## 1. PCA FCIs Factor Loadings

## Source details

**Canonical URL:** [1. PCA FCIs Factor Loadings](https://www.imf.org/-/media/files/publications/cr/2019/1sycea2019003.pdf)

## Other formats

- [Markdown version](/-/media/files/publications/cr/2019/1sycea2019003.pdf.md)
- [Structured JSON version](/-/media/files/publications/cr/2019/1sycea2019003.pdf.json)

---

### Introduction
- Purpose: develop a Financial Conditions Index (FCI) for Seychelles to gauge the operational state of the financial sector and predict real economy activity.
- Sample period: 2003Q2‒2018Q3.
- Motivation and context:
  - FCIs extend monetary conditions indices by incorporating asset prices, long-term interest rates, liquidity indicators, and risk factors to better capture financial stress (notably after the Global Financial Crisis).
  - FCIs can be used to assess whether financial conditions are lax or tight and to forecast economic activity; pooling financial information improves predictive power.
  - Two principal construction approaches: weighted-sum (VAR) and common factor (PCA). VAR links to GDP but may be limited by degrees of freedom; PCA allows many variables but is agnostic about output relationships.

### Methodology
- Variable selection:
  - Global variables: MSCI EUR, VSTOXX.
  - Domestic variables: NEER, average lending rate, Treasury Bill rate (T-bill rate), growth rates of credit to the private sector, currency in circulation.
  - Selection criterion: variables with the highest and significant correlation coefficient with Seychelles’ real GDP growth; excluded variables were often highly correlated with selected ones.
- VAR (weighted-sum) approach:
  - Recursive VAR estimated with the seven financial variables plus annualized quarterly real GDP and CPI.
  - Identification: Cholesky decomposition with ordering: MSCI EUR, VSTOXX, GDP growth, consumer price index, NEER, lending rate, Treasury Bill rate, credit to the private sector, currency in circulation.
  - Stationarity treatment:
    - Augmented Dickey-Fuller tests: all variables in level except the lending rate and VSTOXX are non-stationary.
    - Treasury Bill rate is first-difference stationary.
    - Non-stationary variables (MSCI EUR, NEER, currency in circulation, credit to the private sector, GDP and CPI) enter VAR as year-on-year percentage change.
    - Lending rate and VSTOXX enter in levels; Treasury Bill rate in first difference.
  - Weighted-sum FCI formula:
    - FCI_t = sum_{j=1}^k w_j (x_{jt} − u_j)
    - Here k=7; w_j is the cumulative two-quarter impulse response of real GDP growth to a one-unit shock to f_jt; u_j is the mean of variable j over the sample.
- PCA (principal component) approach:
  - PCA models variance structure using optimal linear combinations of observed financial variables.
  - To avoid over-weighting similar indicators, only one variable per area was chosen (e.g., stock market developments, bond spreads).
  - Static and dynamic PCA implemented; dynamics introduced by adding lags of each financial variable to the observables matrix (X_t replaced by Z_t = [X_t − X_{t−p}] with p selected via AIC; 1 lag included).
  - Number of common factors: Bai and Ng (2002) criteria suggest 7 common factors which account for 100 percent of variation in the data.
  - The 7 common factors were combined into a single common factor using weighted averaging with weights equal to the factors’ respective coefficients of variation.
  - Purging feedback from past economic activity:
    - Combined common factors are stripped using: F_t = A(L) y_t + A(L) π_t + ε_t
    - Where A(L) is lag operator for current and lagged GDP growth (y_t) and inflation (π_t); ε_t are the PCA-based FCIs capturing exogenous financial developments.

### Financial Conditions Index for Seychelles — Constructed Indices
- Constructed FCIs:
  - VAR-based (weighted-sum) FCI.
  - PCA-based FCIs: static PCA and dynamic PCA; purged and unpurged variants.
- Interpretation: increasing FCI implies looser financial conditions; decreasing FCI implies tighter financial conditions.
- Purging rationale: remove feedback from current and past economic activity (GDP growth and inflation) so FCI captures exogenous financial developments.

### Overview of the Constructed FCIs — Factor Loadings and Dynamics
- Factor loadings (correlations between unpurged factor-based FCIs and financial variables):
  - Positive loadings (improve financial conditions if higher): MSCI EUR, NEER, growth in credit, currency in circulation.
  - Negative loadings (improve financial conditions if lower): VSTOXX, domestic lending rate, T-bill rate.
  - Note: all factor loadings are statistically significant except for the average lending rate.
- Financial cycles identified (2003Q2-2018Q3):
  - Seychelles experienced about four financial cycles during 2003Q2-2018Q3.
  - First cycle: started before the 2009 financial crisis and ended with the 2009 recession trough in 2009Q2.
  - Second cycle: ended around 2012Q2; highest peak of loose financial conditions around end-2010 (coinciding with private sector credit recovery and banking sector excess liquidity following rapid public debt reduction).
  - Third cycle: spanned 2012Q2-2015Q2 (PCA-based).
  - Fourth cycle: appears to be nearing its end by the end of the sample, possibly reflecting recent monetary policy tightening.
- Relationship with GDP:
  - Constructed FCIs track GDP growth well and lead GDP accelerations and decelerations.
  - Example: financial conditions tightened in early 2007, about three quarters before Seychelles’ largest post-2000 GDP contraction.

### Forecast Evaluation and Predictive Performance
- Model and estimation:
  - Diffusion index model (equation (4)): y_t = α(L) y_{t−1} + θ(L) F_t + ξ_t, where y is real GDP growth and F is the FCI of interest.
  - The optimal number of lags for each estimation is chosen based on the Bayesian Information Criterion (BIC). Optimal lags based on AIC are higher, and more unstable, than those based on BIC.
  - The model in equation (4) is estimated with and without the FCI, and the relevant root mean squared errors (RMSEs) are compared.
  - GDP growth and purged FCIs span 2007Q4-2018Q3.
  - Real GDP reported as "Real GDP (annualized year-on-year percent change)".
- RMSE comparison (Table 2 findings):
  - Root mean squared errors are better (smaller) when FCIs are included in the estimation.
  - Focus is on FCIs stripped of feedback from current and past economic activity: purged PCA-based FCIs and VAR FCI.
  - VAR-based FCI shows the highest improvement in RMSE when included, suggesting marginally better forecasting power than PCA-based FCIs; therefore the VAR-based FCI is used for the rest of the analysis.
  - Reported RMSEs and optimal lags:
    - Purged Static Common Factor FCI: RMSE - BIC = 0.75; RMSE - AIC = 0.75; Optimal Lags BIC = 2; AIC = 2
    - Purged Dynamic Common Factor FCI: RMSE - BIC = 0.74; RMSE - AIC = 0.98; Optimal Lags BIC = 2; AIC = 4
    - VAR FCI: RMSE - BIC = 0.43; RMSE - AIC = 0.43; Optimal Lags BIC = 2; AIC = 2
    - Simple AR process: RMSE - BIC = 0.79; RMSE - AIC = 0.79; Optimal Lags BIC = 2; AIC = 2
  - Reported RMSE improvements at forecast horizons (t..t+4):
    - Static FCI: t = 0.52*; t+1 = 0.68*; t+2 = 0.72*; t+3 = 0.68*; t+4 = 0.6*
    - Static FCI (purged): t = 0; t+1 = -0.03; t+2 = -0.13; t+3 = -0.24; t+4 = -0.28*
    - Dynamic FCI: t = 0.63*; t+1 = 0.73*; t+2 = 0.73*; t+3 = 0.67*; t+4 = 0.56*
    - Dynamic FCI (purged): t = 0; t+1 = -0.05; t+2 = -0.17; t+3 = -0.22; t+4 = -0.2
    - VAR FCI: t = 0.66; t+1 = 0.76*; t+2 = 0.71*; t+3 = 0.58*; t+4 = 0.42*
    - * = statistically significant at the 10 percent level
  - Source: IMF Staff estimates

### Use of the FCI for Macroprudential Policy (Countercyclical Capital Buffer)
- Regulatory context:
  - Seychelles is in the process of implementing adoption of Basel II; full transition to Basel II and adoption of Basel III capital definition expected to be completed by the end of 2019 as one of reform targets under Seychelles’ authorities program supported by the Fund’s Policy Coordination Instrument (PCI).
  - The CCB is an additional capital buffer built when credit to the private sector grows disproportionately; it serves to absorb mounting losses in an ensuing downturn.
  - The Basel Committee on Banking Supervision (BCBS, 2010) recommends using a specification of the credit-to-GDP gap for activation of the CCB.
  - Several country authorities deviate from the pure BCBS buffer guide and consider other indicators such as asset prices or financial sector conditions more broadly.
  - As Seychelles adopts Basel III capital definition, the FCI could serve as one of the other indicators for setting CCB.
- BCBS (2010) guidance noted:
  - Banks should start building the CCB when the credit-to-GDP gap surpasses 2 percentage points, up to a maximum of 10 percentage points, at which point the maximum size of the CCB of 2.5 percent of risk-weighted assets is normally reached.
  - Banks can reduce the buffer when allowed by the regulator, normally when the credit boom episode is over or when bank losses rise in a downturn.

### FCI as Predictor of Credit Boom-Bust Episodes and Bank Distress
- Assessment focus: whether the FCI predicts unsustainable credit boom conditions followed by downturns more accurately than the credit-to-GDP gap; downturn episodes manifested in rising NPLs and falling bank profits.
- Estimation results: the FCI is a leading indicator for credit accelerations.
- Credit cycles (year-on-year growth in nominal private sector credit) over 2007Q1‒2018Q3:
  - 2007Q1‒2009Q4
  - 2009Q4‒2013Q2
  - 2013Q2‒2017Q1
  - 2017Q1‒2018Q3
- Findings:
  - These four credit growth cycles roughly coincide with the four credit-to-GDP gap cycles whereas the financial cycles identified through the dynamics of the FCI begin and end about two to three quarters before the credit boom-bust cycles.
  - This indicates that in Seychelles the FCI is a better predictor of credit booms and busts than the credit-to-GDP gap.
  - FCI and loan quality / profitability:
    - The rise to the peak of the FCI started in early 2009, a few quarters before the NPL ratio began to rise.
    - The FCI kept rising for about two years, a few quarters ahead of continued deterioration of the loan portfolio.
    - The credit gap either turned positive contemporaneously with rising NPLs or turned negative as the loan portfolio deterioration intensified.
    - Similar pattern observed for FCI, credit gap, and return on equity (ROE).
    - Conclusion: FCI appears to be a better predictor of run-ups in NPLs and deteriorating profitability than the credit gap.
    - Caveat: It is challenging to give concrete guidance on setting the CCB based on only one recent episode of mild banking distress.
  - Note: Consistent information on NPLs is available only from 2008.

### Conclusions and Policy Implications
- The paper develops a financial condition index for Seychelles using different approaches.
- Two sets of FCIs derived from the standard VAR and PCA approaches display a high correlation and strongly agree in their trajectories, including only slight differences between constructed FCI and the one purged from influence of real variables and between static and dynamic FCI that incorporates variable lags.
- Domestic financial conditions dominate in the FCIs:
  - NEER has the highest factor loading and thus contributes the most to the FCIs, followed by currency in circulation (CIC) and private sector credit.
  - Lending and Treasury bill rates feature least prominently.
  - Among external variables, MSCI EUR has the highest importance, followed by VSTOXX.
- Predictive performance:
  - The FCI predicts changes in economic activity well, particularly:
    - 2008-09 and 2011-13 economic downturns were signaled by a marked decline in the FCIs one year before.
    - Subsequent recoveries were signaled when the FCI turned around ahead of real economic activity.
- Policy recommendation / practical implication:
  - The FCI can help inform macroprudential policy decisions:
    - Once Seychelles completes the transition to Basel II and adoption of Basel III capital definition, decisions on setting the countercyclical capital buffer could be informed by scrutinizing FCI developments.
    - The FCI and other relevant information could be used for macroprudential policy purposes (such as activating the countercyclical capital buffer).
  - The economic forecasting framework for Seychelles could be usefully augmented with the FCI, which conveniently aggregates key external and domestic variables shown to influence future domestic real and financial activity.

*Source: IMF staff estimates and analysis in 1sycea2019003 (Seychelles FCI chapter).*

### 1. PCA FCIs Factor Loadings ________________________________________________________ 11

### 1. PCA FCIs Factor Loadings

### Introduction
- Purpose: develop a Financial Conditions Index (FCI) for Seychelles to gauge the operational state of the financial sector and predict real economy activity.
- Sample period: 2003Q2‒2018Q3.
- Motivation and context:
  - FCIs extend monetary conditions indices by incorporating asset prices, long-term interest rates, liquidity indicators, and risk factors to better capture financial stress (notably after the Global Financial Crisis).
  - FCIs can be used to assess whether financial conditions are lax or tight and to forecast economic activity; pooling financial information improves predictive power.
  - Two principal construction approaches: weighted-sum (VAR) and common factor (PCA). VAR links to GDP but may be limited by degrees of freedom; PCA allows many variables but is agnostic about output relationships.

### Methodology
- Variable selection:
  - Global variables: MSCI EUR (Europe Morgan Stanley Capital International index), VSTOXX (Europe’s stock market volatility index).
  - Domestic variables: NEER (nominal effective exchange rate), average lending rate, Treasury Bill rate (T-bill rate), growth rates of credit to the private sector, currency in circulation.
  - Selection criterion: variables with the highest and significant correlation coefficient with Seychelles’ real GDP growth; excluded variables were often highly correlated with selected ones.
- VAR (weighted-sum) approach:
  - Recursive VAR estimated with the seven financial variables plus annualized quarterly real GDP and CPI.
  - Identification: Cholesky decomposition with ordering: MSCI EUR, VSTOXX, GDP growth, consumer price index, NEER, lending rate, Treasury Bill rate, credit to the private sector, currency in circulation.
  - Stationarity treatment:
    - Augmented Dickey-Fuller tests: all variables in level except the lending rate and VSTOXX are non-stationary.
    - Treasury Bill rate is first-difference stationary.
    - Non-stationary variables (MSCI EUR, NEER, currency in circulation, credit to the private sector, GDP and CPI) enter VAR as year-on-year percentage change.
    - Lending rate and VSTOXX enter in levels; Treasury Bill rate in first difference.
  - Weighted-sum FCI formula:
    - FCI_t = sum_{j=1}^k w_j (x_{jt} − u_j)
    - Here k=7; w_j is the cumulative two-quarter impulse response of real GDP growth to a one-unit shock to f_jt; u_j is the mean of variable j over the sample.
- PCA (principal component) approach:
  - PCA models variance structure using optimal linear combinations of observed financial variables.
  - To avoid over-weighting similar indicators, only one variable per area was chosen (e.g., stock market developments, bond spreads).
  - Static and dynamic PCA implemented; dynamics introduced by adding lags of each financial variable to the observables matrix (X_t replaced by Z_t = [X_t − X_{t−p}] with p selected via AIC; 1 lag included).
  - Number of common factors: Bai and Ng (2002) criteria suggest 7 common factors which account for 100 percent of variation in the data.
  - The 7 common factors were combined into a single common factor using weighted averaging with weights equal to the factors’ respective coefficients of variation.
  - Purging feedback from past economic activity:
    - Combined common factors are stripped using: F_t = A(L) y_t + A(L) π_t + ε_t
    - Where A(L) is lag operator for current and lagged GDP growth (y_t) and inflation (π_t); ε_t are the PCA-based FCIs capturing exogenous financial developments.

### Financial Conditions Index for Seychelles — Constructed Indices
- Constructed FCIs:
  - VAR-based (weighted-sum) FCI.
  - PCA-based FCIs: static PCA and dynamic PCA; purged and unpurged variants.
- Interpretation: increasing FCI implies looser financial conditions; decreasing FCI implies tighter financial conditions.
- Purging rationale: remove feedback from current and past economic activity (GDP growth and inflation) so FCI captures exogenous financial developments.

### Overview of the Constructed FCIs — Factor Loadings and Dynamics
- Factor loadings (correlations between unpurged factor-based FCIs and financial variables):
  - Positive loadings (improve financial conditions if higher): MSCI EUR, NEER, growth in credit, currency in circulation.
  - Negative loadings (improve financial conditions if lower): VSTOXX (Europe’s stock market volatility), domestic lending rate, T-bill rate.
  - Note: all factor loadings are statistically significant except for the average lending rate.
- Financial cycles identified (2003Q2-2018Q3):
  - Seychelles experienced about four financial cycles during 2003Q2-2018Q3.
  - First cycle: started before the 2009 financial crisis and ended with the 2009 recession trough in 2009Q2.
  - Second cycle: ended around 2012Q2; highest peak of loose financial conditions around end-2010 (coinciding with private sector credit recovery and banking sector excess liquidity following rapid public debt reduction).
  - Third cycle: spanned 2012Q2-2015Q2 (PCA-based).
  - Fourth cycle: appears to be nearing its end by the end of the sample, possibly reflecting recent monetary policy tightening.
- Relationship with GDP:
  - Constructed FCIs track GDP growth well and lead GDP accelerations and decelerations.
  - Example: financial conditions tightened in early 2007, about three quarters before Seychelles’ largest post-2000 GDP contraction.

### Forecast Evaluation and Predictive Performance
- General finding: most constructed FCIs show forecasting power confirmed by correlation with real GDP.
- Specifics:
  - The VAR-based FCI correlates the most with one- to four-quarter ahead growth rates, indicating potential near-term predictive power.
  - The unpurged dynamic PCA-based FCI has the highest contemporaneous correlation with GDP growth.
  - The purged dynamic PCA-based FCI is uncorrelated with future growth, reflecting that purging removes feedback from current and previous quarters’ real economic activity.

### Policy relevance and applications highlighted
- FCIs can be useful forecasting tools for Seychelles’ economy and potentially useful instruments for macroprudential policy.
- This study introduces the use of computed FCIs for macroprudential purposes in Seychelles, notably as an alternative to the prevailing credit-to-GDP gap in setting the countercyclical capital buffer (CCB) under Basel III adoption plans.

*Prepared by Salifou Issoufou; Approved by David Owen; May 30, 2019.*

### 23.      To evaluate the strength of the constructed FCIs in forecasting GDP, we use the

### 23.      To evaluate the strength of the constructed FCIs in forecasting GDP, we use the

### Model and estimation
- Diffusion index model (equation (4)): y_t = α(L) y_{t−1} + θ(L) F_t + ξ_t, where y is real GDP growth and F is the FCI of interest.
- The optimal number of lags for each estimation is chosen based on the Bayesian Information Criterion (BIC).  Optimal lags based on AIC are higher, and more unstable, than those based on BIC.
- The model in equation (4) is estimated with and without the FCI, and the relevant root mean squared errors (RMSEs) are compared.
- GDP growth and purged FCIs span 2007Q4-2018Q3.
- Real GDP reported as "Real GDP (annualized year-on-year percent change)".

### RMSE comparison (Table 2 findings)
- Root mean squared errors are better (smaller) when FCIs are included in the estimation.
- Focus is on FCIs stripped of feedback from current and past economic activity: purged PCA-based FCIs and VAR FCI.
- VAR-based FCI shows the highest improvement in RMSE when included, suggesting marginally better forecasting power than PCA-based FCIs; therefore the VAR-based FCI is used for the rest of the analysis.

- RMSE - BIC and RMSE - AIC, Optimal Lags (as reported):
  - Purged Static Common Factor FCI: RMSE - BIC = 0.75; RMSE - AIC = 0.75; Optimal Lags BIC = 2; AIC = 2
  - Purged Dynamic Common Factor FCI: RMSE - BIC = 0.74; RMSE - AIC = 0.98; Optimal Lags BIC = 2; AIC = 4
  - VAR FCI: RMSE - BIC = 0.43; RMSE - AIC = 0.43; Optimal Lags BIC = 2; AIC = 2
  - Simple AR process: RMSE - BIC = 0.79; RMSE - AIC = 0.79; Optimal Lags BIC = 2; AIC = 2

- Reported RMSE improvements at forecast horizons (from table listing variable vs t..t+4):
  - Static FCI: t = 0.52*; t+1 = 0.68*; t+2 = 0.72*; t+3 = 0.68*; t+4 = 0.6*
  - Static FCI (purged): t = 0; t+1 = -0.03; t+2 = -0.13; t+3 = -0.24; t+4 = -0.28*
  - Dynamic FCI: t = 0.63*; t+1 = 0.73*; t+2 = 0.73*; t+3 = 0.67*; t+4 = 0.56*
  - Dynamic FCI (purged): t = 0; t+1 = -0.05; t+2 = -0.17; t+3 = -0.22; t+4 = -0.2
  - VAR FCI: t = 0.66; t+1 = 0.76*; t+2 = 0.71*; t+3 = 0.58*; t+4 = 0.42*
- * = statistically significant at the 10 percent level
- Source: IMF Staff estimates

### Use of the FCI for macroprudential policy (Countercyclical Capital Buffer)
- Seychelles is in the process of implementing adoption of Basel II; full transition to Basel II and adoption of Basel III capital definition expected to be completed by the end of 2019 as one of reform targets under Seychelles’ authorities program supported by the Fund’s Policy Coordination Instrument (PCI).
- The CCB is an additional capital buffer built when credit to the private sector grows disproportionately; it serves to absorb mounting losses in an ensuing downturn.
- The Basel Committee on Banking Supervision (BCBS, 2010) recommends using a specification of the credit-to-GDP gap for activation of the CCB.
- Several country authorities deviate from the pure BCBS buffer guide and consider other indicators such as asset prices or financial sector conditions more broadly.
- As Seychelles adopts Basel III capital definition, the FCI could serve as one of the other indicators for setting CCB.

- BCBS (2010) guidance noted in the text:
  - Banks should start building the CCB when the credit-to-GDP gap surpasses 2 percentage points, up to a maximum of 10 percentage points, at which point the maximum size of the CCB of 2.5 percent of risk-weighted assets is normally reached.
  - Banks can reduce the buffer when allowed by the regulator, normally when the credit boom episode is over or when bank losses rise in a downturn.

### FCI as predictor of credit boom-bust episodes and bank distress
- Assessment focus: whether the FCI predicts unsustainable credit boom conditions followed by downturns more accurately than the credit-to-GDP gap; downturn episodes manifested in rising NPLs and falling bank profits.
- Estimation results: the FCI is a leading indicator for credit accelerations.
- Based on year-on-year growth in nominal private sector credit, Seychelles experienced four credit cycles over 2007Q1‒2018Q3:
  - 2007Q1‒2009Q4
  - 2009Q4‒2013Q2
  - 2013Q2‒2017Q1
  - 2017Q1‒2018Q3
- These four credit growth cycles roughly coincide with the four credit-to-GDP gap cycles whereas the financial cycles identified through the dynamics of the FCI begin and end about two to three quarters before the credit boom-bust cycles.
- This indicates that in Seychelles the FCI is a better predictor of credit booms and busts than the credit-to-GDP gap.

- FCI and loan quality / profitability:
  - The rise to the peak of the FCI started in early 2009, a few quarters before the NPL ratio began to rise.
  - The FCI kept rising for about two years, a few quarters ahead of continued deterioration of the loan portfolio.
  - The credit gap either turned positive contemporaneously with rising NPLs or turned negative as the loan portfolio deterioration intensified.
  - Similar pattern observed for FCI, credit gap, and return on equity (ROE).
  - Conclusion: FCI appears to be a better predictor of run-ups in NPLs and deteriorating profitability than the credit gap.
  - Caveat: It is challenging to give concrete guidance on setting the CCB based on only one recent episode of mild banking distress.
- Note: Consistent information on NPLs is available only from 2008.

### Conclusions (summary of substantive findings and policy implications)
- The paper develops a financial condition index for Seychelles using different approaches.
- Two sets of FCIs derived from the standard VAR and PCA approaches display a high correlation and strongly agree in their trajectories, including only slight differences between constructed FCI and the one purged from influence of real variables and between static and dynamic FCI that incorporates variable lags.
- Domestic financial conditions dominate in the FCIs:
  - NEER has the highest factor loading and thus contributes the most to the FCIs, followed by currency in circulation (CIC) and private sector credit.
  - Lending and Treasury bill rates feature least prominently.
  - Among external variables, MSCI EUR has the highest importance, followed by Europe’s stock market volatility index (VSTOXX).
- The FCI predicts changes in economic activity well, particularly:
  - 2008-09 and 2011-13 economic downturns were signaled by a marked decline in the FCIs one year before.
  - Subsequent recoveries were signaled when the FCI turned around ahead of real economic activity.
- The FCI can help inform macroprudential policy decisions:
  - Once Seychelles completes the transition to Basel II and adoption of Basel III capital definition, decisions on setting the countercyclical capital buffer could be informed by scrutinizing FCI developments.
  - Historically Seychelles has experienced one financial crisis since the early 2000s, and the constructed FCIs predict well the boom-bust episodes.
  - The FCI and other relevant information could be used for macroprudential policy purposes (such as activating the countercyclical capital buffer).
- Policy recommendation / practical implication:
  - The economic forecasting framework for Seychelles could be usefully augmented with the FCI, which conveniently aggregates key external and domestic variables shown to influence future domestic real and financial activity.

*Source: IMF staff estimates and analysis in 1sycea2019003 (Seychelles FCI chapter).*

---


_Source: https://www.imf.org/-/media/files/publications/cr/2019/1sycea2019003.pdf_