## _wp1520

## Source details

**Canonical URL:** [_wp1520](https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2015/_wp1520.pdf)

## Other formats

- [Markdown version](/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2015/_wp1520.pdf.md)
- [Structured JSON version](/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2015/_wp1520.pdf.json)

---

### Introduction
- Developed a new monthly World Trade Leading Indicator (WTLI) using nonparametric and parametric approaches.
- Compared to the CPB World Trade Monitor’s benchmark indicator for global trade, the WTLI captures turning points in global trade with an average lead between 2 and 3 months.
- The WTLI tracks the annual growth rate in global trade, suggesting that the recent slowdown is due in part to certain cyclical factors.
- Purpose: provide policy makers foresight into the future direction of economic activity by tracking world trade more efficiently.

### Objectives and Data
- Target series: annual growth rate from the monthly CPB index defined as 100*(log(CPB_t) − log(CPB_t−12)). Denoted CPB annual growth rate and used to identify turning points.
- Turning points for the CPB annual growth rate established with the Bry-Boschan algorithm (k = 5 for monthly data), with rules:
  - disregard turning points within six months of sample ends;
  - require phases to last at least 6 months;
  - complete cycle from peak to peak at least 15 months;
  - in double trough/peak keep the lowest/largest value.
- Seven potential leading indicators chosen based on timeliness and dynamic correlation with the CPB:
  - Baltic Dry Index (BDI) — daily, End of month m.
  - Oil price (Brent) — daily, End of month m.
  - Commodity Research Bureau index (CRB) — monthly, Around the 3rd day of month m+1.
  - Purchasing Managers' Index (PMI) — monthly, Around the 3rd day of month m+1.
  - Ifo Business Climate (Ifo Climate) — monthly, Around the 21st day of month m.
  - Ifo Expectations (Ifo Expectation) — monthly, Around the 21st day of month m.
  - U.S. dollar nominal effective exchange rate (USD) — daily, End of month m.
- CPB index coverage note: built from trade series of 85 countries, covering around 97 percent of the world trade volume.
- Operational advantage: as of the beginning of month m+1, WTLI for month m can be computed — automatic gain in time about 2 months compared to CPB release.
- Analysis period due to data availability: July 1998–July 2014.

### Methodology
- Two approaches implemented to build composite WTLI:
  - Nonparametric indicator (WTLI_np):
    - Based on Conference Board methodology: weighted average of components, with weight of each component inversely related to its standard deviation computed on differentiated (log-differenced) data y_it = Δ log(X_it).
    - Weight formula: α_i = s_i^−1 / Σ_j s_j^−1, where s_i is the standard error of y_it.
    - Composite index CI_t = exp(Σ_i α_i log(X_it)).
    - WTLI_np defined as 100*(CI_t − CI_t−12).
    - WTLI_np tracks CPB annual growth and appears to lead turning points.
  - Parametric, factor-based indicator (WTLI_f):
    - Static Principal Component Analysis (PCA) on normalized differentiated series y_it to estimate factor(s) following Stock and Watson (2002) approach.
    - First estimated factor F̂_t (denoted â_t) intended to reflect global-trade growth rate over one month; integrated with a basis of 100 in July 1998 to form index I_t where I_t = 100 * Σ_{j=0}^{11} â_{t−j}? (paper indicates integration to compare 12-month differences).
    - WTLI_f defined as 100*(I_t − I_t−12).
    - WTLI_f also tracks CPB annual growth and tends to lead peaks and troughs.

### Turning Point Analysis and Key Statistics
- Bry-Boschan algorithm applied to WTLI_np and WTLI_f to detect turning points and compare with CPB annual growth turning points.
- WTLI_np performance:
  - Average lead relative to CPB: 2.8 months (text statement).
  - Table-based Mean lead reported under NP: 2.4.
  - Maximum lead: six months (noted in text).
  - Lead is quite stable over time.
- WTLI_f performance:
  - Average lead relative to CPB: 2.7 months.
  - Maximum lead: five months.
- Both WTLI versions signal a trough in June 2014, indicating a trough in the trade cycle likely to occur in 2014.
- Selected turning point comparisons (CPB date / WTLI_np date / WTLI_np lead; WTLI_f date / WTLI_f lead):
  - Peak: May 2000 / Mar. 2000 / 2 ; Mar. 2000 / 2
  - Trough: Dec. 2001 / Oct. 2001 / 2 ; Oct. 2001 / 2
  - Peak: Dec. 2002 / Oct. 2002 / 2 ; Oct. 2002 / 2
  - Trough: Aug. 2003 / May. 2003 / 5 ; Apr. 2003 / 4
  - Peak: June 2004 / May 2004 / 1 ; Jan. 2004 / 5
  - Trough: July 2005 / May 2005 / 2 ; May 2005 / 2
  - Peak: Mar. 2006 / Apr. 2006 / -1 ; May 2006 / -2
  - Trough: Apr. 2009 / Dec. 2008 / 4 ; Dec. 2008 / 4
  - Peak: May 2010 / Dec. 2009 / 5 ; Dec. 2009 / 5
- Summary statistics from Table 2:
  - Mean lead: NP column 2.4 ; Factor column 2.7
  - Standard deviation: NP column 1.9 ; Factor column 2.2

### Conclusion
- A monthly leading indicator for global trade (WTLI) was developed using both a simple nonparametric approach and a factor model.
- Both WTLI_np and WTLI_f track trade cycles as measured by the CPB and lead turning points by on average between 2 and 3 months, offering more timely signals than the CPB monthly index.

### Section 2 — Key findings on indicator performance
- The two approaches lead to very similar results.
- The indicator closely tracks the trade cycles as measured by the CPB.
- The WTLI leads turning points in the trade cycle with an average lead of two to three months.
- Overall, the cyclical indicator is very effective in tracking the annual growth of trade.
- The effectiveness is particularly true for the last period, which suggests that the recent slowdown is at least partly related to cyclical factors.

*Prepared by Karim Barhoumi and Laurent Ferrara, WP/15/20.*

### Section 1

### A World Trade Leading Index (WTLI)

### Introduction
- Developed a new monthly World Trade Leading Indicator (WTLI) using nonparametric and parametric approaches.
- Compared to the CPB World Trade Monitor’s benchmark indicator for global trade, the WTLI captures turning points in global trade with an average lead between 2 and 3 months.
- The WTLI tracks the annual growth rate in global trade, suggesting that the recent slowdown is due in part to certain cyclical factors.
- Purpose: provide policy makers foresight into the future direction of economic activity by tracking world trade more efficiently.

### Objectives and Data
- Target series: annual growth rate from the monthly CPB index defined as 100*(log(CPB_t) − log(CPB_t−12)). Denoted CPB annual growth rate and used to identify turning points.
- Turning points for the CPB annual growth rate established with the Bry-Boschan algorithm (k = 5 for monthly data), with rules: disregard turning points within six months of sample ends; require phases to last at least 6 months; complete cycle from peak to peak at least 15 months; in double trough/peak keep the lowest/largest value.
- Seven potential leading indicators chosen based on timeliness and dynamic correlation with the CPB:
  - Baltic Dry Index (BDI) — daily, End of month m.
  - Oil price (Brent) — daily, End of month m.
  - Commodity Research Bureau index (CRB) — monthly, Around the 3rd day of month m+1.
  - Purchasing Managers' Index (PMI) — monthly, Around the 3rd day of month m+1.
  - Ifo Business Climate (Ifo Climate) — monthly, Around the 21st day of month m.
  - Ifo Expectations (Ifo Expectation) — monthly, Around the 21st day of month m.
  - U.S. dollar nominal effective exchange rate (USD) — daily, End of month m.
- CPB index coverage note: built from trade series of 85 countries, covering around 97 percent of the world trade volume.
- Operational advantage: as of the beginning of month m+1, WTLI for month m can be computed — automatic gain in time about 2 months compared to CPB release.
- Analysis period due to data availability: July 1998–July 2014.

### Methodology
- Two approaches implemented to build composite WTLI:
  - Nonparametric indicator (WTLI_np):
    - Based on Conference Board methodology: weighted average of components, with weight of each component inversely related to its standard deviation computed on differentiated (log-differenced) data y_it = Δ log(X_it).
    - Weight formula: α_i = s_i^−1 / Σ_j s_j^−1, where s_i is the standard error of y_it.
    - Composite index CI_t = exp(Σ_i α_i log(X_it)).
    - WTLI_np defined as 100*(CI_t − CI_t−12).
    - WTLI_np tracks CPB annual growth and appears to lead turning points.
  - Parametric, factor-based indicator (WTLI_f):
    - Static Principal Component Analysis (PCA) on normalized differentiated series y_it to estimate factor(s) following Stock and Watson (2002) approach.
    - First estimated factor F̂_t (denoted â_t) intended to reflect global-trade growth rate over one month; integrated with a basis of 100 in July 1998 to form index I_t where I_t = 100 * Σ_{j=0}^{11} â_{t−j}? (paper indicates integration to compare 12-month differences).
    - WTLI_f defined as 100*(I_t − I_t−12).
    - WTLI_f also tracks CPB annual growth and tends to lead peaks and troughs.

### Turning Point Analysis and Key Statistics
- Bry-Boschan algorithm applied to WTLI_np and WTLI_f to detect turning points and compare with CPB annual growth turning points.
- WTLI_np performance:
  - Average lead relative to CPB: 2.8 months (Table 2 reporting mean 2.4 for NP column in table header; Table 2 shows Mean 2.4 under NP and 2.7 under Factor; text states WTLI_np average lead of 2.8 months — source contains both figures).
  - Maximum lead: six months (noted in text).
  - Lead is quite stable over time.
- WTLI_f performance:
  - Average lead relative to CPB: 2.7 months.
  - Maximum lead: five months.
- Both WTLI versions signal a trough in June 2014, indicating a trough in the trade cycle likely to occur in 2014.
- Table 2 (dates of turning points and leads in months) — selected entries (CPB date / WTLI_np date / WTLI_np lead; WTLI_f date / WTLI_f lead):
  - Peak: May 2000 / Mar. 2000 / 2 ; Mar. 2000 / 2
  - Trough: Dec. 2001 / Oct. 2001 / 2 ; Oct. 2001 / 2
  - Peak: Dec. 2002 / Oct. 2002 / 2 ; Oct. 2002 / 2
  - Trough: Aug. 2003 / May. 2003 / 5 ; Apr. 2003 / 4
  - Peak: June 2004 / May 2004 / 1 ; Jan. 2004 / 5
  - Trough: July 2005 / May 2005 / 2 ; May 2005 / 2
  - Peak: Mar. 2006 / Apr. 2006 / -1 ; May 2006 / -2
  - Trough: Apr. 2009 / Dec. 2008 / 4 ; Dec. 2008 / 4
  - Peak: May 2010 / Dec. 2009 / 5 ; Dec. 2009 / 5
- Summary statistics from Table 2:
  - Mean lead: NP column 2.4 ; Factor column 2.7
  - Standard deviation: NP column 1.9 ; Factor column 2.2

### Conclusion
- A monthly leading indicator for global trade (WTLI) was developed using both a simple nonparametric approach and a factor model.
- Both WTLI_np and WTLI_f track trade cycles as measured by the CPB and lead turning points by on average between 2 and 3 months, offering more timely signals than the CPB monthly index.

*Prepared by Karim Barhoumi and Laurent Ferrara, WP/15/20.*

### Section 2

### _wp1520 - Section 2

### Key findings on indicator performance
- The two approaches lead to very similar results.
- The indicator closely tracks the trade cycles as measured by the CPB.
- The WTLI leads turning points in the trade cycle with an average lead of two to three months.
- Overall, the cyclical indicator is very effective in tracking the annual growth of trade.
- The effectiveness is particularly true for the last period, which suggests that the recent slowdown is at least partly related to cyclical factors.

*Source: _wp1520 - Section 2*

---


_Source: https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2015/_wp1520.pdf_