Exchange rate policy has played a key role in the changing fortunes of the Philippine economy. Exchange rate rigidity, in combination with expansive and inward-looking policies, fostered economic distortions and eroded external competitiveness in the 1970s, culminating in the debt crisis of 1983. Following two devaluations and a severe further tightening of exchange and trade restrictions, a fixed exchange rate strategy adopted in the wake of the debt crisis again foundered. Confronted by a binding balance of payments constraint, the Philippines moved to a flexible exchange regime in October 1984. This switch aimed at generating a market-determined and export-driven economic recovery and was combined with broad-based policy efforts to open up the economy, reduce widespread structural rigidities, bring the fiscal house in order, and restructure the external debt.
The fruits of these efforts have become evident, as economic growth has taken off, allowing the Philippines to share in the success of the economically most dynamic region in the world. At the same time, since buoyant exports and large capital inflows have alleviated the balance of payments constraint, the options for exchange rate policy have increased.
This paper evaluates the Philippines’ experience with different exchange regimes and suggests three broad directions in which monetary policy could be developed. It argues that the shift to a flexible regime in the mid-1980s was crucial to restoring external viability, but that at times difficulties in meeting money targets and asymmetric policy reactions to exchange rate pressures have been reflected in an uneven inflation performance. Since adoption of a firm nominal anchor for monetary policy would contribute to a more effective control of inflation and thereby to better prospects for sustained growth, the merits of three monetary strategy options are reviewed: stricter adherence to a money supply rule, adoption of an exchange rate commitment, and a switch to direct inflation targeting.
II. The Experience with Different Exchange Regimes
A. A Fixed Exchange Regime: 1970–84
During the 1970s, the Philippines effectively implemented a fixed exchange rate policy vis-a-vis the U.S. dollar.1 However, the peg was not backed by appropriate supporting policies, as an inward-looking strategy was pursued that combined a tightening of trade policies with, at the end of the decade, an easing of macroeconomic policies. In particular, the average tariff rate rose under the main tariff reform program (in 1973), non-tariff barriers were progressively increased, and during 1979-82 monetary policy was accommodating (real interest rates were significantly negative) while the fiscal deficit widened by 4 1/2 percent of GNP.2
Given the insulation from international competition and a loose macroeconomic stance, domestic prices rose significantly faster than those of trading partners, causing a steady erosion of Philippine competitiveness and a stagnation in exports. In the meantime, the inconsistent policy mix—in particular the easing of financial policies and the fixed exchange rate—was sustained only by heavy recourse to foreign borrowing: a doubling of the current account deficit between 1975 and 1983 (to 8 percent of GNP) was mirrored in a doubling of the country’s external debt (to 75 percent of GNP).
Exchange rate policy responded ‘too little, too late’ to evident balance of payments pressures. While some depreciation was allowed during 1981 and 1982 (by 8 and 12 percent respectively), this proved insufficient to halt, let alone reverse, the progressive deterioration of the external position. In practice, the much-needed adjustment towards exports was postponed, the pace of debt accumulation was accelerated (at increasingly short maturities), and official reserves were dissipated. Moreover, attempts to reduce the anti-trade bias in the early 1980s by liberalizing the trade regime exacerbated the effect of the overvalued exchange rate on import demand. Although commendable in their own right, these attempts proved at odds with the further easing of fiscal policy and the reluctance to allow significant exchange rate correction.
During 1979-83, the detrimental impact of these policy imbalances was greatly compounded by a combination of adverse external shocks prompted by the second oil crisis. On the one hand, foreign exchange earnings dwindled as foreign demand decelerated and as commodity prices collapsed (weighted by the Philippines’ commodity exports, these prices fell by more than 30 percent during 1980-82). On the other hand, the need for foreign exchange surged as the oil bill doubled, rising to about one quarter of total imports, and as world interest rates reached record levels, contributing to a debt service ratio that spiraled from 20 percent in 1979 to an unsustainable 37 percent in 1982.
In the face of the impending debt crisis, the Philippines opted for an economic strategy centered around an explicit and rigid exchange rate peg to the U.S. dollar, starting in June 1983, and a severe tightening of exchange and trade restrictions. This included the imposition of a surrender requirement of current receipts to the central bank and the allocation of foreign exchange only to priority uses. In view of the precarious level of official reserves, the first policy task was to maintain essential trade and financing transactions with a minimum of liquidity. Given a lack of confidence in the peso, however, there proved to be insufficient incentives to elicit a substantial increase in the surrender of foreign exchange and thereby to avert the accumulation of external payment arrears. These developments came to a head in October 1983 when the Philippines announced a debt service moratorium.
With foreign exchange reserves depleted and external commercial financing cut off, the Philippines implemented two discretionary devaluations in October 1983 (by 27 percent) and June 1984 (29 percent) to induce a greater inflow of foreign earnings. However, since the U.S. dollar—to which the peso was pegged—experienced considerable appreciation during this period, these adjustments had only a limited impact on the Philippines’ effective exchange rate.
In the event, when the devaluations and exchange restrictions failed to relieve the acute shortage of foreign exchange and the economic crisis deepened, a drastic strategy switch had become inevitable. In October 1984 the exchange rate was allowed to float, base money targets were adopted as the new nominal anchor, and the recently imposed exchange restrictions were reversed.
B. A Flexible Exchange Regime: 1985–96
The switch to a floating exchange regime was propitious. Together with the adoption of far-reaching reforms—particularly in the trade and fiscal areas—aimed at removing distortions and opening up the highly protected economy, this set the stage for an increasingly export-driven recovery that started in the late 1980s and that has taken firm hold in recent years. However, the adoption of a flexible exchange regime has contributed more favorably to the country’s external objectives, which were of overriding concern immediately after the debt crisis, than to its internal objectives, particularly on the inflation front.
From the perspective of the external balance, the flexible regime facilitated adjustment to the major uncertainties in the balance of payments. These stemmed mainly from the weakened capital account, as the standstill on external debt payments effectively barred the Philippines from the global capital markets and initially even from customary trade financing. With a shortage of foreign financing, the exchange rate became a much needed instrument to absorb balance of payments pressures. The ensuing currency depreciation (by a nominal effective 22 percent during 1985-86) promoted expenditure-switching to domestic products, thereby lessening the need for expenditure-reducing policies, especially fiscal contraction.
The flexible regime allowed the exchange rate to settle at a competitive level that would clear the foreign exchange market and would revitalize external performance. The success of this strategy is illustrated by developments in the Philippines’ export volume, in particular relative to the growth of its export markets. Following years of steep decline, the loss of international market share was brought to a halt in 1985 when export growth again managed to keep pace with foreign demand (see Chart 1, top-left panel). In fact, since 1992 export growth has been much faster than the expansion of foreign markets, and there has been an impressive recovery in the Philippines’ export market share.
CHART 1 PHILIPPINESExternal Sector Developments, 1980-1995
Sources: IMF, International Financial Statistics; and the Philippine authorities.
1/ Excludes re-exports and semi-processed imports in electrical equipment exports.
Although a large trade deficit re-emerged a few years after the debt crisis, this mainly reflected rapid growth of imports of capital goods and semi-processed inputs. In particular, when the share of investment in GNP recovered, the share of capital goods in imports and in GNP rose at an even faster pace (Chart 1, top-right panel). Consistent with the apparent comparative advantage in light manufacturing, investment has increasingly relied on imported capital goods. As a result, in comparison with the situation before 1985, the trade deficit is now more directly related to capital goods imports and thus arguably to an increasing debt servicing capacity.
Balance of payments uncertainties have also been reduced by efforts to normalize relations with creditors. Successive debt rescheduling agreements with official creditors (covering 1985-1992), as well as a comprehensive two-step debt reduction deal with commercial creditors (implemented during 1988-94), shook off the remaining shackles of the debt crisis.3 This is evidenced by the regained access to international capital markets, notably the Eurobond market, where annual bond issues averaged US$1.2 billion in 1993-95. It is also vividly illustrated by the decline in the stripped yield spread on Philippine Brady bonds (which is the premium over U.S. Treasury bonds of comparable maturity) from about 700 points in late 1992, when the commercial debt deal was concluded to about 300 basis points in mid 1996. Significantly, a portion of the Brady bonds was exchanged in September 1996. Most tellingly, as no further exceptional balance of payments financing is anticipated, the Philippines effectively regained external viability in 1995.
Viewed from the perspective of the internal balance, however, the assessment of the flexible exchange regime is less favorable, particularly as regards inflation performance. While, with output weak, inflation was quickly reduced in the aftermath of the debt crisis (prices increased by less than 1 percent in 1986), it nonetheless reverted to 10 percent a year on average during the period 1987-96. This reflected a mix of monetary, exchange rate and intervention policies that, overall, imparted an inflationary bias.
Specifically, following the switch to a flexible regime in late 1984, the exchange rate was to be determined by market forces, central bank intervention was to aim only at preventing excessive exchange market fluctuations without countering fundamental trends, and base money targets were to provide a nominal anchor for domestic prices. In practice, policy implementation has been more complicated. Protracted periods of exchange rate stability vis-a-vis the U.S. dollar have revealed episodes of implicit exchange rate targeting. Moreover, overshooting of base money targets in several years (indeed, the initial target was met in only three of the past eight years—1992, 1995 and 1996) has suggested a reluctance at times to tighten monetary policies on the basis of money growth developments alone, especially when this tightening could have prompted significant additional exchange rate appreciation.
The use of base money targets as a nominal anchor since 1985 has thus been uneven and the exchange rate has at times been heavily managed.4 This has adversely affected price performance in several ways in the period since 1985 as a whole, although 1996 saw a successful reduction in the rate of inflation. First and foremost, the combination of a floating exchange rate and a flexible interpretation of base money targets has, for the period since 1985, made the assignment of policy instruments to economic objectives ambiguous. In the absence of a firm nominal anchor, there has been a tendency at various times to assign multiple objectives to monetary policy including, in addition to controlling money growth, stabilizing the exchange rate, accumulating foreign reserves, safeguarding export competitiveness, and keeping interest rates at a level that supports growth and limits budgetary interest costs. As evidenced by the record, the trade-offs between these multiple objectives have not always been settled in favor of the ultimate inflation goal. Second, without a credible commitment to one nominal target, the economy has effectively lacked a focal point to moor inflationary expectations.
Third, there is evidence that exchange rate and intervention policies have been implemented in an asymmetric fashion. When there has been upward pressure on the exchange rate, especially from capital inflows, nominal appreciation has mostly been resisted through sizable foreign exchange purchases that have generally increased base money (as sterilization has usually been only partial, intervention has typically had a direct impact on money market liquidity). Conversely, when there has been downward pressure, there has been a tendency to allow currency depreciation and to limit—although not exclude—foreign exchange sales that would support the currency and contract base money. This asymmetric policy reaction to exchange market pressures has an inflationary bias, as any exchange rate pressure has tended directly to raise either money growth or import prices. This asymmetry did, however, manifest itself differently in different periods. During 1987-90, it was primarily reflected in nominal exchange rate depreciation (the peso fell by 30 percent in nominal effective terms). By contrast, during 1991-95 it was more implicit, since the nominal exchange rate remained broadly constant and exchange rate policy per se no longer had an inflationary bias. But, in a context of sustained appreciation of the real exchange rate, resistance to nominal appreciation created a bias toward higher inflation.
The implementation of exchange rate and intervention policies is illustrated in Chart 2, which plots monthly changes in the exchange rate against monthly central bank intervention (the latter expressed in percent of base money). The survey starts in 1987, when reserves had recovered from their post-debt crisis lows (reaching 3.7 months of imports at end-1986), thus enabling a two-sided intervention policy, and covers the period through 1996. The focus is on the peso/U.S. dollar exchange rate, as the nominal effective exchange rate gives more diffuse observations, confirming the authorities’ tendency to aim exchange rate policy primarily at the dollar. The observations are mostly in the top-right and the bottom-left quadrant, also confirming that the authorities have indeed, in general terms, implemented an intervention policy of leaning against the wind, purchasing foreign exchange when the peso appreciates and selling when it depreciates.5
CHART 2 PHILIPPINESMonthly Exchange Rate Changes and Net Intervention, 1987-1996
A closer analysis of the exchange rate and intervention policy mix reveals the asymmetry in implementation. Specifically, in those instances where the exchange rate appreciates and the central bank purchases foreign exchange, intervention policy is forcefully used to lessen upward pressure on the exchange rate (the line depicting the average ratio between intervention and exchange rate changes in the top-right quadrant is steeply inclined towards net purchases). By contrast, in those instances where the exchange rate depreciates and the central bank sells foreign exchange, intervention is more limited (the observations in the bottom-left quadrant are closer to the origin) and the exchange rate is more readily allowed to fall (the slope of the line is considerably less steep).6 In particular, the ratio of the average size of intervention and the average size of exchange rate changes is more than five times larger in the top-right quadrant than in the bottom-left quadrant. Thus, intervention policy was pursued with much greater vigor when exchange market pressure was upward, than when it was downward.
The inflationary impact of this asymmetry may even be felt in the context of a broadly stable or slightly appreciating nominal exchange rate, as has been the case in the Philippines in recent years. Under such circumstances, price pressures would result when the underlying real exchange rate is appreciating more strongly.
C. A Trend Real Appreciation: 1987-96
While the peso was overvalued in the runup to the debt crisis, the shift to a flexible exchange regime allowed a major downward correction that bottomed out in 1987. Since then, however, the real effective exchange rate, measured by relative CPIs, has appreciated by over 30 percent.
This real appreciation is sometimes attributed to an excessively tight monetary policy, but the evidence suggests otherwise. The continued rapid liquidity growth (for broad money on average 22 percent annually during 1993-96, in the context of financial deepening), the relatively low reserve coverage (2 1/2 months of imports at end-1996), and the setbacks in some years in reducing inflation do not indicate an overly tight monetary stance. Moreover, fiscal policy has been progressively strengthened over recent years (the consolidated public sector accounts were brought from a deficit of 5.5 percent of GNP in 1990 to virtual balance in 1996), reducing the burden on monetary policy.
In searching for the causes behind the real appreciation, a major role seems to have been played by one-off factors linked to the general recovery of confidence in the peso, as economic and political uncertainties have declined and exchange restrictions have progressively been lifted. Also, insofar as the successive debt restructuring agreements have generated substantive debt and debt service reduction, there has been a one-time upward shift in the equilibrium real exchange rate—comparable to the change that takes place upon the discovery of oil reserves.
There are indications, however, that the real appreciation has been driven not only by one-off corrections towards equilibrium, but also by an ongoing upward adjustment of the equilibrium rate itself. Consistent with the Balassa-Samuelson hypothesis, a trend real appreciation can be generated by differential productivity growth rates in the tradable and nontradable sectors. Specifically, given a less elastic supply in the nontradable sector, wage increases in this sector will tend to be similar to those in the tradable sector, even if productivity gains are smaller. In this way, strong productivity growth in the tradables sector (particularly manufacturing) leads to higher prices for nontradables and an appreciation of the real exchange rate. This is one hypothesis to explain the apparent trend appreciation underlying real exchange rate movements during 1987-96, with an acceleration in 1991-96 (see Chart 3).
CHART 3 PHILIPPINESReal and Nominal Effective Exchange Rates, 1987-96
Sources: IMF: Information Notices System; and staff estimates.
1/ Based on relative consumer prices.
Evidence of the Balassa-Samuelson effect is provided by the real exchange rate based on export prices, which indicates that the loss of export competitiveness in recent years, if any, has been limited (see Chart 4).7 This supports the view that the real appreciation was primarily driven by price increases in the nontradables sector. It also suggests that, since export price competitiveness has hardly been eroded, the recent real appreciation in terms of consumer prices is a misleading indicator for the underlying strength of the external sector. In turn, this is confirmed by recent robust export performance (export growth reached a record 29 percent during 1995 and, although lower in 1996, the slowdown was less marked than observed in some other Asian countries). Moreover, the recent increase in foreign investment inflows attests to the Philippines’ strong appeal as a production base. From the viewpoint of keeping up with the regional economic momentum, however, it seems competitiveness should be kept under review—the real appreciation has been substantively larger relative to main regional competitors (Indonesia, Malaysia, and Thailand) than to trading partners (see also Chart 4).
CHART 4 PHILIPPINESReal Effective Exchange Rate Indicators, 1985-95
Sources: IMF: International Financial Statistics, Information Notice System, World Economic Outlook; and staff estimates.
1/ Indonesia, Malaysia, and Thailand, weighted by the value of total exports (excluding oil).
2/ Based on relative consumer price
3/ Export deflator calculated as a five-month moving average using world developments for petroleum, nonfuel commodities, and export unit values for manufactures, weighted by the composition of the Philippines’exports.
D. Lessons of the Past
The Philippines’ experience with fixed and then floating exchange rate regimes contains several important lessons. As regards the first period (1970-1984), the fixed exchange rate regime proved unsustainable because it was not supported by consistent macroeconomic and structural policies. In particular, financial policies were too loose, and exchange rate stability could only be maintained through broad-ranging exchange and trade restrictions, and through large external borrowings, at increasingly short maturities. In the context of tightened exchange restrictions, two large devaluations were unable to bring about the vital return of confidence in the domestic currency and the fixed regime remained unsustainable.
In the period since 1985, the adoption of a flexible regime proved crucial to promoting an external recovery, but it produced mixed results on the inflationary front as the economy was left without a nominal anchor for domestic prices. Although base money targets were to serve this purpose, in practice commitment to these targets was uneven and the exchange rate was not allowed to float freely. As well as imparting an inflationary bias, the implicit shifts between money supply and exchange rate targets sent conflicting signals to markets, affecting the credibility of monetary policy.
While it could be argued that overall macroeconomic performance is robust and that inflation is under control, there is a strong case for greater ambition in reducing inflation. Of the many reasons, three stand out. First, experience indicates that higher levels of inflation are more volatile.8 This also holds for the Philippines, where fluctuations in the inflation rate have been pronounced (annual inflation averaged 14 percent during 1970-95, but with a standard deviation of 10 percentage points). Since high inflation is therefore more likely to trigger an abrupt need for monetary tightening, it also risks undermining a sustained economic take-off. Second, to the extent that exchange rate stability remains an objective—whether implicitly or explicitly—it should be noted that higher inflation, in itself, typically leads to greater exchange rate variability. In this respect, better sustained performance on inflation would also likely reduce the risk premium in Philippine interest rates. Third, there is evidence that once inflation exceeds a “low” threshold level, as it generally has done in the Philippines, it has an increasingly adverse impact on growth.9
Against this background, consideration could be given to adopting a monetary strategy with a firm nominal anchor that would impart a greater degree of stability to domestic prices.
III. The Monetary Strategy Choice
In assessing monetary strategy options, the focus has traditionally been on the choice between pure exchange rate and money based nominal anchors.10 Theory suggests that, in the absence of binding fiscal or balance of payments constraints, the superior regime would partly depend on the nature of shocks affecting an economy. On the one hand, economies experiencing large monetary shocks should adopt an exchange rate anchor, as this would allow the shocks (in particular money demand fluctuations) to be accommodated through the balance of payments without undermining inflation objectives or output growth. On the other hand, economies suffering large real shocks would be advised to adopt money supply rules, as these shocks could then be compensated by exchange rate changes that would adjust the level of external spending on domestic goods and domestic spending on external goods. Especially when nominal wages and prices are downwardly rigid, this would help stabilize domestic output growth.
In addition to the nature of potential shocks, the preferred monetary strategy would depend in part on the structural characteristics of the economy, as these determine the ability to absorb such shocks. For example, the benefits of a money based approach would depend in part on whether there is sufficient real wage flexibility (e.g., an absence of indexation practices) to ensure that nominal exchange rate changes have a lasting impact on real wages and on output. Similarly, the advantages of an exchange rate based approach would hinge in part on whether a high degree of capital mobility allows an easy accommodation of domestic money shocks through the external financial account. Moreover, an exchange rate anchor would generally be more advantageous when an economy is more open—since exchange rate stability will be more closely linked to price stability—and has a traded goods sector that is sufficiently diversified to accommodate industry-specific shocks.
Next to these factors, the merits of each strategy would depend on the authorities’ prioritization of economic objectives, in particular regarding short-term trade-offs between output, inflation, and the balance of payments. From the above, it could be inferred that countries adopting money anchors may experience greater short-term variability in inflation (since money demand shocks would not automatically be accommodated) but less short-term variability in output and the balance of payments (since the exchange rate would be available as an adjustment instrument).11 Beyond this, an exchange rate based approach has the advantage of transparency, which may increase incentives for financial discipline. By providing a focal point for the authorities and economic agents alike, an exchange rate commitment may thereby generate favorable announcement and credibility effects, but at the same time be subject to greater testing by the markets.
While the literature has traditionally weighed money anchors against exchange rate anchors, these two options—when applied rigidly—can be considered extremes. In particular, they will likely be suboptimal for countries experiencing large shocks of both the monetary and real variety. For these countries, a third nominal anchor option could also be considered: direct inflation targeting. Under such an approach, monetary policy would be aimed at the end-objective for inflation with no clear-cut intermediate target (other than the inflation forecast). This unequivocal policy objective may help anchor inflationary expectations and reduce the sacrifice ratio of disinflation. While the monetary stance would be based on all available variables containing information on future inflation, the authorities would be granted some flexibility in tailoring their policy response according to the nature and expected duration of shocks affecting the economy. In order to promote judicious responses when short-term trade-offs are involved, policy decisions would be cast in a transparent framework with a clear target, well-defined escape clauses, and an autonomous and accountable central bank. On the downside, however, it could be argued that this strategy’s broad scope would risk leading to weaker or delayed decision-making when policy conflicts occur between output and inflation objectives.
The advantages and disadvantages in broad terms of each of these three monetary strategies are assessed below specifically for the Philippines. At the outset, however, it should be recognized that the inter-related and varying nature of the driving factors influencing the economy makes it impossible to reach a conclusive verdict for any country on the superior monetary strategy. Moreover, the differences between two of the three strategic alternatives—monetary and direct inflation targeting—may be limited in the actual operation of monetary policy. In practice, under both these strategies, the central bank is likely to base its policy decisions on a wide variety of information variables, with great although not exclusive importance being attached to money growth. Also, the optimal monetary strategy depends greatly on its credibility, which may be difficult to gauge beforehand. Steps to enhance the credibility of each approach are discussed in a subsequent section.
A. A Money Anchor
When the Philippines switched to a flexible exchange regime, the central bank adopted base money as its prime intermediate target. In practice, however, the implementation of monetary policy has been eclectic and until 1995 base money targets were often exceeded. To a certain extent this may have resulted from overambition in the setting of inflation targets. But, more fundamentally, it has reflected limited confidence in the predicting power of base money growth for near-term inflation.
The reluctance to focus monetary policy singularly on base money targets is thus linked to a lack of stability between the economy’s real and monetary variables. This can be traced to temporary and more permanent factors. A recovery in confidence in the banking system—following the collapse during the debt crisis—has prompted a temporary acceleration in financial intermediation. At the same time, an ongoing process of banking sector expansion and innovation is spurring a sustained financial deepening as banks increase their role in the allocation of national savings. Since it is very difficult to distinguish ex ante between these temporary and more permanent factors, the specification of monetary targets has become subject to wide margins of error.
From an operational perspective, the monetary authorities have faced a two-pronged problem. First, in projecting their monetary framework, they have faced an unstable demand for broad money, especially in recent years.12 Specifically, the ratio of broad money to nominal GNP has risen only modestly in some years (1993: 1 1/2 percent), but has surged in others (1994: 12 1/2 percent). This instability of money demand is illustrated in Chart 5, which plots the ratio of broad money to nominal GNP.
CHART 5 PHILIPPINES Inverse Money Velocities and Money Multiplier, 1937-96
Sources: The Philippine outhorities; and IMF staff calculations.
1/ Adjusted for seasonality.
2/ Adjusted for season nifty and reserve requirement changes.
But, second, there has also been an issue of controllability, as the relationship between base money (the variable that is directly influenced by the central bank) and broad money has been similarly unstable, owing largely to sharp swings in the currency-to-deposit ratio. Thus, while broad money has expanded at a rate similar to base money in some years (1991: 15.5 and 13.4 percent, respectively), in other years broad money expansion was about twice as fast (1994: 26.5 and 13.9 percent, respectively).13
In short, abrupt shifts in the level and the composition of money demand have undermined the stability of both money velocity and the multiplier. This greatly complicates the task of predicting a growth rate for broad or base money that would be consistent with a given inflation objective. Consequently, strict adherence to a money anchor may help achieve disinflation, but only at a considerable risk. If money targets were pitched too tight, monetary policy would unnecessarily constrain output, while if these were too loose, monetary policy would not achieve its inflation objectives.
B. An Exchange Rate Anchor
Heavy management of the exchange rate as well as official statements have revealed a strong attachment to exchange rate stability. For example, in July 1996, after the peso had been kept stable for more than six months, the central bank governor expressed the expectation that it would continue to fluctuate within a one percent band of its rate vis-a-vis the U.S. dollar through the end of the year.14 In assessing prospects for a formal exchange rate anchor, four developments stand out that would, at first glance, appear to make a fixed exchange regime more attractive than when such a regime was abandoned in the early 1980s.
First, there has been a substantial opening up of the economy. While the ratio of total trade of goods and services to GNP was reasonably stable at about 50 percent during 1980-85, it had steadily risen to almost 80 percent in 1995 and increased further in 1996 (see Chart 1, bottom-left panel).15 As a result, the exchange rate is now clearly the most important relative price in the economy, making it an effective instrument to help achieve price stability.
Second, some aspects of the balance of payments may have become less susceptible to external shocks. On the payments side, a sustained fall in the external debt burden has reduced the exposure to world interest rate changes (during 1986-96, the debt and debt service ratios have both been broadly halved to 53 percent of GNP and 13 percent of exports of goods and services, respectively). Also, the share of mineral fuels in imports has declined from about one fourth in 1980-85 to less than one-tenth now, thereby lessening the exposure to oil price fluctuations. On the receipts side, the dependence on workers’ remittances and on developments in the electronic sector is heavy, but the export base has been diversified from agriculture and mining to manufactures (see Chart 1, bottom-right panel). This has reduced the vulnerability to commodity-related terms of trade shocks. Meanwhile, the restoration of access to international capital markets and, more generally, the improvement in confidence have strengthened the capital account.
Third, the steady elimination of exchange restrictions has freed the flow of capital in and out of the country, thereby facilitating an autonomous accommodation of monetary shocks. In particular, the foreign currency deposit system has been fully liberalized and all restrictions on current account transactions have been lifted (as evidenced by the acceptance, in September 1995, of the IMF’s Article VIII obligations).
Fourth, fiscal policy has been placed on a firmer footing. Whereas the combined public sector deficit rose to above 5 percent of GNP in the early 1980s, it has been on a steady downward trend in recent years, reaching virtual balance in 1996. Although fiscal policy remains subject to uncertainties—given a small revenue base, reliance on privatization revenues, and an urgent need for infrastructural investments—there are good prospects that it would be compatible with an exchange rate peg.
In addition to these developments, the Philippine labor market continues to be marked by flexibility, as supply is plentiful and institutional arrangements have not introduced rigidities. Labor market flexibility, particularly in the context of productivity increases, reduces the need for nominal exchange rate adjustments.
However, while these factors appear to have made an exchange rate based disinflation strategy more attractive, there would be costs. The main disadvantage would be the loss of monetary policy independence to cope with the large real shocks, of both external and domestic origin, that continue to affect the Philippine economy. On the external side, the economy is vulnerable to shifts in emerging-market sentiment (such as the Mexico-related shocks in 1995), which typically have a contagion effect on the Philippines, and to a potential slowdown in remittances by overseas workers, which have been crucial to finance the large trade gap. On the domestic side, the country is prone to natural calamities (including, in the past decade, major earthquakes, volcanic eruptions, and cyclones) and to inevitable political uncertainties of a regained democracy. Some shocks may have only passing effects and could be addressed by a temporary tightening of financial policies. But if the external outlook changed durably, a timely adjustment of any exchange rate peg would need to be considered.
The effectiveness of an exchange rate anchor in bringing down inflation also depends on whether the real exchange rate continues to appreciate (reflecting the rapid structural and technological improvements). If so, adoption of a nominal exchange anchor could in principle be consistent with some progress on disinflation, but price increases would still tend to be higher than in trading partners. Thus, pegging the nominal rate would elicit higher inflation than otherwise. Looking ahead, however, the pace of real appreciation may be slower than in the recent past, to the extent that the one-off factors have run their course. The ongoing trade liberalization and the strength of regional competitors also suggest that further sizable real appreciation may begin to challenge the favorable external outlook.
C. Inflation Targeting
Given vulnerability to both large monetary and large (and ongoing) real shocks, the superior monetary strategy is not clear-cut. In this case, strict adherence to either a monetary or exchange rate anchor could have high costs in terms of output growth or inflation. A third approach, whereby the authorities are left more room for maneuver and monetary policy is targeted directly at inflation, may be better suited to the economy. By embedding the commitment to more stable prices into a transparent and consistent policy framework, this would enhance the effectiveness of monetary policy, while maintaining some flexibility to cope with unforeseen developments.
To a limited degree, the authorities have already moved in the direction of inflation targeting. Since September 1995, the central bank’s base money targets have been automatically adjusted upward for higher-than-expected accumulation of net foreign assets, subject to inflation being on track. At the same time, if inflation exceeds the programmed path three months in a row, the base money program is revisited to see whether the monthly targets themselves need downward adjustment. However, to be more effective as a nominal anchor, a comprehensive and forward-looking framework for inflation targeting would need to be established. This would involve adapting the institutional framework in several key areas.
First, under such a regime, the inflation target would be made the central bank’s prime statutory objective, granting clear priority to this target above other objectives. Second, the central bank would increase the transparency of its operations by explicitly justifying policy changes in the context of achieving the inflation target. This would involve providing adequate background information on recent developments in inflation and on the derivation of the target. For some information variables (such as money growth, and perhaps the exchange rate) “comfort levels” consistent with the targeted inflation path could also be published. These steps would provide credibility to the new policy by allowing the public to monitor progress toward the target. Third, a greater degree of accountability could be established, for example through periodic reporting to the Congress on performance with respect to the inflation target.
Several difficulties of inflation targeting would also need to be addressed. For example, the central bank may consider inflation to be too far outside its direct control. Given the unpredictable price impact of monetary and real shocks (including, for example, the rice shortage in late 1995), and the difficulty of accurately quantifying the transmission of monetary policy on inflation during periods of rapid structural change, it may feel insufficiently confident to commit itself firmly to a specific inflation target. Also, with this inherent unpredictability, an inflation target could generate too much instrument volatility (in particular of interest rates) relative to the benefits of increased price stability. These issues could be addressed to some extent, however, by defining the inflation target as a relatively wide band (of, say, three percent), by specifying limited escape clauses, and by excluding the impact of government measures (such as VAT changes) or certain price developments (such as oil price changes) from the inflation target.
IV. Establishing Credibility
The benefits of each monetary strategy will crucially depend on the credibility of the new policy resolve, as this will determine the transitional output costs of lowering the inflation rate. Generally, any strategy aimed at lasting disinflation will necessitate some short-term policy tightening and a sustained cautious stance thereafter. The brunt of this adjustment would fall on monetary policy, as fiscal, wage and structural policies are not main contributors to the inflationary process—specifically, the consolidated budget is balanced, wage negotiations are decentralized and limited in scope, and structural reforms are opening up the economy (e.g., the average tariff rate fell from 28 percent in 1990 to 16 percent in 1995, and is scheduled to decline further to 9 percent in 2000). Thus, monetary policy could no longer pursue multiple objectives and domestic interest rate policy would need to be subordinated to the nominal anchor objective, even if this may conflict with other short-term objectives such as for growth or fiscal interest expenditures.
Of the three monetary strategy alternatives, policy credibility may be more difficult to establish for a money based approach, as money targets relate to abstract aggregates, and as experience in some recent years with such targets has been mixed. Nonetheless, confidence in these targets could be buttressed by clearly signaling their derivation, and by justifying any changes in the monetary stance on the basis of developments relative to the targets.
In the case of an exchange rate peg, credibility would mainly depend on two preconditions. First, prior to pegging, inflation should be put firmly on a downward path toward levels prevailing in trading partner countries. This would signal the will to pursue a monetary policy that is sufficiently tight to sustain the peg. Second, official reserves would need to be substantially increased from their current level in order to increase the capacity to deal with the inevitable market testing of the authorities’ resolve. This should be less difficult, however, in the context of continued large capital inflows.
The credibility gains of an inflation targeting approach would be largely determined by the transparency of the institutional set-up supporting the new anchor and by the accountability of the central bank in delivering on its target. An explicit government commitment to the central bank’s inflation objective would be indispensable to buttress these gains.
Finally, it deserves emphasis that a monetary strategy with an explicit nominal anchor should only be adopted if it is genuinely supported by a firm policy commitment to greater price stability, as the costs of failure would be high. In particular, if the newly adopted anchor were to slip, policy credibility would be crucially damaged, risk premia in interest rates would rise, foreign investors would be deterred, and inflationary expectations could become more entrenched.
The switch from a fixed to a flexible exchange regime has served the Philippines well. The floating exchange rate policy was a key element in the adjustment strategy that restored external viability and laid the foundations for sustained growth. However, inflation performance has been mixed, and, despite good progress in 1996, average inflation is still higher than desirable in the medium term. The evidence suggests that inflation performance is linked to a difficulty in some years to achieve monetary targets, and to an asymmetric policy whereby nominal appreciation has been resisted more vigorously than depreciation. In light of the benefits of greater price stability to a sustained economic take-off, policy makers could consider adopting a monetary strategy with an explicit nominal anchor that would provide an unambiguous guidepost for successful disinflation. Monetary policy would then need to give unequivocal precedence to this nominal anchor objective, and other policies would need to provide consistent support.
Of the different monetary strategy alternatives, a pure money supply rule may have the larger drawbacks. With sizable monetary shocks caused by an unpredictable process of financial deepening, adherence to money targets may yield a less certain inflation outcome and may place an excessive adjustment burden on the real economy. Furthermore, given that this approach relates to abstract aggregates and that it has been pursued with mixed success over the years, the credibility gap may be more difficult to bridge.
The strengthened balance of payments and the opening up of the economy suggest that an exchange rate anchor could be an effective instrument to limit inflation, while allowing an automatic accommodation of monetary shocks. However, a formal exchange rate commitment would limit monetary policy independence, thereby reducing the ability to cope with real shocks that repeatedly affect the Philippine economy. Also, to the extent that the equilibrium real exchange rate continues to appreciate, success with disinflation may only be partial. In any event, as this approach would be most rigorously tested by the markets, it would be advisable first to bolster confidence by bringing inflation firmly downward and by substantially augmenting international reserves. Moreover, while credibility considerations would limit the scope for exchange rate changes once the peg is adopted, the authorities would have to remain alert to the need for adjustment if there were a fundamental change in the economy’s external outlook.
In the context of large shocks of both the monetary and real variety, an inflation targeting approach would provide a clear nominal anchor while still allowing room for discretionary policy adjustments. Such a strategy would increase policy flexibility in dealing with such shocks; at the same time the central bank’s unequivocal objective may help anchor inflationary expectations and thereby reduce the output costs of disinflation. Substantive institutional changes would need to be implemented, however, to enhance the transparency and accountability of monetary policy, and thereby to impart credibility to the specified targets.
Aghevli, Bijan, MohsinKhan and PeterMontiel,“Exchange Rate Policy Issues in Developing Countries: Some Analytical Issues”, Occasional Paper No. 78. International Monetary Fund, Washington D.C., March1991
Bredenkamp, Hugh,“Capital Inflows: Are They a “Problem”?”, in: Philippines Background Paper, IMF Staff Country Report 95/114, International Monetary Fund, Washington D.C., December1995
Dohner, Robert, and PoncianoIntal,“The Marcos Legacy: Economic Policy and Foreign Debt in the Philippines”, in: J.Sachs (ed.), Developing Country Debt and Economic Performance. Chicago, 1989.
Fischer, Stanley,“Exchange Rate versus Money Targets in Disinflation,” in: S.Fischer (ed.), Indexing, Inflation and Economic Policy. MIT Press, Cambridge Mass., 1986
Goldsbrough, David, and IqbalZaidi,“Monetary Policy in the Philippines during Periods of Financial Crisis and Changes in Exchange Rate Regime,”Working Paper WP/89/98. International Monetary Fund, Washington D.C., December1989
Ghosh, Atish, Anne-MarieGuide, JonathanOstry and HolgerWolf,“Does the Nominal Exchange Regime Matter?”, Working Paper WP/95/121. International Monetary Fund, Washington D.C., November1995
Krugman, Paul, JamesAim, SusanCollins, and EliRemolona,“Transforming the Philippine Economy”, UNDP/NEDA Manila, 1992
Guitián, Manuel,“The Choice of an Exchange Regime”, in: Barth, Richard and Chorng-HueyWong (eds.), Approaches to Exchange Rate Policy. IMF Institute, International Monetary Fund, 1994
Poole, William,“Optimal Choice of Monetary Policy Instruments in a Simple Stochastic Macro Model”, Quarterly Journal of Economics. Vol. 84, pp. 197-216, May1970
Sarel, Michael,“Nonlinear Effects of Inflation on Economic Growth”, Staff Papers, International Monetary Fund, Vol. 43No.1, pp. 199-215, March1996
At the time of writing the author was in the Policy Development and Review Department; he is now with the Monetary and Economic Policy Department of the Netherlands Bank. The author gratefully acknowledges comments by Hugh Bredenkamp, Joshua Felman, John Hicklin and Mark Stone.
Although a managed float was formally introduced in 1970, for all intents and purposes exchange rate policy maintained the objective of stability against the dollar: changes in this bilateral exchange rate averaged less than 2 percent per year during 1971-1980. Prior to 1970, the peso had been fixed to the dollar, with only one discretionary devaluation in 1963.
Dohner and Intal (1989) elaborate on these policies and their effects.
Given a stronger-than-expected external position, a 1994 Paris Club rescheduling agreement was canceled at the request of the Philippine authorities.
Goldsbrough and Zaidi (1989) apply a monetary model of exchange market pressure to the Philippines and show that, even after the switch to a floating regime, changes in foreign reserves still absorbed most exchange market pressure.
The outliers in the first and fourth quadrant relate to observations where intervention was generally prompted by exchange rate changes in earlier months, or where a precarious level of reserves necessitated purchases notwithstanding a weakening exchange rate.
Bivariate regression analysis of monthly exchange rate changes and net intervention in the period 1987-1996 confirms this asymmetry.
The export deflator was calculated using developments in world prices for nonfuel commodities, petroleum, and export unit values for manufactures, weighted by the composition of the Philippines’ exports.
The relationship between the level and variability of inflation is long-established; see, for example, Taylor (1981).
In a recent study of inflation in 87 countries (including the Philippines and most ASEAN countries) during 1970-90, Sarel (1996) finds a structural break at a rate of 8 percent. Above this level, inflation has a very powerful negative effect on growth. By comparison, Philippine inflation was above this level in 17 of the past 20 years.
In a seminal article by Poole (1970), the analysis is cast in terms of a money supply or interest rate rule. In the presence of full capital mobility, the latter is comparable to an exchange rate rule with interest rates determined exogenously. More recent studies include Aghevli et at (1991), Fischer (1986), and Guitián (1994).
In a comprehensive study of 136 countries over a 30-year period, Ghosh et al. (1995) find that fixed exchange regimes are indeed associated with lower inflation, but more variable output than flexible regimes.
Stone (1995) presents regression analysis that suggests an upward shift in money demand, starting in 1991 and accelerating in late-1994.
Base money data are adjusted for changes in reserve requirements.
As reported in “Peso seen to remain stable”, The Manila Chronicle, July 4, 1996.
It is sometimes argued (e.g., in Krugman et al., 1992) that the rapid trade growth and export diversification mainly reflects goods that are only assembled in the Philippines, with low domestic value added. However, after adjusting for re-exports and semi-processed imports in electrical equipment trade, the ratio of trade to GNP still shows an impressive rise from 40 percent in the mid-1980s to almost 60 percent in 1995.