1989: Jordanian dinar devalued by 21 percent in real effective terms (previously pegged to SDRs since 1973).
  - 1989: economy contracted by 13 percent in real terms.
  - 1990: JD lost another 10 percent of its value in real terms.
  - External debt exceeded 190 percent of GDP before rescheduling completed in 1993/94.
  - 1990s: macroeconomic adjustment under IMF arrangements; external debt reduced by more than half over a decade.
  - Since October 1995: JD effectively pegged to U.S. dollar.
  - Early 1990s temporary boost from repatriation of workers from Kuwait.
  - 1993–1999: economic growth averaged 4 percent.
  - Inflation reduced to single digits by 1991 and averaged 3 percent thereafter.
  - Structural reforms: privatization, trade liberalization, removal of generalized price subsidies replaced by direct income transfers, pension reform, strengthened financial supervision.
- Late 1990s pressures:
  - Late 1998–early 1999 speculative pressure related to King Hussein’s deteriorating health; reserves drained; authorities tightened policy and sought IMF arrangement; interest rates raised and fiscal policy tightened to avert currency crisis.
  - Subsequent restoration of confidence and accelerating economic growth.
- Selected economic indicators (Table 1, 1993–2002) — key entries preserved exactly:
  - Real GDP at market prices (annual percent changes): 4.5, 5.0, 6.2, 2.1, 3.3, 3.0, 3.1, 4.2, 4.2, 4.9
  - GDP deflator at market prices (annual percent changes): 2.8, 6.9, 1.9, 2.1, 1.2, 6.0, -0.2, -0.1, 0.0, 0.5
  - Nominal GDP at market prices (annual percent changes): 7.4, 12.2, 8.2, 4.2, 4.6, 9.2, 2.8, 4.1, 4.3, 5.3
  - Consumer price index (annual average): 3.3, 3.6, 2.3, 6.5, 3.0, 3.1, 0.6, 0.7, 1.8, 1.8
  - Consumer price index (end of period): 1.6, 5.1, 4.1, 2.6, 6.3, -0.8, 2.8, -1.9, 3.8, 0.5
  - Nominal GDP at market prices (in millions of JD): 3,925; 4,401; 4,774; 4,982; 5,138; 5,610; 5,767; 6,002; 6,260; 6,594
  - Gross domestic investment (percent of GDP): 36.6, 33.3, 33.0, 30.5, 25.7, 21.8, 21.6, 22.1, 22.4, 23.0
    - Government (percent of GDP): 6.3, 6.0, 7.0, 7.3, 5.6, 6.0, 5.8, 5.2, 5.9, 6.8
    - Other (percent of GDP): 30.4, 27.3, 25.9, 23.2, 20.1, 15.8, 15.9, 16.9, 16.5, 16.3
  - Gross national savings (percent of GDP): 30.4, 31.3, 33.5, 32.0, 30.4, 25.8, 29.8, 22.8, 22.4, 27.9
    - Government (percent of GDP): 5.7, 4.6, 3.1, 4.4, 3.1, 0.1, 2.2, 0.5, 2.2, 1.7
    - Other (percent of GDP): 24.7, 26.6, 30.4, 27.6, 27.3, 25.7, 27.7, 22.3, 20.2, 26.2
  - Savings/investment balance (percent of GDP): -11.5, -6.4, -3.8, -3.2, 0.4, 0.3, 5.0, 0.7, -0.1, 4.9
    - Government: -0.5, -1.4, -3.9, -2.8, -2.5, -6.0, -3.5, -4.7, -3.7, -5.0
    - Other: -10.9, -5.0, 0.1, -0.4, 2.9, 6.2, 8.5, 5.4, 3.6, 10.0
  - Revenue and grants (percent of GDP): 34.8, 32.6, 34.4, 33.6, 31.5, 30.4, 31.0, 30.0, 30.8, 30.5
    - Of which: grants (percent of GDP): 4.1, 4.0, 3.6, 4.5, 4.4, 3.7, 3.5, 4.2, 4.4, 5.2
  - Expenditure and net lending (incl. off-budget accounts) (percent of GDP): 36.4, 34.5, 36.0, 36.6, 34.5, 36.8, 34.9, 34.7, 34.5, 35.5
  - Overall fiscal balance (after grants) (percent of GDP): -0.5, -1.4, -3.9, -2.8, -2.5, -6.0, -3.5, -4.7, -3.7, -5.0
  - Government and government guaranteed debt (percent of GDP) 1/: ............, 94.3, 104.5, 111.3, 99.7, 97.8, 101.4
    - Of which: external: ............, 89.2, 89.3, 95.5, 84.0, 79.4, 81.2
  - Current account balance (after grants) (percent of GDP): -11.5, -6.4, -3.8, -3.2, 0.4, 0.3, 5.0, 0.7, -0.1, 4.9
  - Domestic merchandise exports (annual percent changes in U.S. dollar terms): 7.0, 13.8, 26.2, 2.3, 2.6, -1.9, 0.5, 2.8, 25.1, 13.7
  - Merchandise imports (annual percent changes in U.S. dollar terms): -6.6, -4.6, 9.5, 16.2, -4.4, -6.7, -3.3, 23.7, 5.6, 2.7
  - Net foreign assets (changes in percent of beginning of period broad money): -1.7, 2.8, 2.5, 0.4, 8.1, 1.9, 10.4, 12.6, 1.8, 5.4
  - Broad money (annual percent changes): 8.3, 8.1, 6.5, 0.3, 7.8, 7.6, 12.0, 10.2, 5.8, 7.0
  - Nominal per capita GDP (in U.S. dollars): 1,404.1; 1,506.7; 1,568.6; 1,559.0; 1,575.3; 1,663.8; 1,660.1; 1,679; 1,703; 1,746
  - In months of prospective imports of GNFS 4/1.: 1.9, 1.2, 1.1, 1.8, 4.6, 3.3, 4.8, 6.4, 6.1, 7.5
  - As percent of Jordanian dinar broad money: 10.6, 6.8, 6.6, 11.1, 25.0, 16.9, 25.6, 33.2, 30.1, 38.1
  - Net International Reserves (in millions of U.S. dollars) 2/: 612; 463; 534; 442; 1,508; 967; 1,463; 2,275; 2,111; 3,032
  - U.S. dollar per Jordanian dinar (period average): 1.44, 1.43, 1.43, 1.41, 1.41, 1.41, 1.41, 1.41, 1.41, 1.41
  - Notes:
    - 1/ Domestic debt is net of government deposits with the banking system, and external debt includes collateralized Brady bonds.
    - 2/ Net of short-term foreign liabilities and excluding commercial banks' foreign currency deposits with the Central Bank of Jordan.
    - 3/ Excludes pledged assets under the 1993 commercial debt rescheduling agreement and the yearly change in foreign currency swaps.
    - 4/ Imports of goods and nonfactor services, excluding imports for re-export, in subsequent 12 months.
- Figure 1: Economic Performance, 1980–2002 (visual showing Real GDP growth, Inflation, Exchange rate and periods labeled I. Boom of the 1980s; II. Exchange Rate Crisis; III. The Recovery). Source: IMF International Financial Statistics.
- Near-term note in text:
  - Despite negative shocks (renewed Palestinian-Israeli conflict and September 11, 2001 events), economic growth accelerated to 4.2 percent in 2001.

*Source: _wp03170 - References (IMF staff document).*

### 4.9 percent in 2002. This acceleration is mainly been driven by export growth, particularly to

### _wp03170 - 4.9 percent in 2002. This acceleration is mainly been driven by export growth, particularly to

### Export-driven acceleration and trade links
- Growth detail:
  - 4.9 percent in 2002.
- Factors driving export surge:
  - Accession to the WTO in 2000.
  - Duty- and quota-free access to U.S. markets under the 1994 Peace Agreement with Israel (Qualified Industrial Zones require a minimum value added from both Israeli and Jordanian sources).
  - The 2001 Free Trade Agreement with the United States.
- Major export destination highlighted:
  - The United States.

### Business cycle characteristics for Jordan
- Overall characterization:
  - Business cycle has been quite irregular and dominated by the exchange rate crisis of the late 1980s.
  - Remainder of the period characterized by somewhat uneven growth; regular booms and troughs typical of advanced economies cannot be discerned.
  - Periods of higher and slower growth suggest some pattern of business cycle activity, similar to other emerging markets dominated by sudden macroeconomic crises.
- Comparative evidence:
  - Consistent with Agenor, McDermott, and Prasad (2000) on emerging-market cycles.

### Data sample for composite indicators
- Sample composition:
  - Monthly observations of 40 variables from all sectors of the economy (real, fiscal, monetary, and external sectors).
  - Sample period: January 1996 to December 2002 (84 observations).
- Data limitations and adjustments:
  - Most variables prior to 1996 available only on annual or quarterly basis; unsuitable for monthly composite indicators.
  - Change in exchange rate regime in October 1995 affects stability of relationships and limits extending sample further back.
  - Exports and imports deflated by their respective price indexes; all other nominal variables deflated by the consumer price index.
  - The net usable reserves of the CBJ, denominated in U.S. dollars, were deflated by the U.S. CPI index.
- External influences included:
  - U.S. and Euro area producer price indexes and OECD composite leading indicators included as potential coincident and leading indicators for Jordan.

### Seasonal adjustment and treatment of Muslim holidays
- Seasonal adjustment procedure:
  - All variables seasonally adjusted to account for both Gregorian calendar and Muslim holidays.
  - U.S. Census Bureau X12 procedure used for Gregorian-calendar seasonal adjustment.
  - X12 procedure leaves residual seasonality associated with two major Muslim holidays: Eid-ul-Fitr and Eid-ul-Adha (their Gregorian dates shift year-to-year).
- Dummy-variable adjustment:
  - Following Alper and Aruoba (2001), two dummy variables created for Gregorian dates of the two major Muslim holidays; estimated equation for each X12-seasonally adjusted variable:
    - tAdhaulEidFitrulEidt UDDV+++= −−−−21 βαβ
  - Regression results (summary):
    - Eid-ul-Fitr dummy significant only for: industrial production, construction activity, reserve money, real imports, and the trade balance.
    - Eid-ul-Adha dummy insignificant for all variables.
  - Further seasonal adjustment applied to variables where the Eid-ul-Fitr coefficient was significant.
- Illustration:
  - Figure 2 presents seasonal adjustment of industrial production (X12 and X12 adjusted for Eid-ul-Fitr).

### Stationarity, transformations, and reference series choice
- Unit-root testing:
  - All seasonally adjusted variables tested in logarithmic form for nonstationarity using Phillips-Perron tests.
  - Unit root hypothesis could not be rejected for some variables in logarithms.
  - All variables in first differences (i.e., growth rates) were found to be stationary.
  - Regressions expressed in first differences to avoid spurious correlations associated with non-stationary variables.
- Choice of reference series for CEI:
  - Real GDP preferable in theory but available only quarterly and with substantial lag.
  - Industrial production index chosen as reference series for the state of the economy (common in NBER and OECD methodologies).
  - Validation for Jordan:
    - Positive correlation between real GDP and real index of production in the sample period: 88 percent trended, 45 percent detrended.
    - Visual similarity between series (see Figure 3).
  - Rationale for composite index:
    - Composite index preferable to a single series because it reflects a broader spectrum of the economy (real, monetary, fiscal, external) and reduces measurement error associated with individual cyclical indicators.

*Source: _wp03170 - 4.9 percent in 2002. This acceleration is mainly been driven by export growth, particularly to*

### Section II, both the NBER-CB and OECD methodologies follow an ad hoc procedure to

### _wp03170 - Section II, both the NBER-CB and OECD methodologies follow an ad hoc procedure to

### Methodological approach: selection and estimation
- The paper departs from ad hoc selection procedures by using an econometric approach based on statistical inference.
- Variable selection uses a principle of parsimony and recursive elimination of the variable with the lowest T-ratio (as in Stock and Watson (1989)), with attention to avoiding multicollinearity among close proxies.
- A reduced form linear regression is estimated rather than a dynamic factor model because the sample for Jordan is limited.
- The estimated reduced form for coincident indicators is:
  - ∆LRIIP_t = α + β' ∆LCI_t + u_t
  - u_t = ε_t + θ ε_{t-1} (MA(1) error term)
- As the error term is not normally distributed, standard errors and covariance matrices are estimated using the Newey-West heteroskedastic-consistent procedure.

### Estimation of the Coincident Economic Indicator (CEI)
- Selection procedure identified five coincident indicators from all sectors of the economy: TOT, TB, IMPCG, EMPTAX, COPERM.
- All five variables are significant at the 1 percent significance level.
- These variables explain about two thirds of the variation in the growth rate of industrial production.
- The fitted values closely track the actual data (see Figure 4 in source).

- Table 2 — Estimation of Coincident Economic Indicators (dependent variable: IIP)
  - Sample (adjusted): 1996:02 2002:12
  - Included observations: 83 after adjusting endpoints
  - Convergence achieved after 10 iterations
  - Newey-West HAC Standard Errors & Covariance (lag truncation=3)
  - Backcast: 1996:01
  - Variables and estimated coefficients:
    - C: Coefficient 0.004075  Std. Error 0.001332  t-Statistic 3.058719  Prob. 0.0031
    - TOT: Coefficient 0.506743  Std. Error 0.138262  t-Statistic 3.665089  Prob. 0.0005
    - TB: Coefficient 0.561191  Std. Error 0.173907  t-Statistic 3.226965  Prob. 0.0018
    - IMPCG: Coefficient 0.067893  Std. Error 0.024072  t-Statistic 2.820390  Prob. 0.0061
    - EMPTAX: Coefficient 0.053902  Std. Error 0.018522  t-Statistic 2.910185  Prob. 0.0047
    - COPERM: Coefficient 0.158926  Std. Error 0.024182  t-Statistic 6.572115  Prob. 0.0000
    - MA(1): Coefficient -0.739768  Std. Error 0.062502  t-Statistic -11.835910  Prob. 0.0000
  - Goodness-of-fit and diagnostics:
    - R-squared 0.667357
    - Adjusted R-squared 0.641095
    - Mean dependent var 0.003213
    - S.D. dependent var 0.065715
    - S.E. of regression 0.039369
    - Akaike info criterion -3.551120
    - Schwarz criterion -3.347122
    - Log likelihood 154.3715
    - Sum squared resid 0.117792
    - F-statistic 25.41216
    - Prob(F-statistic) 0.000000
    - Durbin-Watson stat 1.790883
    - Inverted MA Roots .74
- Note: while θ is close to unity in absolute value, the null hypothesis θ = 1 can be rejected at the 5 percent confidence interval.

### Economic interpretation and CEI derivation
- Economic interpretation:
  - For a relatively open small economy (Jordan), external factors (terms of trade and trade balance) are important determinants of current economic activity.
  - Imports of capital goods (IMPCG) are interpreted as a proxy for overall investment; higher IMPCG signals improvement.
  - Employee payroll deductions (EMPTAX) act as a fiscal-sector coincident indicator: higher economic activity generates higher salaries and payroll deductions.
  - Construction permits (COPERM) proxy construction investment, known to be highly procyclical.
  - Global business cycle indicators (U.S. and Euro area industrial production indexes) were not significant, suggesting Jordanian export demand is driven more by market access than global demand.
- CEI construction:
  - Fitted monthly growth rates from the regression are interpreted as growth rates of the composite index.
  - Initial observation of the CEI set equal to the equivalent observation of industrial production; subsequent observations derived by multiplying the previous observation by the fitted monthly growth rate.
- Performance:
  - CEI picks up most turning points in industrial production but sometimes overstates cyclical fluctuations, possibly due to the small sample.
  - In the latter sample, CEI significantly understates activity, probably reflecting uncertainty surrounding the impending war in Iraq.

### Estimation of the Leading Economic Indicator (LEI)
- Leading horizon chosen: six months (aligned broadly with NBER-CB and OECD methodologies) to provide policy-relevant lead time.
- Reduced form estimated for a six-month lead:
  - ∆LRIIP_t+6 = α + β' ∆LLI_t + u_t
  - u_t = ε_t + θ ε_{t-1} (MA(1) error term)
- Variable selection uses the same recursive elimination and multicollinearity avoidance as for CEI; MA(1) included to capture short-term endogenous dynamics.

### Estimation results for LEI
- Selection procedure identified variables: CLAIMSPS, SPREAD3M, RESERVUSD, DEXP, ASTOCEXCH.
- Statistical significance:
  - All variables significant at the 5 percent confidence interval except ASTOCEXCH, which is significant at the 15 percent level.
- These variables explain about 50 percent of the variation in the growth rate of industrial production.
- Table 3 — Estimation of Leading Economic Indicators (dependent variable: IIP(+6))
  - Sample (adjusted): 1996:02 2002:06
  - Included observations: 77 after adjusting endpoints
  - Convergence achieved after 19 iterations
  - Newey-West HAC Standard Errors & Covariance (lag truncation=3)
  - Backcast: 1996:01
  - Variables and estimated coefficients:
    - CLAIMSPS: Coefficient 0.967311  Std. Error 0.156717  t-Statistic 6.172354  Prob. 0.0000
    - SPREAD3M: Coefficient -0.001210  Std. Error 0.000348  t-Statistic -3.480252  Prob. 0.0009
    - RESERVUSD: Coefficient 0.113004  Std. Error 0.023433  t-Statistic 4.822525  Prob. 0.0000
    - DEXP: Coefficient 0.176229  Std. Error 0.055984  t-Statistic 3.147841  Prob. 0.0024
    - ASTOCEXCH: Coefficient 0.084349  Std. Error 0.056545  t-Statistic 1.491731  Prob. 0.1402
    - MA(1): Coefficient -0.980252  Std. Error 0.016004  t-Statistic -61.24882  Prob. 0.0000
  - Goodness-of-fit and diagnostics:
    - R-squared 0.489537
    - Adjusted R-squared 0.453589
    - Mean dependent var 0.003078
    - S.D. dependent var 0.067135
    - S.E. of regression 0.049626
    - Akaike info criterion -3.093884
    - Schwarz criterion -2.911250
    - Log likelihood 125.1145
    - Sum squared resid 0.174855
    - Durbin-Watson stat 1.933549
    - Inverted MA Roots .98
- Economic interpretation:
  - Monetary sector: increases in real credit to the private sector (CLAIMSPS) or reductions in the interest spread (SPREAD3M) forecast higher activity six months ahead.
  - External sector: accumulation in net usable reserves (RESERVUSD) and increases in demand for domestic exports (DEXP) have positive near-term impacts.
  - Equity market: rises in share prices (ASTOCEXCH) indicate increased confidence and positively affect activity six months ahead.
  - U.S. and Euro area leading indicators were not significant, again suggesting limited influence of the global business cycle on Jordan’s future activity.

### Derivation and forecasting of the LEI; out-of-sample forecast
- LEI construction:
  - Fitted values from the regression determine statistically significant weights (coefficients).
  - Initial LEI observation set equal to the equivalent industrial production observation; subsequent observations derived by multiplying the previous observation by the fitted monthly growth rate.
- Out-of-sample forecasting:
  - Regression estimates used to forecast growth rates of industrial production from January to June 2003; LEI extended using the recursive procedure.
  - Standard error bands around the forecast calculated to provide a confidence interval (shown in Figure 8 in source).
- Forecast interpretation and caveats:
  - Projections in Figure 8 suggest a slowdown in economic growth (seasonally adjusted) for Q1 2003.
  - Actual data available after the forecast: industrial production contracted by 14 percent and real GDP growth slowed to 2.8 percent in Q1 2003, compared to 4.7 percent in Q4 2002.
  - Two caveats:
    - The confidence interval is relatively wide and increases with forecast horizon, reflecting that about 50 percent of variation in activity is unexplained by the regression.
    - Some value judgment is still required to assess LEI forecasts; geopolitical shocks (e.g., war in Iraq) are difficult to capture in a purely quantitative framework.
  - Note: The contraction in industrial production in February 2003 was exacerbated by a snow storm, which halted production for about a week.

### Conclusions and operational guidance
- Composite indexes of coincident and leading economic indicators provide useful summary statistics for analyzing current and future economic direction.
- The econometric approach produces statistically significant and economically intuitive CEI and LEI for Jordan despite a relatively small sample.
- Limitations:
  - A larger sample is likely necessary before the indexes can be relied upon for more than qualitative assessment.
  - The proposed indexes should be considered experimental and used operationally only after a testing period that confirms out-of-sample performance.
  - Forecasts must account for additional qualitative information that is hard to quantify, including geopolitical uncertainties and macroeconomic policy changes.
- Operational recommendation:
  - Regular updating of these indexes could provide a useful tool for the Jordanian authorities for policy formulation.
  - The econometric roadmap here may be useful for analyzing coincident and leading indicators in other emerging markets.

*Source: _wp03170 - Section II, both the NBER-CB and OECD methodologies follow an ad hoc procedure to (IMF PDF content provided).*

### Appendix II. Classical Business Cycle versus Growth Cycles Approaches to Estimating

### Appendix II. Classical Business Cycle versus Growth Cycles Approaches to Estimating Coincident and Leading Indicators

### Classical versus growth-cycle approaches
- Classical approach: analysis of booms and busts in output and other indicators over a given time period independently of underlying trend (Burns and Mitchell (1946)).
- Growth-cycle approach: focuses on cyclical fluctuations around long run trends; identified by filtering series from its long run trend (Lucas (1987)).
- Trade-offs:
  - Advantages of growth-cycle approach: less sensitive to underlying trend growth rate; can reveal cycles in fast-growing economies that have few absolute declines (Stock and Watson, 1999).
  - Advantages of classical approach: directly analyzes turning points in levels of economic activity (NBER definition); avoids loss of information from detrending (Burns and Mitchell, 1946).
- Conclusion in the source: benefits seem to weigh in favor of the classical approach for the analysis of small open economies like Jordan.

### Definition of business cycles and decomposition
- Classical definition (NBER): turning points in the level of economic activity; local maxima and minima in Yt or yt = log(Yt).
- Alternative decomposition: Yt = Tt + Ct + It, where Tt = trend, Ct = cycle, It = error term; business cycles identified with Ct (deviations from trend).
- Argument against routine detrending:
  - Burns and Mitchell (1946) argue that detrending may involve loss of critical information.
  - Separation of trend and cyclical components may be inconsistent with models where shocks affect both long-run growth and fluctuations (Stock and Watson, 1999).
- Situations favoring detrending:
  - When deterministic trend dominates and yt rarely declines (example: post-World War II Germany and Japan; some Asian economies prior to 1978), studying deviations from trend can be more informative.

### What is a trend and what is a cyclical component?
- Tt, Ct, and It are unobservable; many trend-cycle decompositions exist.
- Common filters referenced: Hodrick and Prescott (1997), Baxter and King (1999); debates include Canova (1998a, 1998b, 1999) and Burnside (1998).
- Key issues:
  - Fundamental disagreements on trend properties and relation to cyclical component.
  - In small samples the appropriate statistical representation of trend is unresolved; choice of relationship between cyclical and secular components is arbitrary (Canova).
  - "Measurement without theory" concern: dynamic theory does not dictate the exact type of trend or trend-cycle relationship, creating circularity and potential specification errors.

### Problems with filtering (especially for developing countries)
- Different detrending methods extract different business-cycle information, leading to qualitative and quantitative differences.
- No independent criterion exists to choose filters (Canova 1998a and 1998b).
- Specific concerns about commonly used filters:
  - Band-Pass filter (BP): can produce spurious correlations between cyclical components; cointegration relationships can reverse inferred cyclical sign (Benati, 2001).
  - Hodrick-Prescott (HP) filter: can induce business-cycle periodicities and comovement in series that do not contain cycles (Harvey and Jaeger, 1993; Cogley and Nason, 1995).
- Reliance on filter parameters calibrated for industrial countries when analyzing developing countries is ad hoc and can produce misleading summary statistics or validate inappropriate numerical models (Rand and Tarp (2001)).
- Classical analysis of coincident and leading indicators typically uses unconditional correlations between cyclical components; such correlations do not imply causality (Agenor, McDermott, and Prasad, 2000).

### Appendix IV — Muslim holidays (dates) and significance testing
- Table of Gregorian dates for Eid-ul-Fitr and Eid-ul-Adha during the sample period:
  - Eid-ul-Fitr (End of Ramadan) / Eid-ul-Adha (Day of Sacrifice)
  - 1996: 20 February / 28 April
  - 1997: 9 February / 18 April
  - 1998: 30 January / 8 April
  - 1999: 19 January / 28 March
  - 2000: 8 January and 27 December / 17 March
  - 2001: 16 December / 6 March
  - 2002: 6 December / 23 February
- Dummy variables defined:
  - EIDULFITR = 1 if eid-ul-fitr falls in the month, 0 otherwise.
  - EIDULADHA = 1 if eid-ul-adha falls in the month, 0 otherwise.
  - EIDSDUMMY = 1 if eid-ul-fitr or eid-ul-adha falls in the month, 0 otherwise.
- Table 2 reports significance tests on X12 seasonally adjusted variables for residual seasonality associated with these dummies; significance markers are 10 percent (*), 5 percent (**), and 1 percent (***). Examples of reported results (variable — significance columns: EIDULFITR / EIDULADHA / EIDSDUMMY):
  - COAREA: **--*
  - COPERM: ***--***
  - IIP: **--**
  - IMP: **--*
  - RESMON: **--*
  - TB: **--*
  - TOTTAX: *----
  - Many variables show "------" indicating no significance for those dummies.

### Appendix V — Unit root tests (Phillips-Perron)
- Table 1 presents Phillips-Perron tests for unit roots for variables expressed in logarithms, both in levels and in first differences (growth rates).
- Significance notation: * = reject unit root at 10 percent; ** = reject at 5 percent; *** = reject at 1 percent.
- Tests performed with an autoregressive spectral-OLS procedure and a lag length of 9 months.
- Selected test results (Variable — Test on levels (in logarithms) — Test on First Differences (in growth rates)):
  - AQABA: -7.872***  -101.101***
  - ASTOCEXCH: -7.075***  -9.917***
  - BMON: -3.724***  -29.532***
  - BMONJD: -0.941  -44.111***
  - CBYCAP: -7.648***  -155.049***
  - CEXP: -13.311***  -56.630***
  - CLAIMSPS: -1.792  -10.503***
  - COAREA: -4.239***  -72.566***
  - COPERM: -6.504***  -128.229***
  - CPIJ: -1.927  -10.438***
  - CPIUSA: -1.317  -6.539***
  - CUREXP: -4.063***  -54.965***
  - DEXP: -2.162  -25.897***
  - ELEC: -0.865  -26.199***
  - EMPTAX: -4.202***  -24.501***
  - EUROCLI: -1.968  -3.105**
  - EUROIIP: -1.527  -12.534***
  - EUROPPI: -1.371  -3.812***
  - IIP: -1.385  -74.564***
  - IMP: -7.830***  -55.049***
  - IMPCG: -6.489***  -84.053***
  - IMPCMI: -4.998***  -13.524***
  - IR3MJ: -0.804  -8.508**
  - IR3MUSA: 2.221  -3.764**
  - MANUFAC: -33.561***  -156.707***
  - MINING: -5.755***  -83.967***
  - PIEXP: -1.769  -21.390***
  - PIIMP: -5.366***  -20.246***
  - RESERVUSD: -1.471  -6.596***
  - RESMON: -3.820***  -36.377***
  - REXP: -7.350***  -116.516***
  - SALTAX: -3.089***  -14.217***
  - SPREAD3M: -3.140**  -12.203***
  - TB: -8.753***  -71.011***
  - TOT: -2.114  -24.581***
  - TOTDEP: -3.546***  -42.740***
  - TOTTAX: -10.450***  -162.243***
  - TRADTAX: -2.733*  -14.652***
  - USCLI: -2.248  -3.609***
  - USIIP: -2.360  -6.095***
  - USPPI: -2.323  -7.964***
- General pattern: many variables do not reject unit root in levels but reject in first differences, consistent with I(1) behavior for numerous series in the dataset.

*Source: Appendix II–V from the provided IMF content unit.*

### References

### _wp03170 - References

### Methodological foundations and theory
- Burns, Arthur. F., and Wesley C. Mitchell, 1946, “Measuring Business Cycles,” NBER Working Paper No. 2, (Cambridge, Massachusetts: National Bureau of Economic Research).
- Koopmans, J.T., 1947, “Measuring Without Theory,” Review of Economics and Statistics, Vol. 29, pp. 161–172.
- Lucas, R. E., 1987, Models of Business Cycles, Blackwell, New York.
- Plosser, Charles I., and G. William Schwert, 1977, “Estimation of a Non-Invertible Moving Average Process: The Case of Overdifferencing,” Journal of Econometrics, No. 6, pp. 199–224.
- Hamilton, James D., 1994, “Time Series Analysis,” Princeton University Press, Princeton, New Jersey.
- Enders Walter, 1995, “Applied Econometric Time Series,” John Wiley & Sons, New York City, New York.

### Filters, detrending, and business cycle measurement
- Hodrick, Robert J., and Edward C. Prescott, 1997, “Postwar U.S. Business Cycles: An Empirical Investigation,” Journal of Monetary, Credit, and Banking, Vol. 29, pp.1–16.
- Baxter, Marianne, and Robert G. King, 1999, “Measuring Business Cycles: Approximate Band-Pass Filters for Economic Time Series,” The Review of Economic and Statistics, Vol. 81, No. 4, pp. 575–593.
- Canova, Fabio, 1998a, “Detrending and Business Cycles Facts,” Journal of Monetary Economics, Vol. 41, pp. 475–512.
- Canova, Fabio, 1998b, “Detrending and Business Cycles Facts: A User’s Guide,” Journal of Monetary Economics, Vol. 41, pp. 533–540.
- Canova, Fabio, 1999, “Does Detrending Matter for the Determination of the Reference Cycle and the Selection of Turning Points?.” Economic Journal, Vol. 109, pp. 126–150.
- Cogley, Timothy, and James M. Nason, “Effects of the Hodrick-Prescott Filter on Trend and Difference Stationary Time Series: Implications for Business Cycle Research,” Journal of Economic Dynamics and Control, Vol. 19, Nos. 1–2, pp. 253–78.
- Burnside, Craig, 1998, “Detrending and Business Cycles Facts: A Comment,” Journal of Monetary Economics, Vol. 41, pp. 513–32.
- Harvey, A. C., and A. Jaeger, “Detrending, Stylized Facts and the Business Cycle,” Journal of Applied Econometrics, Vol. 8, No. 3, pp. 231–247.
- Kozicki, Sharon, 1999, “Multivariate Detrending under Common Trend Restriction: Implications for Business Cycle Research,” Journal of Economic Dynamics and Control, Vol. 23, pp. 997–1028.
- Benati, Luca, 2001, “Band-Pass Filtering, Cointegration, and Business Cycle Analysis,” Working Paper (London: The Bank of England).
- Zornowitz, Victor, and Ataman Ozyildirim, 2002, “Time Series Decomposition and Measurement of Business Cycles, Trend and Growth Cycles” NBER Working Paper No. 8736 (Cambridge, Massachusetts: National Bureau of Economic Research).

### Leading, coincident, and composite indicators
- Stock, James H., and Mark W. Watson, 1989, “New Indexes of Coincident and Leading Economic Indicators,” NBER Macroeconomic Annual Report (Cambridge, Massachusetts: National Bureau of Economic Research), pp. 351–394.
- Stock, James H. and Mark W. Watson, 1992, “A Procedure for Predicting Recessions With Leading Indicators: Econometric Issues and Recent Experience,” NBER Working Papers Series, (Cambridge, Massachusetts: National Bureau of Economic Research), No. 4014.
- Stock, James H., and Mark W. Watson, 1999, “Business Cycle Fluctuations in U.S. Macroeconomic Time Series,” Handbook of Macroeconomics, Vol. 1A. Chapter 1. pp. 3–65.
- Stock, James H. and Mark W. Watson, 2001, “Forecasting Output and Inflation: the Role of Asset Prices,” NBER Working Paper 8180 (Cambridge, Massachusetts: National Bureau of Economic Research).
- Conference Board, Inc., 1997, Business Cycle Indicators. Available via the Internet: http://www.globalindicators.org/GeneralInfo/bci4.pdf.
- Benoît, Arnaud, 2002, “The OECD System of Leading Indicators: Recent Efforts to Meet Users’ Needs.” Available via the Internet: http://www.oecd.org/pdf/M00003000/M00003284.pdf.
- Organization for Economic Cooperation and Development, 1998, “Composite Leading Indicators: a Tool for Short-Term Analysis. Available via the Internet: http://www.oecd.org/pdf/M00003000/M00003226.pdf
- Phillips, Keith R., 1998, “The Composite Index of Leading Economic Indicators: A Comparison of Approaches,” Journal of Economic and Social Measurement, Vol. 25, No. 3–4, pp. 141–162
- McGuckin, Robert H., Ataman Ozyildirim, and Victor Zarnovitz, “The Composite Index of Leading Economic Indicators: How to Make it More Timely,” NBER Working Paper No. 8430 (Cambridge, Massachusetts: National Bureau of Economic Research).
- Nilsson Ronny, 2000, “Confidence Indicators and Composite Indicators,” OECD. Available via the Internet: http://www.oecd.org/pdf/M00018000/M00018836.pdf
- Mall, O. P., 1999, “Composite Index of Leading Indicators for Business Cycles in India,” Reserve Bank of India Occasional Papers, Vol. 20, No. 3, pp. 373–414.
- Dua, Pami and Anirvan Banerji, 1999, “An Index of Coincident Economic Indicators for the Indian Economy,” Journal of Quantitative Economics, Vol. 15, pp. 177–201.
- Dua, Pami and Anirvan Banerji, 2001, “A Leading Index for the Indian Economy.” Available via the Internet: http://www.cdedse.org/papers/work90.pdf.
- McGuckin, Robert H., Ataman Ozyildirim, and Victor Zarnovitz, “The Composite Index of Leading Economic Indicators: How to Make it More Timely,” NBER Working Paper No. 8430 (Cambridge, Massachusetts: National Bureau of Economic Research).
- Moore, Geoffrey H. and Julius Shiskin, 1967, “Indicators of Business Expansions and Contractions,” NBER Occasional Paper No. 103, (Cambridge, Massachusetts: National Bureau of Economic Research).

### Country studies, calendar effects, and applications
- Agenor, Pierre-Richard, C. John McDermott, and Eswar S. Prasad, 2000, “Macroeconomic Fluctuations in Developing Countries: Some Stylized Facts,” The World Bank Economic Review, Vol. 14, No. 2, pp. 251–85.
- Leigh, Daniel and Rossi, Marco, (2002), “Leading Indicators of Growth and Inflation in Turkey,” IMF Working Paper No. 02/231 (Washington: International Monetary Fund).
- Simone, Alejandro, 2001, “In Search of Coincident and Leading Indicators of Economic Activity in Argentina,” IMF Working Paper, WP/01/30.
- Rand, John., and Finn Tarp, 2001, “Business Cycles in Developing Countries: Are They Different?,” CREDIT Research Paper, No. 01/21.
- Kibritcioglu, Benji, Bulent Kose, and Gamze Ugur, “A Leading Indicators Approach to the Predictability of Currency Crises: the Case of Turkey.” Available via the Internet: http://www.econturk.org/Turkisheconomy/bengi.pdf.
- Riazuddin, Riaz and Mahmoud-ul-Hasan Khan, 2002, “Detection and Forecasting of Muslim Calendar Effects in Time Series Data,” State Bank of Pakistan Working Paper, Available via the Internet: http://www.sbp.org.pk/Wpaper/wp02.pdf.
- Alper, Emre C., and S. Boragan Aruoba, 2001, “Moving Holidays and Seasonality: An Application in the Time and the Frequency Domains For Turkey.” Available via the Internet: http://www.econ.boun.edu.tr/papers/wp-01-07.pdf.
- Maciejewski, Edouard, and Ahsan Mansur, eds., 1996, Jordan: Strategy for Adjustment and Growth, IMF Occasional Paper No. 136 (Washington: International Monetary Fund).
- Central Bank of Jordan: Monthly Statistical Bulletin, Various Issues.
- Ministry of Finance of Jordan, Government Finance Bulletin, various issues.

### Empirical tools, software, and data resources
- Lilien, D. and others, 2002, Eviews 4.1 Software Package, Quantitative Micro Software LLC, Irvine, CA, Available via the Internet: http://www.eviews.com/index.html.
- United Nation Economic and Social Commission for Asia and the Pacific (UNESCAP). Available via the Internet: http://www.unescap.org/stat/meet/libts/libts.htm
- Harding, D., and A. Pagan, 2001, “Extracting, Analyzing and Using Cyclical Information,” Discussion Paper Series (Hong Kong: University of Hong Kong. School of Economics and Finance), No. 338, pp. 1–34.
- Pallage, Stephane, and Michel A. Robe, 1998, “Foreign Aid and the Business Cycle,” IMF Seminar Series, No. 1998–24, pp. 1–22.

*References list from _wp03170 - References*

---


_Source: https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2003/_wp03170.pdf_