## 2. Bank of Mauritius Foreign Exchange Market Intervention, 1996/97–2003/04

## Source details

**Canonical URL:** [2. Bank of Mauritius Foreign Exchange Market Intervention, 1996/97–2003/04](https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2005/_wp05172.pdf)

## Other formats

- [Markdown version](/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2005/_wp05172.pdf.md)
- [Structured JSON version](/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2005/_wp05172.pdf.json)

---

### I. Introduction — context and headline outcomes
- Mauritius’s monetary policy regime since 1996 characterized as “inflation targeting lite,” combining a public annual CPI inflation objective with continued managed exchange rate intervention.
- Inflation (year-on-year) fell from around 8 percent in mid-1996 to around 4 percent in the year to March 2004.
- Output growth averaged more than 5 percent per year over the same period.
- Shift toward inflation targeting lite associated with a fall in inflation principally through the reduction of inflationary expectations.
- Institutional and macro-financial context:
  - Government debt estimated at almost 60 percent of GDP in 2003/04.
  - Moderate (and declining) external debt.
  - Since 2002 the Bank of Mauritius (BOM) has been a net debtor of the government (the government has maintained substantial (net) deposits with the BOM).
  - Developed on-shore and growing off-shore banking sector and a primary dealer network for treasury bills.
  - Central statistical office produces good and timely macroeconomic data.
- Practical implication: Mauritius’s experience offers an example for small open island economies of controlling inflation while retaining exchange rate management flexibility.

### II. Development of the inflation targeting lite framework
- Historical evolution and institutional change:
  - BOM established by Bank of Mauritius Act in 1967; expanded supervisory powers in 1971.
  - Prior to the 1990s monetary policy relied on direct instruments (credit ceilings, administered rates).
  - Financial liberalization in late 1980s and early 1990s: liberalization of exchange controls, inter-bank FX market established, basket-peg replaced by more flexible exchange regime (completed in 1994).
  - Credit ceilings abolished in July 1993.
  - In 1996 BOM introduced an “informal inflation targeting regime” (inflation targeting lite).
- Characteristics of inflation targeting lite (aligned with Stone (2003) definition):
  - Multiple monetary policy objectives with increasing commitment to price stability.
  - Public announcement of an annual aggregate (CPI) inflation target since 1996/97.
  - Flexible monetary policy including exchange rate management and FX intervention.
  - Integrated operational strategy using many variables (including FX intervention) and reserve money program.
  - Relatively transparent monetary policy operations (annual reports, monthly statistical bulletin, public communications).
- Limitations relative to full-fledged inflation targeting:
  - Central bank mandate and institutional framework do not fully subordinate other objectives to price stability.
  - Only a forward-looking 12-month target is announced; lack of a medium-term trajectory and detailed analytical inflation-forecasting framework at the time.
  - BOM planned to further refine inflation-forecasting methodology and analysis of transmission mechanisms.

### III. Public CPI inflation targets and outcomes (1996/97–2003/04)
- BOM adopted a gradualist strategy to lower announced CPI targets, aiding credibility by setting realistic sequential targets and accommodating known one-off effects (example: VAT introduction on September 7, 1998).
- BOM met preannounced inflation objectives in all years except 2001/02 (missed due to unanticipated VAT increases and a cyclone).
- Table: CPI Inflation Targets and Actual Outcomes (as presented)
  - 1996/97 — Target inflation: 8.0; Actual inflation: 7.9
  - 1997/98 — Target inflation: 6.0; Actual inflation: 5.4
  - 1998/99 — Target inflation: 8.0; Actual inflation: 7.9
  - 1999/00 — Target inflation: 6.0; Actual inflation: 5.3
  - 2000/01 — Target inflation: 5.0-5.5; Actual inflation: 4.4
  - 2001/02 — Target inflation: 6.0; Actual inflation: 6.3
  - 2002/03 — Target inflation: 6.0; Actual inflation: 5.1
  - 2003/04 — Target inflation: 4.0; Actual inflation: 4.0

### IV. Monetary operations and exchange rate policy
- Integrated operational strategy:
  - Reserve money program announced in 1996 to maintain the monetary base on a path consistent with the inflation target and growth forecasts (reserve money targets were never announced due to difficulty achieving them).
  - BOM instruments: primary government bond auctions, repo/reverse repo, occasional secondary market operations.
  - Treasury (and BOM) bills issued at weekly auctions in 3-, 6-, 12-, and 24-month maturities; Development Loan Stock up to 15 years issued annually; five-year bonds issued quarterly; BOM began issuing its own bills from July 2003.
- Signaling:
  - Lombard facility used to signal stance of monetary policy; Lombard rate historically influenced market rates though the facility was less used when money market rates were much lower than the Lombard.
- Exchange rate management and FX intervention:
  - BOM maintains FX interventions to target real effective exchange level and to minimize excessive volatility in a very thin FX market.
  - Interventions are integrated with open market operations through the reserve money program and have generally declined over the period, implying increased policy flexibility and a shift toward inflation targeting.
- Nominal and real effective exchange rate behavior:
  - Over the last eight years prior to the study the nominal effective exchange rate depreciated steadily while the real effective exchange rate remained broadly stable.

### V. Bank of Mauritius foreign exchange market interventions (1996/97–2003/04)
- Reported interventions (In millions of U.S. dollars) — Table 2 figures (as presented)
  - Sales:
    - 1996/97: 392.1
    - 1997/98: 252.4
    - 1998/99: 126.6
    - 1999/00: 94.0
    - 2000/01: 118.3
    - 2001/02: 19.0
    - 2002/03: 0
    - 2003/04: 0
  - Purchases:
    - 1996/97: 38.9
    - 1997/98: 37.0
    - 1998/99: 11.0
    - 1999/00: 15.7
    - 2000/01: 0
    - 2001/02: 21.1
    - 2002/03: 189.0
    - 2003/04: 68.0
- Interpretation:
  - General decline in FX intervention over the period, with evidence of one-sided intervention early on and more symmetric, infrequent interventions later.
  - Given a thin FX market, some intervention may remain necessary to smooth excessive volatility.

### VI. Transparency and further institutional steps
- Transparency measures in place:
  - Frequent consultations between BOM and commercial banks; informative BOM website; comprehensive annual report; monthly statistical monetary bulletin; public speeches by the Governor.
  - Introduction of Lombard facility, repo and reverse repo frameworks intended to increase transparency.
- Steps planned/needed to formalize the regime:
  - Further refinement of inflation-forecasting methodology.
  - Further analysis of the transmission mechanism between monetary policy instruments and inflation.
  - Additional transparency consistent with formal inflation-targeting frameworks.

### VII. Open-economy macrofinance model used for empirical analysis
- Modeling approach:
  - New open-economy macrofinance model combines an affine no-arbitrage term structure (yield curve) with a standard open-economy new-Keynesian macroeconomic model.
  - Explicit modeling of the yield curve allows extraction of information on inflation expectations and the link between policy interest rates and longer-term yields.
- Key structure and interpretation:
  - Short-term policy rate captured by two latent term-structure factors tL and tS in equation (1): it = δ0 + tL + tS.
    - tL interpreted as the level of the yield curve and proxy for inflationary expectations.
    - tS interpreted as the slope of the yield curve (spread between long- and short-term rates); movements in tS reflect monetary policy choices and real interest rate fluctuations.
  - Aggregate demand equation (equation (2)): output gap responds to expected output, real exchange rate q, and real interest rate (tL − i).
  - Expectations updating (equation (3)): tL evolves with weight on past expectations and actual inflation, with stochastic influences.
  - Phillips curve (equation (4)): current inflation reflects forward-looking expectations (tL), backward-looking lags of inflation, the output gap, and changes in competitiveness.
  - Real exchange rate dynamics and expectations (equations (5) and (6)/(6’)): current real exchange rate determined by expected future real exchange rate and ex ante real interest differential; inertia in q formation included.
  - Monetary policy rule for BOM (equation (7)/(9)): policy interest rate rule depends on the output gap, inflation expectations tL, the deviation (πt − tL), the real exchange rate q, and smoothing parameters; reformulated in terms of tS in equation (9).
- Empirical estimation:
  - Model estimated by maximum likelihood over July 1996 to March 2004.
  - Data: monthly government bond yields for 3-, 6-, and 12-month maturities (provided by BOM); annualized CPI inflation from the IMF’s International Financial Statistics; real exchange rate from the IMF’s Information Notice System.
  - Due to lack of high-frequency GDP data, detrended, seasonally adjusted, real monthly credit to the private sector (deflated by the CPI) used as a proxy for the output gap.

### Appendix II — parameter estimates and statistical findings
- Dynamics of latent factors and inflation expectations:
  - Lρ 0.9921 [3.3exp(-57)] (tL extremely persistent).
  - Sρ 0.9890 [5.1exp(-25)] (tS highly persistent).
  - Unanticipated monetary shocks, Stu, uncorrelated (uρ 1.17exp(-5) [1.0000]).
  - Average expected inflation and its fluctuation declined significantly, especially since late-2000, consistent with improved BOM credibility.
- Monetary policy reaction function (estimated textual form):
  - 1 0.9890.011 0.64740.93010.0096 ttt t tSt SSqyuπ − =+++ + (as presented in source).
- Key parameter estimates (selected, with p-values in square brackets):
  - Monetary policy:
    - Sρ 0.9890 [5.1exp(-25)]
    - qγ 0.6474 [0.9985]
    - πγ 0.9301 [0.8678]
    - yγ 0.0096 [0.9999]
    - uρ 1.17exp(-5) [1.0000]
  - Aggregate supply (Phillips curve):
    - πμ 0.87414 [0.2054]
    - 1πα -0.0123 [0.9941]
    - 2πα 0.5791 [0.9110]
    - qα -0.2509 [0.9969]
    - yα -0.076 [0.9976]
  - Aggregate demand:
    - yμ 0.0663 [0.4243]
    - 1yβ 0.0632 [0.0000]
    - 2yβ 0.9368 [0.0000]
    - qyβ -0.0003 [0.9992]
    - rβ -0.0085 [0.3687]
  - Real exchange rate dynamics:
    - 1qβ 0.9447 [0.6841]
    - 2qβ 0.9885 [0.0018]
    - qSβ 0.0038 [0.7554]
  - Risk pricing:
    - LL1λ -0.1338 [0.6173]
    - LS1λ 0.5793 [0.1659]
    - SL1λ -1.06exp(-5) [0.9999]
    - SS1λ -0.0014 [0.9976]
- Standard deviations (selected):
  - Lσ 0.2950 [0.0120]
  - Sσ 0.2631 [0.0505]
  - πσ 5.3179 [1.9exp(-16)]
  - qσ 0.0607 [0.1399]
  - yσ 0.0222 [4.6exp(-14)]
  - Measurement error for the 6-month bond yield: 6σ 0.1956 [1exp(-20)]
- Measurement, estimation approach, and no-arbitrage conditions:
  - 6-month yield assumed measured with error (measurement error distributed N(0, 6σ)).
  - State-space representation and Sims (2001) algorithm used to transform the system; state follows ttt FFρεΣ += −1.
  - No-arbitrage affine bond-pricing recursion applied following Rudebusch and Wu (2003) with short rate expressed as a linear function of the state (tt FBAi ' 11 1 −−=).

### Interpretation and policy-relevant implications
- Expectations and inflation:
  - Expectations play a preeminent role in driving inflation in Mauritius (πμ 0.87414).
  - Inflation expectations are extremely persistent (Lρ 0.9921), implying reputation effects of the BOM are long-lasting and actual inflation carries a very small weight in expectation formation.
- Monetary policy stance and objectives:
  - High persistence in the slope factor (Sρ 0.9890) implies considerable inertia in the BOM’s stated monetary policy setting and relatively weak control over the short-term interest rate (−Sρ1 0.011 as reported in text).
  - Estimated policy parameters suggest possible dual objectives (inflation and exchange rate considerations), but parameter estimates are imprecise and none are significant.
- Real exchange rate and output:
  - Impact of real exchange rate changes on output is weak (qyβ -0.0003), possibly reflecting a flat real exchange rate over the inflation targeting lite period and administered prices for imports.
- Practical policy conclusions:
  - Mauritius’s inflation targeting lite regime is associated with a fall in inflation and a pronounced fall in inflation expectations, indicating improved BOM credibility.
  - Anchoring inflation expectations is particularly important in Mauritius due to centralized wage bargaining; anchoring can influence inflation far into the future.
  - Inflation targeting lite, applied consistently and credibly, can be a feasible alternative to a hard peg for small island economies, but performance may be sensitive to global inflationary environments and the time required to build central bank credibility.

*Source: Bank of Mauritius Annual Reports and material from the IMF working paper chapter titled "2. Bank of Mauritius Foreign Exchange Market Intervention, 1996/97–2003/04."*

### Appendix I .............................................................................................................

### Appendix I

### Appendix I
- Appendix I ..............................................................................................................................19

### Appendix II
- Appendix II .............................................................................................................................21

### References
- References...............................................................................................................................23

### Figures
- 1. Annual CPI Inflation.....................................................................................................4
- 2. Nominal and Real Effective Exchange Rates ...............................................................9
- 3. The Level and Slope of the Yield Curve ....................................................................17

### Tables
- 1. CPI Inflation Targets and Outcomes.............................................................................8

*https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2005/_wp05172.pdf*

### 2. Bank of Mauritius Foreign Exchange Market Intervention, 1996/97–2003/04...........11

### 2. Bank of Mauritius Foreign Exchange Market Intervention, 1996/97–2003/04

### I. Introduction — context and headline outcomes
- Mauritius’s monetary policy regime since 1996 is best characterized as “inflation targeting lite,” combining a public annual CPI inflation objective with continued managed exchange rate intervention.
- Inflation (year-on-year) fell from around 8 percent in mid-1996 to around 4 percent in the year to March 2004.
- Output growth averaged more than 5 percent per year over the same period.
- The shift toward inflation targeting lite appears associated with a fall in inflation principally through the reduction of inflationary expectations.
- Institutional and macro-financial context highlighted:
  - Government debt estimated at almost 60 percent of GDP in 2003/04.
  - Moderate (and declining) external debt.
  - Since 2002 the Bank of Mauritius (BOM) has been a net debtor of the government (the government has maintained substantial (net) deposits with the BOM).
  - Mauritius has a developed on-shore and growing off-shore banking sector and a primary dealer network for treasury bills.
  - Central statistical office produces good and timely macroeconomic data.
- Practical implication: Mauritius’s experience offers an example for small open island economies of controlling inflation while retaining exchange rate management flexibility.

### II. Development of the inflation targeting lite framework
- Historical evolution (high-level):
  - BOM established by Bank of Mauritius Act in 1967; expanded supervisory powers in 1971.
  - Prior to the 1990s monetary policy relied on direct instruments (credit ceilings, administered rates).
  - Financial liberalization in late 1980s and early 1990s: liberalization of exchange controls, inter-bank FX market established, basket-peg replaced by more flexible exchange regime (completed in 1994).
  - Credit ceilings abolished in July 1993.
  - In 1996 BOM introduced an “informal inflation targeting regime” (inflation targeting lite).
- Characteristics of Mauritius’s inflation targeting lite (aligned with Stone (2003) definition):
  - Multiple monetary policy objectives with increasing commitment to price stability.
  - Public announcement of an annual aggregate (CPI) inflation target since 1996/97.
  - Flexible monetary policy including exchange rate management and FX intervention.
  - Integrated operational strategy using many variables (including FX intervention) and reserve money program.
  - Relatively transparent monetary policy operations (annual reports, monthly statistical bulletin, public communications).
- Limitations relative to full-fledged inflation targeting:
  - Central bank mandate and institutional framework do not fully subordinate other objectives to price stability.
  - Only a forward-looking 12-month target is announced; lack of a medium-term trajectory and detailed analytical inflation-forecasting framework at the time.
  - BOM planned to further refine inflation-forecasting methodology and analysis of transmission mechanisms.

### III. Public CPI inflation targets and outcomes (1996/97–2003/04)
- BOM adopted a gradualist strategy to lower announced CPI targets, aiding credibility by setting realistic sequential targets and accommodating known one-off effects (example: VAT introduction on September 7, 1998).
- BOM met preannounced inflation objectives in all years except 2001/02 (missed due to unanticipated VAT increases and a cyclone).
- Table: CPI Inflation Targets and Actual Outcomes (as presented)
  - 1996/97 — Target inflation: 8.0; Actual inflation: 7.9
  - 1997/98 — Target inflation: 6.0; Actual inflation: 5.4
  - 1998/99 — Target inflation: 8.0; Actual inflation: 7.9
  - 1999/00 — Target inflation: 6.0; Actual inflation: 5.3
  - 2000/01 — Target inflation: 5.0-5.5; Actual inflation: 4.4
  - 2001/02 — Target inflation: 6.0; Actual inflation: 6.3
  - 2002/03 — Target inflation: 6.0; Actual inflation: 5.1
  - 2003/04 — Target inflation: 4.0; Actual inflation: 4.0

### IV. Monetary operations and exchange rate policy
- Integrated operational strategy:
  - Reserve money program announced in 1996 to maintain the monetary base on a path consistent with the inflation target and growth forecasts (but reserve money targets were never announced due to difficulty achieving them).
  - BOM uses a mix of instruments (primary government bond auctions, repo/reverse repo, occasional secondary market operations).
  - Treasury (and BOM) bills issued at weekly auctions in 3-, 6-, 12-, and 24-month maturities; Development Loan Stock up to 15 years issued annually; five-year bonds issued quarterly; BOM began issuing its own bills from July 2003.
- Signaling:
  - Lombard facility used to signal stance of monetary policy; Lombard rate historically influenced market rates though the facility was less used when money market rates were much lower than the Lombard.
- Exchange rate management and FX intervention:
  - BOM maintains FX interventions to target real effective exchange level and to minimize excessive volatility in a very thin FX market.
  - Interventions are integrated with open market operations through the reserve money program and have generally declined over the period, implying increased policy flexibility and a shift toward inflation targeting.
- Nominal and real effective exchange rate behavior:
  - Over the last eight years prior to the study the nominal effective exchange rate depreciated steadily while the real effective exchange rate remained broadly stable.

### V. Bank of Mauritius foreign exchange market interventions (1996/97–2003/04)
- Reported interventions (In millions of U.S. dollars) — Table 2 figures (as presented)
  - Sales:
    - 1996/97: 392.1
    - 1997/98: 252.4
    - 1998/99: 126.6
    - 1999/00: 94.0
    - 2000/01: 118.3
    - 2001/02: 19.0
    - 2002/03: 0
    - 2003/04: 0
  - Purchases:
    - 1996/97: 38.9
    - 1997/98: 37.0
    - 1998/99: 11.0
    - 1999/00: 15.7
    - 2000/01: 0
    - 2001/02: 21.1
    - 2002/03: 189.0
    - 2003/04: 68.0
- Interpretation:
  - General decline in FX intervention over the period, with evidence of one-sided intervention early on and more symmetric, infrequent interventions later.
  - Given a thin FX market, some intervention may remain necessary to smooth excessive volatility.

### VI. Transparency and further institutional steps
- Transparency measures already in place:
  - Frequent consultations between BOM and commercial banks; informative BOM website; comprehensive annual report; monthly statistical monetary bulletin; public speeches by the Governor.
  - Introduction of Lombard facility, repo and reverse repo frameworks intended to increase transparency.
- Steps BOM planned/needed to formalize the regime (as described in the content):
  - Further refinement of inflation-forecasting methodology.
  - Further analysis of the transmission mechanism between monetary policy instruments and inflation.
  - Additional transparency consistent with formal inflation-targeting frameworks.

### VII. Open-economy macrofinance model used for empirical analysis
- Modeling approach:
  - A new open-economy macrofinance model combines an affine no-arbitrage term structure (yield curve) with a standard open-economy new-Keynesian macroeconomic model.
  - Explicit modeling of the yield curve allows extraction of information on inflation expectations and the link between policy interest rates and longer-term yields.
- Key modeling structure and interpretation (equation references as in source):
  - Short-term policy rate captured by two latent term-structure factors tL and tS in equation (1): it = δ0 + tL + tS.
    - tL interpreted as the level of the yield curve and proxy for inflationary expectations.
    - tS interpreted as the slope of the yield curve (spread between long- and short-term rates) and movements in tS reflect monetary policy choices and real interest rate fluctuations.
  - Aggregate demand equation (new-Keynesian open economy) — equation (2): output gap responds to expected output, real exchange rate q, and real interest rate (tL − i).
  - Expectations updating — equation (3): tL evolves with weight on past expectations and actual inflation, with stochastic influences.
  - Phillips curve (open-economy supply) — equation (4): current inflation reflects forward-looking expectations (tL), backward-looking lags of inflation, the output gap, and changes in competitiveness.
  - Real exchange rate dynamics and expectations — equations (5) and (6)/(6’): current real exchange rate determined by expected future real exchange rate and ex ante real interest differential; inertia in q formation included.
  - Monetary policy rule for BOM — equation (7)/(9): the policy interest rate rule depends on the output gap, inflation expectations tL, the deviation between actual inflation and inflation expectation (πt − tL), the real exchange rate q, and smoothing parameters; reformulated to express in terms of tS in equation (9).
- Empirical estimation:
  - Model estimated by maximum likelihood over July 1996 to March 2004.
  - Data used: monthly government bond yields for 3-, 6-, and 12-month maturities (provided by BOM); annualized CPI inflation from the IMF’s International Financial Statistics; real exchange rate from the IMF’s Information Notice System.
  - Due to lack of high-frequency GDP data, detrended, seasonally adjusted, real monthly credit to the private sector (deflated by the CPI) used as a proxy for the output gap.

*Source: Bank of Mauritius Annual Reports and material from the IMF working paper chapter titled "2. Bank of Mauritius Foreign Exchange Market Intervention, 1996/97–2003/04."*

### Appendix II presents the parameter estimates of the open economy macrofinance model.

### _wp05172 - Appendix II presents the parameter estimates of the open economy macrofinance model.

### Dynamics of latent factors and inflation expectations
- The factor representing inflationary expectations, tL, is extremely persistent with Lρ 0.9921 [3.3exp(-57)].
- With inflation expectations so persistent, actual inflation carries a very small weight in inflation expectation formation; reputation effects of the Bank of Mauritius (BOM) are long-lasting.
- The slope factor tS is also highly persistent with Sρ 0.9890 [5.1exp(-25)], implying considerable inertia in the BOM’s stated monetary policy setting and relatively weak control over the short-term interest rate (−Sρ1 0.011 as reported in text).
- Unanticipated monetary shocks, Stu, are uncorrelated (uρ 1.17exp(-5) [1.0000]).
- The level of the yield curve (tL) is time-varying; average expected inflation and its fluctuation declined significantly, especially since late-2000, consistent with an improvement in BOM credibility.

### Monetary policy reaction function (estimated)
- Estimated BOM monetary policy rule (textual form): 1 0.9890.011 0.64740.93010.0096 ttt t tSt SSqyuπ − =+++ + (as presented in source).
- Parameter estimates (Appendix II, monetary policy reaction function):
  - Sρ 0.9890 [5.1exp(-25)]
  - qγ 0.6474 [0.9985]
  - πγ 0.9301 [0.8678]
  - yγ 0.0096 [0.9999]
  - uρ 1.17exp(-5) [1.0000]
- The estimates suggest possible dual monetary policy objectives (consistent with an inflation targeting lite regime), but the parameters are imprecisely estimated and none are significant.

### Aggregate supply (Phillips curve) and demand estimates
- Aggregate supply (Phillips curve) findings:
  - πμ 0.87414 [0.2054] (current inflation depends primarily on inflationary expectations)
  - 1πα -0.0123 [0.9941]
  - 2πα 0.5791 [0.9110]
  - qα -0.2509 [0.9969]
  - yα -0.076 [0.9976]
- Aggregate demand curve estimates (Appendix II):
  - yμ 0.0663 [0.4243]
  - 1yβ 0.0632 [0.0000]
  - 2yβ 0.9368 [0.0000]
  - qyβ -0.0003 [0.9992]
  - rβ -0.0085 [0.3687]
- Interpretation:
  - Expectations play a preeminent role in driving inflation in Mauritius (πμ 0.87414).
  - Forward-looking output expectations receive negligible weight at monthly frequency (yμ 0.0663).
  - The impact of real exchange rate changes on output is weak (qyβ -0.0003), possibly reflecting a flat real exchange rate over the inflation targeting lite period and administered prices for imports.

### Real exchange rate dynamics and risk pricing
- Real exchange rate dynamics (Appendix II):
  - 1qβ 0.9447 [0.6841]
  - 2qβ 0.9885 [0.0018]
  - qSβ 0.0038 [0.7554]
- Risk pricing matrix (Appendix II):
  - LL1λ -0.1338 [0.6173]
  - LS1λ 0.5793 [0.1659]
  - SL1λ -1.06exp(-5) [0.9999]
  - SS1λ -0.0014 [0.9976]

### Measurement, estimation approach, and no-arbitrage conditions
- Model estimation uses data on inflation, private sector credit, real exchange rate, and yields on 3-, 6-, and 12-month treasury securities; the 6-month yield is assumed measured with error (measurement error distributed N(0, 6σ)).
- State-space representation and Sims (2001) algorithm are used to transform the system; the state follows ttt FFρεΣ += −1.
- No-arbitrage affine bond-pricing recursion is applied following Rudebusch and Wu (2003) with short rate expressed as a linear function of the state (tt FBAi ' 11 1 −−=).

### Parameter estimates, standard deviations, and measurement error (Appendix II)
- Selected parameter estimates (values with p-values in square brackets):
  - Aggregate demand:
    - yμ 0.0663 [0.4243]
    - 1yβ 0.0632 [0.0000]
    - 2yβ 0.9368 [0.0000]
    - qyβ -0.0003 [0.9992]
    - rβ -0.0085 [0.3687]
  - Aggregate supply:
    - πμ 0.87414 [0.2054]
    - 1πα -0.0123 [0.9941]
    - 2πα 0.5791 [0.9110]
    - qα -0.2509 [0.9969]
    - yα -0.076 [0.9976]
  - Inflation dynamics:
    - Lρ 0.9921 [3.3exp(-57)]
  - Real exchange rate dynamics:
    - 1qβ 0.9447 [0.6841]
    - 2qβ 0.9885 [0.0018]
    - qSβ 0.0038 [0.7554]
  - Monetary policy:
    - Sρ 0.9890 [5.1exp(-25)]
    - qγ 0.6474 [0.9985]
    - πγ 0.9301 [0.8678]
    - yγ 0.0096 [0.9999]
    - uρ 1.17exp(-5) [1.0000]
  - Risk pricing:
    - LL1λ -0.1338 [0.6173]
    - LS1λ 0.5793 [0.1659]
    - SL1λ -1.06exp(-5) [0.9999]
    - SS1λ -0.0014 [0.9976]
- Standard deviations (Appendix II):
  - Lσ 0.2950 [0.0120]
  - Sσ 0.2631 [0.0505]
  - πσ 5.3179 [1.9exp(-16)]
  - qσ 0.0607 [0.1399]
  - yσ 0.0222 [4.6exp(-14)]
- Standard deviation of measurement error for the 6-month bond yield:
  - 6σ 0.1956 [1exp(-20)]

### Conclusions and policy-relevant implications (from source)
- Mauritius’s inflation targeting lite regime is associated with a fall in inflation and a pronounced fall in inflation expectations, indicating improved BOM credibility.
- Inflationary expectations are particularly important in Mauritius due to centralized wage bargaining; anchoring expectations can influence inflation far into the future.
- Inflation targeting lite, applied consistently and credibly, can be a feasible alternative to a hard peg for small island economies, but performance may be sensitive to global inflationary environments and the time required to build central bank credibility.

*Source: Appendix II and related sections of the provided IMF Working Paper content unit.*

---


_Source: https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2005/_wp05172.pdf_