## _wp05132

## Source details

**Canonical URL:** [_wp05132](https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2005/_wp05132.pdf)

## Other formats

- [Markdown version](/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2005/_wp05132.pdf.md)
- [Structured JSON version](/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2005/_wp05132.pdf.json)

---

### I. Introduction — purpose and key conclusions
- Purpose:
  - Analyze the role of exchange rate policy and regulations in Chile’s macroeconomic framework over the last decade.
  - Section breakdown: II — nature and evolution of exchange rate policy and capital account regulations and effects of external financing shocks; III — empirical analysis on the real exchange rate (RER); IV — policy recommendations and concluding remarks.
- Key conclusions:
  - Gradual liberalization of Chile’s capital account was carried out opportunistically, in parallel with increased exchange rate flexibility and after important conditions had been met.
  - Capital account regulations were applied to support an independent monetary policy.
  - The policy framework aimed at stabilizing the RER appears to have been of limited effectiveness; ample swings in the RER were induced by surges and sudden stops in capital flows.
  - Financial policy variables did not appear to have affected the RER trend.
  - The unremunerated reserve requirement (URR) and foreign exchange intervention had only temporary effects on the RER trajectory toward a new equilibrium.
  - Given the weak and short-lived effect of foreign exchange intervention on the RER, devoting intervention policy to attain a specific level of the RER appears inconvenient.
  - Capital account restrictions can be used only transitorily to smooth adjustment and not as a permanent policy variable.
  - Fiscal policy that stabilizes the growth rate of government spending throughout the business cycle and trade opening are significant potential tools to moderate RER appreciation induced by capital flow surges.

### II. External financing shocks, exchange-rate regimes, and capital-account regulations
- Late 1980s to 1999 exchange-rate management:
  - Peso kept within a crawling band around an implicit RER target (CPER) from mid-1980s until September 1999.
  - Band width history:
    - +/- 5% in 1990
    - +/- 10% in 1992
    - +/- 12.5% in 1997
    - June 1998 transitorily narrowed to +2% and -3.65% around the CPER
    - September 1998 continuous widening from an initial band width of +/-4%
    - December 1999 free floating of the Chilean peso
  - CPER redefinitions:
    - 1992: basket including deutsche mark (DM) and Japanese yen (JY)
    - 1997: increased U.S. dollar weight; recalculated CPER was 2% more appreciated
- Inflation targeting:
  - Central Bank implemented inflation targeting in 1991.
  - Long-term objective: reduce inflation (historically around 30 percent) to single-digit rates.
  - Inflation fell to single digits in 1996.
  - Steady-state inflation target beginning in 1999: a range between 2 and 4 percent.
- Foreign exchange intervention and reserves:
  - Central Bank increased net foreign exchange position by more than US$17 billion from the beginning of the 1990s to 1997 peak.
  - NIR peaking at US$ 20 billion, representing more than 12 months of imports and 25% of GDP.
  - Sterilized intervention contributed to quasi-fiscal losses; additional operating losses estimated at around 0.5% of GDP.
- Exchange-market segmentation and enforcement:
  - Formal exchange market (MCF) for registered transactions; a legal but informal market operated at a freely floating rate.
  - Spread between formal and informal market exchange rates:
    - 1980s: around 20 percent
    - Early 1990s: around 5 percent
    - 1995: below 2 percent
    - April 1997: spread disappeared when all exchange restrictions were discontinued

### Box 1 — Unremunerated Reserve Requirement (URR) and related FX restrictions
- URR design and coverage:
  - Introduced in 1991 as an unremunerated reserve requirement on capital inflows.
  - Required a compulsory and non-remunerated deposit in foreign exchange at the Central Bank for one year.
  - Applied to almost all debt-creating flows and to some portfolio inflows.
  - Coverage:
    - Immediately after 1991 imposition: less than 40% of gross inflows covered.
    - 1992 adjustments: increased coverage to 60% of gross capital inflows.
    - 1994: coverage fell to 30%.
    - 1995–1996 adjustments: coverage increased to 40% and remained around that value for the remainder of the URR period.
- Circumvention and enforcement:
  - Major legal exemptions: foreign direct investment (FDI) and import suppliers’ loans.
  - Illegal evasion possible; balance of payments errors and omissions remained significant but stable, not showing a circumvention pattern.
  - Closing FDI and suppliers’ credit loopholes would have doubled URR coverage — additional flows valued at US$ 4,000 million per year in 1996 and 1997 would have been covered.
- Liberalization timeline:
  - 1990: all export proceeds to be sold in formal market within 90 days.
  - Exporter liberalization completed in April 1995.
  - Compulsory sale of FX proceeds from capital inflows gradually eliminated and transformed into information obligations.
  - Import financing minimums and quantitative limits phased out; quantitative limit for market access eliminated in April 1997 (previous limit US$3,000 per month per individual for foreign travel and other services in 1990).
  - 1997 onward: restrictions on institutional investors and purchases of foreign exchange for investments abroad gradually lifted or transformed into limits to exchange risk exposure.
- Phase-out:
  - After the Asian crisis and end of the surge in inflows, generalized URR application discontinued.
  - 1998: URR rate reduced first to 10% and then to zero.
  - 1999: minimum withholding period for foreign investment eliminated.
  - 2001: all remaining restrictions discontinued and transformed into statistical information requirements.
- Outcomes:
  - Degree of international financial integration (foreign assets and liabilities basis) higher than average for emerging market economies with investment grade at liberalization end.
  - Chile’s debt indicators improved continuously and remained at the top of emerging markets.
  - 1996 specific data point: balance of payment surplus amounted to only US$ 1000 millions; minimized by prepayments of public foreign debt for about US$ 3000 million — without them, surplus would have been precisely US$ 4 billions.

### Box 2 — Capital inflow measures, RER volatility, and macro implications
- Comparative RER volatility (in percent) — sample 1990 - 2003 (author calculations on IMF monthly RER series):
  - Colombia 15.17%, Chile 10.67%, Malaysia 9.08%, Philippines 10.86%, Poland 21.49%, Singapore 5.10%.
  - Sample 1990 - 1997: Chile 10.42%.
  - Sample 1998 - 2003: Chile 10.62%.
  - Max for Chile: 12.72%; Min for Chile: -9.80%; Std Deviation Chile: 7.11%.
- Chile capital flows and macro indicators:
  - Three-year moving average net capital flows (NKF) peaked in 1981 and again in 1997; minimum inflows observed in 1988 and again in 2001 and 2003.
  - Surge period (first half of 1990s): annual net capital inflows averaged 7.3 percent of GDP.
  - Cost of external financing for domestic borrowers fell from 1.8% over LIBOR in 1991 to 0.7% above LIBOR in 1996-97.
  - Private domestic expenditure (¾ of domestic demand) doubled after growing at an annual average rate of 10% in 1991-1997.
  - Domestic 90-day deposit rate average in 1991-97: 6.5%.
  - Fiscal surpluses averaged 2 percent of GDP in 1991-1997.
  - After mid-1998 end of surge: large peso depreciation, sharp monetary tightening; fiscal balance deteriorated to a small deficit mainly due to lower revenues from recession.
- Capital inflow regulations in addition to URR:
  - Minimum withholding periods for FDI and portfolio flows.
  - Limits on issuance of publicly traded instruments in foreign markets (ADRs authorized in 1990 with issuance amount and minimum credit rating restrictions).
  - Restrictions on denomination of foreign liabilities; 1997 amendment authorized denomination in Chilean pesos and UF.
  - Exemptions for individual foreign operations below US$ 200.000 when accumulated registered capital inflows under US$ 500.000 in last 12 months.
- URR evolution and cost:
  - Initially URR rate 20% tied to loan maturity; soon unified to one year and rate raised to 30% (kept until 1998 when reduced to 0).
  - From 1995 URR deposit had to be constituted in U.S. dollars.
  - Maximum URR rate 30% with one-year deposit could deter arbitrages for interest differentials up to 350 basic points for one-year operations (calculation based on a 7 percent relevant external interest rate).
- Policy trade-offs:
  - Sterilized intervention to defend exchange rate level:
    - Does not change private incentives to hold FX liabilities; can favor speculation, generate large open foreign asset positions for the central bank and significant losses.
  - Non-sterilized intervention:
    - Lowers interest rates and can contain inflow surges by reducing attractiveness of foreign liabilities, but may conflict with inflation targets and imply an expansionary stance inconsistent with inflation control.
    - Chile used non-sterilized intervention after the 1998 sudden stop to defend the narrowed band, at the cost of an extremely high real interest rate, drop in activity, and a sharp reduction in core inflation from 6.0 percent in June 1998 to 2.1 percent in December 1999 (target was slightly above 4.3 percent).
  - URR effectiveness:
    - Despite limitations, URR was effective in containing capital inflows and had a positive and statistically significant effect on the real interest rate, giving monetary policy additional room.
    - URR’s effect concentrated on short-term capital flows but was strong enough to modify total capital flows.
  - Political constraints limited use of very contractionary fiscal policy to curb excessive private spending.

### D. Preconditions and sequencing for liberalization and float
- Preconditions for successful float and financial opening:
  - Advance financial opening in parallel with exchange rate flexibility to preserve monetary independence (impossible trinity).
  - Opportunity timing (when risk of large appreciation/depreciation is minimal).
  - A credible nominal anchor alternative to the exchange rate (credible monetary policy).
  - Limited exposure to exchange rate risk in the financial system.
  - Development of liquid local FX and hedging markets.
- Chile’s sequencing:
  - Liberalized capital account after macroeconomic and financial stability: improved external solvency and liquidity metrics, low inflation, fiscal consolidation, and strengthened financial system.
  - Limits imposed on bank currency asset-liability mismatches; supervision included provisions for direct and indirect exchange-rate exposure.
  - Development of liquid local FX instruments from 1995 onward.
  - Limited financial dollarization; domestic markets developed around domestic-currency short-term instruments and inflation-indexed long-term instruments.
  - Emphasis on transparency and prospective reporting: monetary policy report thrice a year, quarterly fiscal reports, Article IV documents.

### III. Empirical analysis — arbitrage, macro fundamentals, and ECM results
- Arbitrage expression and stylized returns:
  - Arb[LRER(t)] = E[LRER(t+1)] − (rdom − rext)/12
  - rext = 90libo − *ΠE + srcl
  - On average 1990-2003, annual returns on domestic currency assets exceeded foreign currency assets by around 240 basis points.
  - During URR application (1991.07-1998.09), the excess exceeded 300 basis points.
- Two-stage least squares monthly econometrics (1990.1-2003.11) — preferred specification (Equation 1.2) for LRER:
  - Arb[rer] coefficient: 1.0589 (p-value 0.000).
  - Fc coefficient: 0.4358 (p-value 0.002).
  - Adjusted R2: 0.966.
  - Durbin-Watson statistic: 1.333.
  - Dickey-Fuller estimation residuals: -9.102 (cointegration significant).
  - Excluding fc is rejected at 1 percent significance; fc is a significant explanatory variable for short-term RER.
  - Foreign exchange intervention (Lfexint) does not show a statistically significant effect on RER level within the arbitrage model.
- Macro RER model (quarterly, two-stage least squares, 1990.2-2003.3) — parsimonious Equation 2.3 (all coefficients significant at 1%):
  - Constant: 16.0374 (p 0.000)
  - LPROD: -1.2907 (p 0.000)
  - NFA: -0.2176 (p 0.005)
  - LTT: -0.3281 (p 0.003)
  - GOBMY: -1.4774 (p 0.013)
  - LTARI: -4.7295 (p 0.000)
  - FKN_ABM: -0.0014 (p 0.000)
  - Volrdom: -4.4666 (p 0.000)
  - Adjusted R2: 0.832
  - Durbin-Watson statistic: 1.941
  - Dickey-Fuller estimation residuals: -7.014 (cointegration significant at 1%)
- Cointegration and model selection diagnostics:
  - Equation 2.2 yields F statistic value of 3.65, and Chi Squared statistic value of 25.57.
  - Exclusion of five variables from 2.2 to form 2.3: F statistic 1.71 and Chi-Squared 8.53 when comparing Eq. 2.2 and 2.3; exclusion not rejected at 10% significance.
  - Critical values referenced: Engel-Granger test -4.348 (10% for five variables and 50 observations); extrapolation to 13 variables yields -8.839. Critical at 1% for 8 variables and 50 observations is -6.814; ADF residual for 2.3 is -7.014.
- Error-Correction Model (ECM) based on Equation 2.3 (OLS, 1991.1-2003.3) — reported coefficients and p-values preserved verbatim:
  - dRER(t) = -0.5608*dProdl (0.000) - 0.7634 *dprodl(t-1) (0.0000) - 0.1992* [dtt(t)+dtt(t-1)] (0.000)
  - - 0.1524*[dnfa (t) + dnfa (t-2)] (0.0000) - 0.8136*[dgobmy + dgobmy(t-2)] (0.0000) - 1.5064*dgobmy (t-1) (0.000)
  - - 3.715*dtari (t-2) (0.0000)
  - - 0.0006*[df_abm (t)+df_abm(t-2)+df_abm(t-3)] (0.0000)
  - +0.6065*[drdom (t-3)- drext (t-3)- dfc (t-3)+ drdom (t-4)- drext (t-4)- dfc(t-4)] (0.0000)
  - -2.0631*[drext(t)- drext (t-2)] (0.0000)
  - - 0.1022 *[dfexi (t-1)- dfexi (t-3)] (0.0001)
  - - 0.3074*RESID(-1) .(0.0000)
  - Model diagnostics: Adj R-Squared  0.78; Durbin Watson 2.06
  - Interpretation: adjustment velocity of the RER toward equilibrium is rather slow, allowing persistence.
- Specific dynamics and magnitudes from ECM:
  - URR financing cost (fc) sustained change of 100 basic points:
    - Depreciated the RER by 0.6 percent after three quarters.
    - After ten quarters total effect fell to 0.1 percent, disappearing afterwards.
  - Real domestic interest rate had effect of same magnitude but opposite sign to fc.
  - Foreign exchange intervention (sustained change defined as 10 percent of intervention stock):
    - Depreciated the RER by 1 percent after one quarter.
    - Effect vanishes very rapidly and reversed to an appreciation of 0.5 percent after three quarters.
- Cost-effectiveness and fiscal implications:
  - URR yielded 0.2 to 0.3 percent of GDP in revenue to the Central Bank during its application.
  - Net return of sterilizing net purchases of foreign exchange financed at the domestic real rate historically negative, with historical average -400 basis points excluding the effect of RER changes.
  - In recent periods the cost of domestically financing NIR has fallen reflecting lower sovereign spreads and more accommodative domestic monetary policy.

### IV. Effects of sustained changes and policy implications
- Table — Effects on RER of sustained changes in key variables (unit: percent). Policy shock definitions: DFC = 100; DFEXI = 10; DGOBMY = 1; DF_ABM = 10; DTARI = 1; DTT = 1.
  - Period 1:
    - DFC: 0.00
    - DFEXI: 1.02
    - DGOBMY: -0.45
    - DF_ABM: -0.80
    - DTARI: -1.45
    - DTT: -0.44
  - Period 3:
    - DFC: 0.61
    - DFEXI: -0.53
    - DGOBMY: -4.94
    - DF_ABM: -2.04
    - DTARI: -5.42
    - DTT: -0.38
  - Period 6:
    - DFC: 0.49
    - DFEXI: -0.18
    - DGOBMY: -4.80
    - DF_ABM: -1.59
    - DTARI: -4.96
    - DTT: -0.35
  - Period 10:
    - DFC: 0.11
    - DFEXI: -0.04
    - DGOBMY: -4.75
    - DF_ABM: -1.42
    - DTARI: -4.78
    - DTT: -0.33
  - Period 20:
    - DFC: 0.00
    - DFEXI: 0.00
    - DGOBMY: -4.73
    - DF_ABM: -1.37
    - DTARI: -4.73
    - DTT: -0.33
- Key quantitative findings:
  - An additional US$10 billion in nkf_abm would have appreciated the RER by a cumulative 2 percent after 3 quarters, and by 1.4 percent over the long-term.
  - A 1 percent increase in the terms of trade resulted in a real appreciation of 0.4 percent in the short-run and 0.3 percent in the longer term.
  - Structural policies:
    - A 1 percent reduction in the tariff index may depreciate the RER by 5 percent over the long term.
    - A reduction of 1 percent of GDP in government spending would depreciate the RER by almost 5 percent over the long term.
    - Note: reported magnitudes for tariffs and government spending may represent overestimation but indicate relative importance of structural policies over financial policies for medium- and long-term RER determination.
- Policy lessons and recommendations:
  - Independent monetary policy is essential for Chile facing idiosyncratic shocks and inflation control objectives.
  - With a pegged exchange rate within a band, independent monetary policy is possible only by limiting international financial integration; capital account regulations complemented the quasi-pegged regime.
  - Gradual liberalization carried out opportunistically in parallel with increased exchange rate flexibility; elimination of outflow restrictions during inflow surges and elimination of inflow restrictions after sudden stops.
  - URR had short-run effects on the RER but only compensated for independent monetary policy; effect limited and transitory.
  - Foreign exchange intervention had weak and short-lived effects; inconvenient to defend RER levels against new surges.
  - Selective capital account restrictions used in Chile are transitorily effective to smooth adjustment but cannot be permanent instruments to control the RER.
  - Fiscal policy stabilizing government spending growth through the cycle and trade openness/tariff reductions are powerful tools to moderate RER swings induced by capital flows; structural policies matter more for medium- and long-term RER than financial policies.

### Appendix highlights — unit-root and data notes
- ADF unit-root test results (1990:1-2003:4) — select results preserved verbatim:
  - LRER: ADF With Constant, Without Trend = -1.95; ADF Constant and Trend = -1.42/
  - dRER: ADF With Constant, Without Trend = -6.76 1/; ADF Constant and Trend = -7.19 1/
  - Fc: ADF With Constant, Without Trend = -1.53; ADF Constant and Trend = -2.03
  - d Fc: ADF With Constant, Without Trend = -7.11 1/; ADF Constant and Trend = -7.20 1/
  - Ltari: ADF With Constant, Without Trend = 0.57; ADF Constant and Trend = -2.57
  - dtari: ADF With Constant, Without Trend = -9.16 1/; ADF Constant and Trend = -4.88 1/
- Notes on significance annotations (as presented):
  - 1/ The null hypothesis of a unit root is rejected at 1% significance.
  - 2/ Rejected at 5%.
  - 3/ Rejected at 10%
  - 4/ The null hypothesis of a unit root cannot be rejected using the ERS test at 10% significance level.
  - 5/ Using the Phillips Perron the unit root hypothesis is rejected at the 1% significance level.
  - 6/ The null hypothesis of a unit root cannot be rejected using the DF-GLS test at 10% significance level.

*Source: _wp05132 - References*

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

### _wp05132 - References

### I. Introduction
- Trade integration enhances efficiency in resource allocation and opens opportunities for growth and diversification; financial integration optimizes intertemporal consumption and risk management by increasing the array of assets available in local markets.
- Volatile capital flows can induce swinging financial conditions and crises; this motivates a prudent approach to financial liberalization.
- Repeated financial crises in developing countries that liberalized without preconditions (including Chile in 1982) cast doubt on immediate, unconditional liberalization.
- IMF and other studies (Eichengreen and others (1998), International Monetary Fund (1998), Prasad and others (2003)) specify costs, benefits, and preconditions for financial opening.
- Financial booms fueled by capital inflows can:
  - Fuel expansion in domestic aggregate demand that exceeds potential output.
  - Result in an unsustainably high current account deficit.
  - Cause swinging real exchange rate (RER) and a vulnerable banking system.
- Literature emphasizes current account reversals and sudden stops in capital flows as sources of instability (Edwards (2003); Milesi-Ferretti and Razin (1996); Calvo and Reinhardt (2000)).
- Purpose of the paper:
  - Analyze the role of exchange rate policy and regulations in Chile’s macroeconomic framework over the last decade.
  - Section II: nature and evolution of exchange rate policy and capital account regulations and effects of external financing shocks.
  - Section III: empirical analysis on the real exchange rate, including effects of fundamentals, financing conditions, and policy intervention.
  - Section IV: policy recommendations and concluding remarks.
- Key conclusions summarized in the introduction:
  - Gradual liberalization of Chile’s capital account was carried out opportunistically, in parallel with increased exchange rate flexibility and after important conditions had been met.
  - Capital account regulations were applied to support an independent monetary policy.
  - The policy framework aimed at stabilizing the RER appears to have been of limited effectiveness; ample swings in the RER were induced by surges and sudden stops in capital flows.
  - Financial policy variables did not appear to have affected the RER trend.
  - The unremunerated reserve requirement (URR) and foreign exchange intervention had only temporary effects on the RER trajectory toward a new equilibrium.
  - Given the weak and short-lived effect of foreign exchange intervention on the RER, devoting intervention policy to attain a specific level of the RER appears inconvenient.
  - Capital account restrictions can be used only transitorily to smooth adjustment and not as a permanent policy variable.
  - Fiscal policy that stabilizes the growth rate of government spending throughout the business cycle and trade opening are significant potential tools to moderate RER appreciation induced by capital flow surges.

### II. External Financing Shocks and Macroeconomic Policy
- Exchange rate insurance against volatile capital flows has limits; under extreme conditions, runs on domestic assets can break exchange rate arrangements and lead to macroeconomic instability and financial system failures (example: Chile in 1982).
- To enhance exchange rate insurance effectiveness, attempts at moderating exchange rate volatility have sometimes been accompanied by restrictions on capital movements.
- In the late 1980s Chile:
  - Capital flows and holdings of foreign assets by residents were severely restricted.
  - Exchange rate was managed within a narrow crawling band, with the rate of crawl following the difference between domestic and external inflation.
- RER misalignments are associated with:
  - Boom-and-bust cycles in external financing, reversals in the current account deficit, slow-downs in economic activity, and potential financial crises.
  - Resource misallocation during excessive appreciation phases.
  - Inflationary pressures and development of non-competitive tradable activities during excessive depreciation phases.

### A. The Crawling Exchange Rate Band and Inflation Targeting
- From the mid 1980s until September 1999, the Chilean peso was kept within a crawling band defined around an implicit RER target, the central parity exchange rate (CPER).
- The CPER crawled with daily adjustments to reflect estimated inflation differentials between Chile and main trading partners.
  - Initially used last month’s CPI inflation for domestic inflation and a projection for trading partners’ WPI measured in U.S. dollar for external inflation.
  - In 1998 the lagged inflation rate was replaced by the inflation target in CPER calculation.
- The band often failed to allow floating; the exchange rate repeatedly hit band edges:
  - 1980s: band ceiling.
  - Most of 1991-97: band floor.
  - Briefly the ceiling in 1998 after transitory narrowing.
- Band width changes:
  - Increased from +/- 5% in 1990 to +/- 10% in 1992, and to +/- 12.5% in 1997.
  - June 1998: transitorily narrowed to +2% and -3.65% around the CPER.
  - September 1998: continuous widening from an initial band width of +/-4%.
  - December 1999: free floating of the Chilean peso.
- CPER redefinitions:
  - 1992: CPER redefined from U.S. dollar to a basket including deutsche mark (DM) and Japanese yen (JY) in addition to the dollar.
  - The DM and JY gained relative importance, increasing CPER volatility against the U.S. dollar.
  - 1997: currency basket modified sharply, increasing the U.S. dollar weight; resulting recalculated CPER was 2% more appreciated — this helped release pressure toward peso appreciation.
  - Despite basket changes, market exchange rate still often reached band edges.
- Inflation targeting:
  - 1991: Central Bank implemented an inflation targeting system.
  - Long-term objective: reduce inflation (historically around 30 percent) to single-digit rates.
  - Annual inflation targets were consistently met.
  - Inflation rate fell to single digits in 1996.
  - Steady-state inflation target beginning in 1999: a range between 2 and 4 percent.
- Conflict between pegged exchange rate and independent monetary policy led to capital flow controls (NKF).
- Foreign exchange intervention and reserves:
  - Central Bank increased its net foreign exchange position by more than US$17 billion from the beginning of the 1990s to the maximum point in 1997.
  - NIR peaking at US$ 20 billion, representing more than 12 months of imports and 25% of GDP.
  - Sterilized intervention contributed to quasi-fiscal losses; additional operating losses estimated at around 0.5% of GDP.

### B. Foreign Exchange and Capital Account Regulations
- Capital account regulations evolved opportunistically with changes in external financing and exchange rate policy to support monetary objectives.
- Regulations were backed by an institutional system able to enforce them; rules were transparent and non-discriminatory, published in the Central Bank Compendium of Foreign Exchange Regulations.
- System included registration of foreign exchange operations in the formal exchange market (MCF); all capital account transactions had to be channeled through MCF.
- Forex market segmentation existed: a legal but informal foreign exchange market operated at a freely floating rate often significantly different from the official rate.
- Main exchange restrictions included obligation to sell foreign exchange proceeds from exports and capital inflows in the banking system within strict time limits (see Box 1 in the source for measures related to foreign exchange restrictions).
- Spread between formal and informal market exchange rates as indicator of exchange restrictions:
  - 1980s: spread typically around 20 percent.
  - Early 1990s: spread fell to around 5 percent.
  - Following liberalization measures: spread fell below 2 percent in 1995.
  - April 1997: spread disappeared completely when all exchange restrictions were discontinued.

*Source: _wp05132 - References*

### Box 1. Measures Related to Foreign Exchange Restrictions

### Box 1. Measures Related to Foreign Exchange Restrictions

### Design and scope of the unremunerated reserve requirement (URR)
- URR introduced in 1991 as an unremunerated reserve requirement on capital inflows.
- URR required a compulsory and non-remunerated deposit in foreign exchange to be maintained at the Central Bank for a period of one year.
- URR applied to almost all debt creating flows and to some portfolio inflows, increasing the cost of external financing through the covered channels, particularly that of short-term indebtedness.
- Coverage of the URR fluctuated around 50% of capital inflows.
- Immediately after imposition in 1991, less than 40% of gross inflows were covered.
- Extensions and adjustments in 1992 increased URR coverage to 60% of gross capital inflows.
- Coverage fell to 30% by 1994.
- Policy adjustments in 1995 and 1996 increased coverage to 40%, and it remained around that value for the remainder of the URR period.

### Circumvention, leaks, and enforcement
- Legal circumvention reduced URR effectiveness; major exemptions included foreign direct investment (FDI) and import suppliers’ loans.
- Illegal evasion (capital flowing into Chile outside of the regulations) was possible; however, balance of payments errors and omissions and other assets flows, which represent an estimation of the repatriation of non-identified assets and are a proxy for non-registered inflows, remained significant but stable over time, not showing any pattern indicative of their use for circumventing the URR.
- Partial coverage may have originated microeconomic costs by further differentiating access to external financing.
- Closing main loopholes on FDI, suppliers’ credit and direct trade financing would have doubled URR coverage, as additional flows for a value of US$ 4,000 million per year in 1996 and 1997 would have been covered by the mechanism.

### Related exchange restrictions and their liberalization
- In 1990, all export proceeds were to be sold in the formal foreign exchange market within 90 days of shipment, without deductibles or minimum amounts.
- Over time, exporters were allowed to keep an increasing share of their foreign exchange proceeds and the period for compulsory sale in the formal market was extended; liberalization completed in April 1995.
- Compulsory sale of foreign exchange proceeds from capital inflows was gradually eliminated and transformed into an obligation to inform on the capital account transaction.
- Exchange restrictions on purchase of foreign exchange in the formal market included minimum financing periods and quantitative limits for importing goods and services:
  - In 1990 imports of goods were subject to 120 days of minimum financing.
  - Foreign travel and other imports of services were subject to a quantitative limit for accessing the formal exchange market of US$3,000, per month, per individual.
- Import financing requirements were promptly abolished.
- The quantitative limit for market access was gradually increased and completely eliminated in April 1997.
- Institutional investors (banks, mutual funds, pension funds and insurance companies) were subject to strict limitations in holdings of foreign assets; beginning in 1997, all restrictions limiting the purchase of foreign exchange in the formal markets to finance investments abroad were eliminated and restrictions on institutional investors were gradually lifted or transformed into limits to the exposure to exchange risk.

### Phase-out, timing, and broader outcomes
- After the Asian crisis and the end of the surge in capital inflows, efforts toward generalized application of the URR were discontinued.
- In 1998, after a very sharp reduction in capital inflows, the URR rate was reduced first to 10% and then to zero.
- In 1999 the minimum withholding period for foreign investment was eliminated.
- In 2001 all remaining restrictions and regulations were discontinued and transformed into a system of statistical information requirements.
- At the end of the liberalization process, the degree of international financial integration of the Chilean economy, measured on the basis of foreign assets and liabilities, was higher than the average for emerging market economies with investment grade.
- Chile’s debt indicators improved continuously over the period and remained at the top of emerging markets.
- Specific 1996 data point: balance of payment surplus amounted to only US$ 1000 millions; this figure is minimized by prepayments of public foreign debt for about US$ 3000 million that took place that year — without them, the surplus would have been precisely US$ 4 billions.

*Source: _wp05132 - Box 1. Measures Related to Foreign Exchange Restrictions_*

### Box 2. Capital Inflow Measures

### Box 2. Capital Inflow Measures

### C. Policy Responses to Surges and Sudden Stops

- Real exchange rate (RER) volatility is a central macro-relative price issue for emerging market economies; less-advanced integration raises vulnerability. Comparative RER volatility (in percent, sample periods shown):
  - 1990 - 2003: Colombia 15.17%, Chile 10.67%, Malaysia 9.08%, Philippines 10.86%, Poland 21.49%, Singapore 5.10%.
  - 1990 - 1997: Colombia 16.62%, Chile 10.42%, Malaysia 5.02%, Philippines 10.83%, Poland 20.09%, Singapore 5.58%.
  - 1998 - 2003: Colombia 13.02%, Chile 10.62%, Malaysia 4.85%, Philippines 7.50%, Poland 8.18%, Singapore 4.09%.
  - Max: Colombia 17.81%, Chile 12.72%, Malaysia 8.80%, Philippines 12.54%, Poland 22.63%, Singapore 6.71%.
  - Min: Colombia -15.37%, Chile -9.80%, Malaysia -9.74%, Philippines -8.56%, Poland -26.02%, Singapore -3.99%.
  - Std Deviation: Colombia 9.83%, Chile 7.11%, Malaysia 6.51%, Philippines 6.20%, Poland 13.62%, Singapore 3.21%.
  - Note: With IMF data, an increase in the RER index represents an appreciation (author calculation on IMF monthly RER series).

- Capital inflow regulations used in Chile (in addition to the URR) included:
  - Minimum withholding periods applied to foreign direct investment and to portfolio flows.
  - Limitations to issue publicly traded financial instruments in foreign markets (including ADRs initially authorized in 1990 with issuance amount and minimum credit rating restrictions).
  - Restrictions limiting denomination of foreign liabilities to authorized foreign currency; 1997 amendment authorized denomination of foreign indebtedness in Chilean pesos and the Chilean indexed unit, UF.
  - Coverage extensions to secondary ADR operations and many portfolio flows; loans used to pay expenses abroad were covered; loans to prepay foreign obligations or extend debt duration were exempt.
  - Exemptions: individual foreign operations below US$ 200.000 were exempt when the participating foreign investor had registered capital inflows below US$ 500.000 accumulated over the last 12 months.

- URR (Unremunerated Reserve Requirement) design and evolution:
  - Initially term of URR deposit tied to loan maturity; rate stood at 20 percent.
  - Term soon unified to one year irrespective of loan maturity and rate raised to 30% (kept at that level until 1998 when it was reduced to 0).
  - URR deposit could be substituted by payment of an upfront fee equivalent to the financial cost of the URR.
  - From 1995 the URR deposit had to be constituted in U.S. dollars, irrespective of currency of the loan.
  - Reductions in nominal interest rates in yens and in gold had created circumvention opportunities.
  - The URR financial cost varied with loan maturity (one-year term made fc higher for shorter maturities); with maximum URR rate 30% and deposit term of one year, the URR financial cost could deter arbitrages for interest differentials up to 350 basic points for one-year operations (calculation based on a 7 percent relevant external interest rate).

- Chile experience, capital flows and macro implications:
  - Chile experienced a large RER depreciation over 60 percent from 1981 to end of the 1980s following the early 1980s Latin American debt crisis, then strong appreciation reaching near early-1980s levels by 1997, then strong depreciation through March 2003 and a subsequent appreciation episode.
  - Three-year moving average net capital flows (NKF) peaked in 1981 and again in 1997; minimum capital inflows (end of sudden stops) observed in 1988 and again in 2001 and 2003 under this metric.
  - A surge in net capital inflows in the first half of the 1990s: annual net capital inflows averaged 7.3 percent of GDP during the surge period.
  - Cost of external financing for domestic borrowers fell from 1.8% over LIBOR for the average Chilean borrower in 1991 to 0.7% above LIBOR in 1996-97.
  - Rapid private domestic expenditure growth in 1991-1997: private domestic expenditure (¾ of domestic demand) doubled after growing at an annual average rate of 10%.
  - Domestic 90-day deposit rate average in 1991-97 stood at 6.5%, higher than in the second half of the 1980s or since the Asian crisis (1998-2000).
  - Fiscal surpluses averaged 2 percent of GDP in the same period (1991-1997).
  - After the abrupt end of the surge in mid-1998: large peso depreciation, sharp monetary tightening; fiscal balance deteriorated to a small deficit mainly due to recession-induced lower revenues.

- Policy responses and trade-offs:
  - Sterilized intervention to defend exchange rate level: does not change private incentives to hold FX liabilities and can favor speculation, generate large open foreign asset positions for the central bank, significant risks and sizeable losses.
  - Non-sterilized intervention lowers interest rates and can contain inflow surges by reducing attractiveness of foreign liabilities, but may conflict with inflation targets and could imply a more expansionary stance than consistent with inflation control. Chile used non-sterilized intervention after the 1998 sudden stop successfully to defend the narrowed band but at the cost of an extremely high real interest rate, drop in economic activity, and a sharp reduction in core inflation from 6.0 percent in June 1998 to 2.1 percent in December 1999 (target was slightly above 4.3 percent).
  - URR effectiveness: despite limitations, the URR was effective in containing capital inflows and had a positive and statistically significant effect on the real interest rate, giving monetary policy additional room. The URR’s effect concentrated on short-term capital flows but was strong enough to modify total capital flows.
  - Political constraints limited the use of very contractionary fiscal policy (steep expenditure cuts or pro-cyclical tax increases) to curb excessive private spending.

### D. Conditions to Liberalize and Float

- Institutional preconditions and sequencing emphasized before liberalization and floating:
  - Financial opening should advance in parallel with exchange rate flexibility to preserve monetary independence under the impossible trinity.
  - Preconditions for a successful float include: opportunity (preferably when risk of large appreciation/depreciation is minimal), a nominal anchor alternative to the exchange rate (credible monetary policy), limited exposure to exchange rate risk in the financial system, and financial markets development (liquid local FX and hedging markets).
  - Chile liberalized and completed capital account opening after achieving macroeconomic and financial stability: improved external solvency and liquidity metrics, low inflation, fiscal consolidation, and strengthened domestic financial system.
  - To limit balance-sheet FX mismatches, Chile imposed limits for asset-liability mismatches of banks in currencies as well as liquidity and interest rate mismatches relative to banks’ capital and reserves; supervision and regulation included special provisions for direct and indirect exposure of banks to exchange rate risk.
  - Development of liquid local FX instruments (spot and forward) expanded from 1995 onward with larger daily average transactions and longer contract maturities, enabling risk sharing and transfer to better-equipped counterparties.
  - Chilean financial dollarization was limited; local markets developed around domestic-currency nominal short-term instruments and inflation-indexed long-term instruments. Still, private sector external foreign debt created mismatches; reducing these requires deeper domestic peso markets or international issuance denominated in Chilean pesos.
  - Transparency and prospective reports on macro and external vulnerabilities were emphasized to reduce information problems that can trigger surges and sudden stops: publication of monetary policy report three times a year, quarterly fiscal reports, and dissemination of documents associated with the annual Article IV Consultation with the IMF.

### III. An Empirical Approach to Real Exchange Rate Dynamics

A. Arbitrage and Real Exchange Rate

- Financial arbitrage expression (log RER):
  - Arb[LRER(t)] = E[LRER(t+1)] − (rdom − rext)/12   (equation (1))
  - rext = 90libo − *ΠE + srcl   (equation (2))
  - rdom: short-term real interest rate in domestic currency (short-term CPI-indexed instruments).
  - rext: short-term real interest in foreign currency relevant for Chilean sovereign borrower; libo90 is 90-day LIBO, *ΠE expected US inflation, srcl Chile sovereign risk premium.

- Stylized facts on returns:
  - On average 1990-2003, annual returns on domestic currency assets exceeded foreign currency assets by around 240 basis points.
  - During URR application (1991.07-1998.09), the excess exceeded 300 basis points.
  - Excess return not associated with higher counterpart risk (rates represented relevant borrowing rates for Chilean sovereign).

- Econometric findings (monthly two-stage least squares, 1990.1-2003.11):
  - Preferred specification (Equation 1.2) for LRER includes Arb[rer] and the URR financing cost (fc).
    - Arb[rer] coefficient: 1.0589 (p-value 0.000).
    - Fc coefficient: 0.4358 (p-value 0.002).
    - Adjusted R2: 0.966.
    - Durbin-Watson statistic: 1.333.
    - Dickey-Fuller estimation residuals: -9.102 (co-integration significant).
  - Excluding fc is rejected at 1 percent significance; fc therefore a significant explanatory variable for short-term RER.
  - Interpretation: higher URR financing cost generates a more depreciated RER than implied by pure arbitrage conditions; foreign exchange intervention (Lfexint) does not show a statistically significant effect on RER level within the arbitrage model.
  - Error-correction dynamics (ECM based on Equation 1.2) show a strong and highly significant error-correction response to the lagged residual of the cointegration regression (ECM estimation: Adj. R2 0.93; strong significance on error-correction term).

B. Real Exchange Rate (RER) as a Relative Price — Macroeconomic Approach

- Five fundamentals for the equilibrium RER (quarterly model):
  - Net foreign asset position (NFA): higher NFA → higher domestic demand → appreciation (lower relative price of tradables).
  - Terms of trade (LTT): higher LTT → higher domestic income and demand → appreciation, though substitution effects can complicate sign.
  - Import tariffs (LTARI): higher protection → relative demand for tradables falls → appreciation.
  - Relative productivity (LPROD): higher productivity (primarily in tradables) → appreciation.
  - Government spending (GOBMY): more intensive in non-tradables → appreciation.

- Macro model estimations (Two-stage least squares, 1990.2-2003.3):
  - Equation 2.1 (fundamentals only): Adjusted R2 0.702; some coefficients significant but residuals show error persistence and not cointegrated at 10%.
  - Equation 2.2 (fundamentals + seven financial variables: nkf_abm, rdom-rext-fc, Lfexint, Dumband, Lcpr, Volrdom, Volrer): inclusion of financial variables rejected for exclusion at 1% significance, but multicollinearity and non-cointegration issues remained (Adj. R2 0.707; DW 1.843; Dickey-Fuller residuals -6.798).
  - Preferred parsimonious specification, Equation 2.3 (selected exclusions from 2.2) — all explanatory variables have expected signs and are statistically different from zero at 1%:
    - Constant: 16.0374 (p 0.000).
    - LPROD: -1.2907 (p 0.000).
    - NFA: -0.2176 (p 0.005).
    - LTT: -0.3281 (p 0.003).
    - GOBMY: -1.4774 (p 0.013).
    - LTARI: -4.7295 (p 0.000).
    - FKN_ABM: -0.0014 (p 0.000).
    - Volrdom: -4.4666 (p 0.000).
    - Adjusted R2: 0.832.
    - Durbin-Watson statistic: 1.941.
    - Dickey-Fuller estimation residuals: -7.014 (co-integration significant at 1%).
  - Interpretation: macro fundamentals and selected financial variables together can explain RER dynamics over the medium term; net foreign inflows (FKN_ABM) and domestic volatilities enter with statistically significant effects in the parsimonious specification.

- Overall empirical conclusions:
  - The URR financing cost (fc) has a statistically significant effect on the RER in a short-run financial-arbitrage framework, generating a more depreciated RER relative to arbitrage-only predictions.
  - Foreign exchange intervention did not show a statistically significant direct effect on the RER level in the arbitrage model, though it may affect RER volatility (requires separate study).
  - Macroeconomic fundamentals are necessary to explain medium- and long-run RER dynamics; adding financial-policy related variables improves explanatory power but requires careful model selection to avoid multicollinearity and non-cointegration.

*Source: Box 2. Capital Inflow Measures, _wp05132 - Box 2. Capital Inflow Measures*

### 2.2 yields F statistic value of 3.65, and Chi Squared statistic value of 25.57.

### _wp05132 - 2.2 yields F statistic value of 3.65, and Chi Squared statistic value of 25.57.

### Cointegration tests and model selection
- Equation 2.2 yields F statistic value of 3.65, and Chi Squared statistic value of 25.57.
- The critical value of the Engel Granger test at 10% of significance for five variables and 50 observations reported in Enders (2004) Table C is -4.348. A simple extrapolation to 13 variables yields a critical value of -8.839.
- The exclusion of five variables from 2.2 leads to specification 2.3. Such exclusion is not rejected by the data at 10 percent significance, with the F statistic value 1.71 and the Chi-Squared value 8.53 when comparing Eq. 2.2 and 2.3.
- The critical value at 1% of significance of the Engel Granger test for 8 variables and 50 observations is -6.814 while the ADF test for the residuals in equation 2.3 is -7.014.

### Error-Correction Model (ECM) for the Real Exchange Rate (based on Equation 2.3)
- Estimation method: Ordinary least squares (1991.1-2003.3). Source: Estimations by the author. Figures in parentheses correspond to P-values.
- Dependent variable: dRER (first difference of Real Exchange Rate in logs). Explanatory variables: first differences of fundamentals and financial variables; RESID(-1) is the lagged estimated residual from the cointegration equation 2.3.
- All reported estimated coefficients associated to the selected explanatory variables are statistically significant at 1 percent and the regression presents no autocorrelation.
- Reported model coefficients and associated P-values (preserved verbatim from source):
  - dRER(t) = -0.5608*dProdl (0.000) - 0.7634 *dprodl(t-1) (0.0000) - 0.1992* [dtt(t)+dtt(t-1)] (0.000)
  - - 0.1524*[dnfa (t) + dnfa (t-2)] (0.0000) - 0.8136*[dgobmy + dgobmy(t-2)] (0.0000) - 1.5064*dgobmy (t-1) (0.000)
  - - 3.715*dtari (t-2) (0.0000)
  - - 0.0006*[df_abm (t)+df_abm(t-2)+df_abm(t-3)] (0.0000)
  - +0.6065*[drdom (t-3)- drext (t-3)- dfc (t-3)+ drdom (t-4)- drext (t-4)- dfc(t-4)] (0.0000)
  - -2.0631*[drext(t)- drext (t-2)] (0.0000)
  - - 0.1022 *[dfexi (t-1)- dfexi (t-3)] (0.0001)
  - - 0.3074*RESID(-1) .(0.0000)
- Model diagnostics:
  - Adj R-Squared  0.78
  - Durbin Watson 2.06
- Interpretation: The adjustment velocity of the real exchange rate towards its equilibrium trend is rather slow, allowing for some persistence of effects.

### Policy effectiveness and dynamics of financial variables
- Financial policy variables explicitly considered: rdom, fc and Lfexint (foreign exchange intervention stock). From the results, rdom, fc and Lfexint did not play an active role in explaining the RER swings by compensating for the effects of the capital flow variables.
- Variables included only in the ECM but not in the cointegration relationship (like fc and lfexint) have only transitory effects on the trajectory of the RER and do not modify its equilibrium value.
- Specific reported dynamics and magnitudes:
  - For the URR financing cost (fc): a sustained change defined as 100 basic points:
    - Depreciated the real exchange rate by 0.6 percent after three quarters.
    - After ten quarters the total effect on the RER fell to 0.1 percent, completely disappearing afterwards.
  - The real domestic interest rate had an effect of the same magnitude but in the opposite direction of fc.
  - Foreign exchange intervention (sustained policy change defined as 10 percent of the intervention stock):
    - Depreciated the RER by 1 percent after one quarter.
    - Effect vanishes very rapidly and reversed to an appreciation of 0.5 percent after three quarters.
- Cost-effectiveness and fiscal implications:
  - The URR yielded 0.2 to 0.3 percent of GDP in revenue to the Central Bank during its application.
  - The net return of sterilizing net purchases of foreign exchange financed at the domestic real rate has been negative historically, with the historical average -400 basis points excluding the effect of changes in the real exchange rate.
  - In recent periods the cost of domestically financing NIR has fallen reflecting lower sovereign spreads and a more accommodative domestic monetary policy.

### Effects on RER of sustained changes in key variables (Table 6 summary, in percent)
- Table header: Chile : Effects on RER of Sustained Changes in Key Variables. Unit: In percent.
- Policy shock definitions (delta): DFC = 100; DFEXI = 10; DGOBMY = 1; DF_ABM = 10; DTARI = 1; DTT = 1.
- Periods and reported effects (preserved exactly):
  - Period 1:
    - DFC: 0.00
    - DFEXI: 1.02
    - DGOBMY: -0.45
    - DF_ABM: -0.80
    - DTARI: -1.45
    - DTT: -0.44
  - Period 3:
    - DFC: 0.61
    - DFEXI: -0.53
    - DGOBMY: -4.94
    - DF_ABM: -2.04
    - DTARI: -5.42
    - DTT: -0.38
  - Period 6:
    - DFC: 0.49
    - DFEXI: -0.18
    - DGOBMY: -4.80
    - DF_ABM: -1.59
    - DTARI: -4.96
    - DTT: -0.35
  - Period 10:
    - DFC: 0.11
    - DFEXI: -0.04
    - DGOBMY: -4.75
    - DF_ABM: -1.42
    - DTARI: -4.78
    - DTT: -0.33
  - Period 20:
    - DFC: 0.00
    - DFEXI: 0.00
    - DGOBMY: -4.73
    - DF_ABM: -1.37
    - DTARI: -4.73
    - DTT: -0.33
- Source note: Estimations by the author on the basis of Table 5 and Equation (2.3).

### Key empirical findings on capital flows, terms of trade, and structural policies
- Capital flows:
  - An additional US$10 billion in nkf_abm would have appreciated the RER by a cumulative 2 percent after 3 quarters, and by 1.4 percent over the long-term.
- Terms of trade:
  - A 1 percent increase in the terms of trade resulted in a real appreciation of 0.4 percent in the short-run and 0.3 percent in the longer term.
- Structural policies (relative effectiveness):
  - Import tariffs: a 1 percent reduction in the tariff index may depreciate the RER by 5 percent over the long term.
  - Government spending: a reduction of 1 percent of GDP in government spending would depreciate the RER by almost 5 percent over the long term.
  - Note: The reported magnitudes for tariffs and government spending may represent overestimation but indicate the relative importance of structural policies over financial policies for medium- and long-term RER determination.

### Conclusions and policy lessons (summary)
- An independent monetary policy is essential for a country like Chile facing idiosyncratic shocks and where inflation control is the goal of an independent central bank.
- With a pegged exchange rate within a band, an independent monetary policy is possible only by limiting international financial integration; hence capital account regulations complemented the quasi-pegged exchange rate in Chile.
- Gradual liberalization of the capital account was opportunistic and carried out in parallel with increased exchange rate flexibility; elimination of outflow restrictions occurred during inflow surges while elimination of inflow restrictions took place after sudden stops.
- The exchange rate band and capital account regulations were of limited effectiveness in stabilizing the RER because surges and sudden stops in capital flows played an important role in RER swings.
- Financial policy variables often associated with RER stabilization:
  - URR had short-run effects on the RER but could only compensate for the effect of independent monetary policy on the RER; its effect was limited and transitory.
  - Foreign exchange intervention had particularly weak and short-lived effects and is judged inconvenient to defend RER levels in the face of new surges in capital inflows.
  - Selective capital account restrictions of the type used in Chile are transitorily effective and can smooth adjustment but cannot act as a permanent policy to control the RER.
- Fiscal and trade policy recommendations:
  - A fiscal policy that stabilizes the growth rate of government spending throughout the business cycle has significant potential to moderate RER swings induced by capital flows, though might be insufficient alone during very strong surges.
  - Trade openness and tariff reductions can have meaningful RER effects, justifying more aggressive trade openness and tariff reductions during periods of strong capital inflows.

### Appendix I — Glossary of key variables (selected entries)
- Real exchange rate (LRER= log RER). An increase in the RER Index represents a real depreciation. Source: Central Bank of Chile.
- Arbitrage Real Exchange Rate: ()()]1([)]([rextrdomtREREtRERarb−−+=). Where E[RER(t+1)] represents the anticipated RER for next period. Source: Calculations by the author.
- Domestic real interest rate (rdom). Average real interest rate of the Central Bank short-term financial instrument indexed to the CPI (denominated in UF). Source: Central Bank of Chile.
- Foreign real interest rate (rext 90libo=-*ΠE+srcl). Real interest rate for loans to the Chilean sovereign. Source calculation by the author.
- Financial cost of the Unremunerated Reserve Requirement on Capital Inflows (fc). Calculated assuming a loan with the same maturity of the required deposit, one year. Parameters: URR rate for the period (30%), tax on foreign interest payments (4%). Source: Le Fort and Sanhueza, 1997.
- Foreign Exchange Intervention Stock (LFEXINT=log(FEXINT): Fexint represent the cumulative net Purchases of foreign exchange by the Central Bank, minus the stock of exchange rate linked instruments issued by the C. Bank. Source: Central Bank of Chile.
- nkf_abm: Net capital flows to main Latin-American countries (sum of annual net capital flows to Argentina, Brazil and Mexico). Source: IMF Balance of Payments Statistics.
- LTARI: Import tariff Index (Ltari=log(1+Av tariff)). Source: Budget office, Ministry of Finance (calculated by the author).
- GOBMY: Government spending with Macroeconomic Relevance (Central Government Spending excluding interest payments to the Central Bank), presented as a percent of GDP. Source: Budget Office, Ministry of Finance.

*Source: _wp05132 - 2.2 yields F statistic value of 3.65, and Chi Squared statistic value of 25.57.*

### APPENDIX I

### Appendix II. Chile: Augmented Dickey-Fuller (ADF) Unit Root Test (1990:1-2003:4)

### Unit-root test results by variable
- LRER: ADF With Constant, Without Trend = -1.95; ADF Constant and Trend = -1.42/
- dRER: ADF With Constant, Without Trend = -6.76 1/; ADF Constant and Trend = -7.19 1/
- Rdom: ADF With Constant, Without Trend = -1.89; ADF Constant and Trend = -2.22
- D(rdom: ADF With Constant, Without Trend = -7.84 1/; ADF Constant and Trend = -7.76 1/
- REXT: ADF With Constant, Without Trend = -1.48; ADF Constant and Trend = -1.02
- d(REXT: ADF With Constant, Without Trend = -5.58 1/; ADF Constant and Trend = -5.54 1/
- Fc: ADF With Constant, Without Trend = -1.53; ADF Constant and Trend = -2.03
- dFc: ADF With Constant, Without Trend = -7.11 1/; ADF Constant and Trend = -7.20 1/
- Lfexint: ADF With Constant, Without Trend = (blank) ; ADF Constant and Trend = -1.36 / -1.47  (as presented)
- Dfexi: ADF With Constant, Without Trend = -3.81 1/; ADF Constant and Trend = -4.91 1/
- Lprod: ADF With Constant, Without Trend = -2.04; ADF Constant and Trend = -1.73
- Dprod: ADF With Constant, Without Trend = -2.61 3/ 5/; ADF Constant and Trend = -3.13 5/
- Rdom-rext-fc: ADF With Constant, Without Trend = -3.13 2/ 6/; ADF Constant and Trend = -3.23 2/6/
- d(Rdom-rext-fc): ADF With Constant, Without Trend = -7.17 1/; ADF Constant and Trend = -7.17 1/
- Volrdom: ADF With Constant, Without Trend = -3.68 1/ 4/; ADF Constant and Trend = -3.71 2/ 4/
- Dvolrdom: ADF With Constant, Without Trend = -7.68 1/; ADF Constant and Trend = -7.61 1/
- Volrer: ADF With Constant, Without Trend = -5.14 1/; ADF Constant and Trend = -5.26 1/
- Dvolrer: ADF With Constant, Without Trend = -10.00 1/; ADF Constant and Trend = -5.05 1/
- Volrext: ADF With Constant, Without Trend = -6.08 1/4/; ADF Constant and Trend = -5.96 1/4
- Dvolrext: ADF With Constant, Without Trend = -6.26 1/; ADF Constant and Trend = -6.21 1/

Additional variables (presented in the same table layout)
- LCPR: ADF With Constant, Without Trend = -1.97; ADF Constant and Trend = -2.95
- dCPR: ADF With Constant, Without Trend = -5.03 1/; ADF Constant and Trend = -5.00 1/
- NKF_ABM: ADF With Constant, Without Trend = -2.03; ADF Constant and Trend = -1.24
- dNKF_ABM: ADF With Constant, Without Trend = -5.76 1/; ADF Constant and Trend = -6.42 1/
- Dumband: ADF With Constant, Without Trend = -0.64; ADF Constant and Trend = -1.96
- dDumband: ADF With Constant, Without Trend = -7.35 1/; ADF Constant and Trend = -7.34 1/
- Ltt: ADF With Constant, Without Trend = -3.61 1/; ADF Constant and Trend = -3.69 2/
- Dtt: ADF With Constant, Without Trend = -8.58 1/; ADF Constant and Trend = -8.50 1/
- Nfa: ADF With Constant, Without Trend = -1.74; ADF Constant and Trend = -2.77
- Dnfa: ADF With Constant, Without Trend = -1.65 5/; ADF Constant and Trend = -1.58 5/
- Gobmy: ADF With Constant, Without Trend = -1.69; ADF Constant and Trend = -2.26
- Dgobmy: ADF With Constant, Without Trend = -3.31 2/; ADF Constant and Trend = -3.28 3/
- Arb[ltcr]: ADF With Constant, Without Trend = -1.75; ADF Constant and Trend = -1,35
- dArb[ltcr]: ADF With Constant, Without Trend = -6.80 1/; ADF Constant and Trend = -7.25 1/
- Ltari: ADF With Constant, Without Trend = 0.57; ADF Constant and Trend = -2.57
- dtari: ADF With Constant, Without Trend = -9.16 1/; ADF Constant and Trend = -4.88 1/

### Notes on significance annotations (as presented)
- 1/ The null hypothesis of a unit root is rejected at 1% significance.
- 2/ Rejected at 5%.
- 3/ Rejected at 10%
- 4/ The null hypothesis of a unit root cannot be rejected using the ERS test at 10% significance level.
- 5/ Using the Phillips Perron the unit root hypothesis is rejected at the 1% significance level.
- 6/ The null hypothesis of a unit root cannot be rejected using the DF-GLS test at 10% significance level.

*Source: Prepared by the author on the basis of the data presented in Appendix A.*

---


_Source: https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2005/_wp05132.pdf_