作者 | Agustín Carstens，国际清算银行总经理
Exchange rates and monetary policy frameworks in emerging market economies
Lecture by Mr Agustín Carstens, General Manager of the BIS, at the London School of Economics, London, 2 May 2019.
Good afternoon. It is a great pleasure to be here today, at the LSE. I am particularly happy to share the floor with Andrés Velasco, for whom I have great admiration as an academic and as a policymaker. My interaction with Andrés was remarkably fruitful back in 2008 and 2009, when we were finance ministers of our respective countries - both well run, open economies in Latin America. At the time, we commiserated with each other about massive spillovers from crisis-stricken advanced economies (AEs) during the Great Financial Crisis (GFC). Regretfully, emerging market economies (EMEs) still face this type of problem, as I hope to illustrate during this lecture.
Monetary policy frameworks are being actively discussed again as central banks around the world attempt to chart their course in the unfamiliar post-crisis economic waters. The Federal Reserve has embarked on a review of its monetary policy framework, and other AE central banks are acutely aware of the need to adapt their own frameworks so that they are fit to meet new challenges.
EME central banks find themselves in similar circumstances. Even though EMEs were not at the epicentre of the GFC, they suffered its shockwaves; and since then, they have been dealing with the spillovers of the needed remedial policies that AEs have been implementing. The unconventional monetary policies that AE central banks have embarked on in the past 10 years have created unprecedented amounts of liquidity in the international financial system, part of which has been channelled to practically all EMEs - induced by the mantra of "search for yield". Such flows have taken the form of portfolio investments, including corporate and sovereign debt foreign direct investment, and bank lending, but also trade financing.
These flows act through different channels and therefore have different consequences. But the common thread is that they affect the exchange rate, making it more volatile and subject to large swings as the capacity of EMEs to handle massive capital flows and huge stocks of foreign resources is tested. This has been a phenomenon during the past several years of inordinately accommodative monetary policies in AEs, and will continue into the future as accommodation persists but also when it is reversed. This situation has challenged, and will continue to challenge, the monetary policy frameworks adopted by EMEs during the last two decades. The main reason is that the behaviour of the exchange rate can fundamentally affect the dynamics of inflation and the capacity of monetary policy to produce the expected results.
Thus, in the course of my lecture today, I will concentrate on how EME central banks have fared in the post-crisis era, and what challenges they face in shaping the contours of a monetary policy framework that can serve them well in the coming years.
In some respects, EMEs have had a head start in charting their monetary policy strategies as they revamped their policy frameworks in the wake of the severe financial crises that hit them in the 1990s. The most notable change was the exchange rate regime. There was a general move away from fixed or heavily managed exchange rate arrangements towards flexible exchange rates. The vast majority of EME central banks today operate under an explicit inflation targeting regime with flexible exchange rates, while only a few use an explicit exchange rate anchor. This framework has taken deep root and has served central banks well. Last summer, EMEs suffered from financial turbulence. By and large, however, most EMEs came out of this turbulence unscathed. I believe their resilience owes much to the flexibility built into their monetary policy frameworks, together with stronger fundamentals, including more prudent fiscal policies and enhanced financial regulation and supervision.
But reality reflects the importance of the exchange rate. Irrespective of the official labelling, EME central banks have, in practice, attached a significant weight to the exchange rate in the conduct of their monetary policy, as reflected for instance in the evolution of FX reserves in EMEs over the past two decades. "Benign neglect of the exchange rate" has been a dictum honoured more in its breach than in its observance as a guide for monetary policy. As a consequence, many EMEs have a quasi-managed floating exchange rate regime where central banks lean against swings in the exchange rate, both on the way up and on the way down. I shall argue that this approach is one where the practice outruns the theory, and it is arguably the theory that needs to catch up. That said, the enduring challenge for EME central banks is to design their monetary policy frameworks in a way that incorporates in a rigorous way the precise role of the exchange rate for their domestic economic outcomes. The BIS is expending a great deal of analytical effort in this direction, and I would like to share a few of the key findings today.
Challenges posed by swings in exchange rates
Why do EME central banks care so much about the exchange rate? At a basic level, it's because the exchange rate is a core determinant of the nominal anchor in a small open economy. It goes to the heart of what central banks are meant to do: preserve the value of money. Large swings in the exchange rate, and especially large depreciations, can destabilise prices, and can do so in non-linear and even discontinuous ways. Monetary stability, usually accompanied by at least moderate exchange rate stability in the medium term, is the cornerstone of orderly economic activity. And for monetary stability to prevail, it is vital to maintain the hard-won credibility of the monetary framework.
Exchange rates and the nominal anchor
Exchange rates impact domestic inflation through their effect on the price of tradables. However, the ultimate effect of exchange rate changes on the broader price level depends crucially on the characteristics of the domestic inflation process, and specifically on the propagation of the initial impact through "second-round" effects. In particular, the initial impact on the price of tradables can feed into the non-tradables sector and to the general price level. The bigger the initial shock, the greater the chance of inflation expectations being set adrift from the central bank's target. The unmooring of inflation has occurred throughout history. It is non-linear and sometimes discontinuous. It's worth remembering that a large depreciation and economic downturn are often followed by a sharp rise in inflation, despite the slowdown in economic activity. The extent of such second-round effects depends, in turn, on how well inflation expectations are anchored to the central bank's target. The less well anchored inflation expectations are, the more pronounced second-round effects will be.
To anchor inflation expectations in the face of destabilising domestic currency depreciations, central banks tend to tighten their monetary policy stance, usually by adjusting a short-term reference interest rate. In technical terms, this type of reaction function would be equivalent to having a Taylor rule that would include exchange rate considerations, reflecting precisely the influence of the exchange rate on inflation dynamics.
In circumstances of extreme market volatility though, traditional monetary policy adjustment through short-term interest rates might not be enough to preserve the anchoring of inflation expectations. In such cases, it would be appropriate to use other instruments, such as foreign exchange interventions and/or macroprudential policies. Given the possibility of massive stock adjustments in capital markets as part of the settling process for the unprecedented liquidity in the international financial system, the need to stabilise using multiple instruments might very well prove to be the norm rather than the exception. As a matter of fact, various episodes in multiple countries have been providing evidence in this direction. Here is where the theory needs to catch up with the reality. Having said all this, it is true that exchange rate pass-through to inflation in many EMEs has decreased considerably over the past two decades, although it often remains larger than in AEs. The decline in exchange rate pass-through to prices is one of the notable achievements of the inflation targeting regimes put in place since the 1990s. There are, however, major regional differences, with pass-through in Latin America being considerably higher than in emerging Asia, reflecting differences in the significance of second-round effects.1 To some extent, the hard-won gains in reducing pass-through may well reflect the monetary policy practice of limiting swings in exchange rates. In this respect, I believe that the departures of monetary policy practice from the textbook prescriptions have been important in consolidating the gains.
So, while exchange rate pass-through has declined, it would be too complacent of us to assume that inflation has been vanquished. In fact, many of the vulnerabilities that lie just below the surface can be exposed at times of economic and financial stress.
Before we delve deeper into the discussion about the appropriate role of the exchange rate in EME monetary frameworks, it is important to understand in greater detail the main avenues through which exchange rate fluctuations work across different layers of the economy.
Exchange rates and export volumes
Let me start with exports. Generally, a depreciation of the currency improves international competitiveness and boosts economic activity. However, this benign picture might be missing some key elements. In particular, as a consequence of widespread dollar invoicing,2 there is a financial channel of exchange rates through global trade financing that weakens the traditional trade channel.
The role of trade finance has increased as global value chains (GVCs) have lengthened, requiring greater financial resources to underpin their expansion. One summary measure of the prevalence of GVC activity is the ratio of global trade to global GDP. Since trade measures gross output while GDP is about value added, the ratio of trade to GDP is a useful proxy for GVC activity. As a stronger US dollar is usually associated with tighter credit conditions for EMEs, this financial dimension weakens the expansionary effect of currency depreciation on a country's exports. In the extreme, currency depreciation could even have a contractionary effect on exports if GVCs are curtailed due to tighter credit conditions.3 One indication of the relevance of such mechanisms is that the ratio of global trade to global GDP is negatively linked to the strength of the US dollar. At the BIS, we are currently looking into how much the recent downturn in manufacturing and trade can be attributed to this channel.
Exchange rates and domestic economic activity and financial conditions
In EMEs, the exchange rate also affects domestic economic activity through financial conditions, further complicating the central bank's task. EMEs' exposure to financial channels of the exchange rate arises from two key features of their financial structure: (i) EME borrowers, especially corporates, rely heavily on foreign currency borrowing; and (ii) foreign investors' large holdings of EME local currency sovereign debt. Through both channels, exchange rate appreciation tends to loosen domestic financial conditions, exerting an expansionary effect on domestic economic activity. Since monetary policy works through financial markets, central banks understandably care about exchange rates in the context of their domestic demand conditions. More broadly, looser financial conditions lead to the build-up of financial vulnerabilities that may pose risks to price stability over longer horizons.
Over the past two decades, the stock of foreign currency debt of EMEs has surged, driven in particular by the non-financial corporate sector, and in many cases it has not been matched with FX assets and revenues, giving rise to currency mismatches.4 Corporates have incurred long-term debt in foreign currencies, often to finance long-term real investment, or simply to accumulate financial assets, such as loans to other, less creditworthy companies, often in domestic currency.
Such currency mismatches on borrower balance sheets make financial conditions a function of the exchange rate. For instance, an appreciation of the domestic exchange rate against the funding currency reduces debt servicing costs and debt burdens, lowering EME borrowers' credit risk, attracting more capital inflows and loosening financial conditions. These mechanisms work in reverse when the currency depreciates, but are then potentially amplified through higher foreign currency debt burdens accumulated in the appreciation phase.
Even in the absence of currency mismatches on borrower balance sheets, exchange rate swings influence EME domestic financial conditions. EME sovereigns have increasingly relied on local currency debt issuance, facilitated by the rapid development of local currency bond markets. However, the ability to borrow in domestic currency has not alleviated the exposure to exchange rate swings. It has rather shifted it to a different form, as a large share of EME local currency sovereign bonds is held by foreign investors, reflecting the search for yield by asset managers.5
Since foreign investors are subject to risk constraints in global currencies, the currency risk merely migrates from borrowers' to lenders' balance sheets. Currency and rollover risks on the borrower side have been replaced by duration and currency risk on the lender side. Exchange rate moves then tend to amplify investors' gains and losses, so that exchange rate fluctuations amplify portfolio flows. Exchange rate appreciation increases credit supply from foreign investors, pushing down bond yields. The same mechanism plays out in reverse when the exchange rate depreciates.6
Since central banks care about domestic activity, exchange rates matter because they influence long-term interest rates above and beyond the textbook transmission channels of monetary policy. A strongly appreciating domestic currency is associated with compressed term premia, while a sharply depreciating domestic currency is associated with widening term premia. Even for a central bank that doesn't worry about financial stability, these swings in long-term rates matter for demand conditions. When financial stability considerations are factored in, the impact of exchange rates is greater still. External borrowing - from both banks and capital markets - and domestic borrowing interact. There is ample evidence that external borrowing increases relative to domestic borrowing during credit booms.7 And that strong credit expansion coupled with strong exchange rate appreciations has preceded financial crisis. In this way, global financial conditions and domestic financial cycles reinforce each other.8
The strong presence of global investors in EME markets further implies that financial shocks, such as a change in AEs' monetary policy or a change in investor sentiment, can trigger portfolio flows that are so large that they can become a driver of the exchange rate. The exchange rate may hence increasingly act as a transmitter and amplifier of financial shocks, rather than as an absorber of real shocks.
Taking account of exchange rates in monetary policy conduct
The link between exchange rates and domestic financial conditions and the weakening of the traditional trade channel have important implications for monetary policy. A depreciation of the domestic currency would push up inflation through exchange rate pass-through, but would have little effect on domestic output through traditional trade channels, at least in the near term. Through financial channels, exchange rate depreciation would further lead to a tightening of financial conditions across the board, exerting a contractionary effect on the domestic economy. As a consequence, the central bank may face the dilemma of rising inflation and a weak real economy when the exchange rate depreciates. A short-term trade-off between inflation and output stabilisation may arise which would complicate the conduct and the communication of monetary policy.
Moreover, thanks to the presence of powerful financial channels, exchange rate fluctuations push inflation and debt in opposite directions, potentially raising an intertemporal trade-off for the central bank in the pursuit of price stability. This trade-off is best described in the context of an appreciating currency. Exchange rate appreciation tends to push down inflation, but fuels the accumulation of debt by loosening financial conditions, raising vulnerabilities over the medium run. Since financial stability risks also imply risks to longer-run price stability, this raises, for central banks, an intertemporal trade-off between short-term and medium-term for both output and price stability.
In the face of these difficult trade-offs, traditional monetary policy response through short-term interest rates runs the risk of falling short of what is needed. Thus, EME central banks have responded to these trade-offs and challenges through the activation of additional supporting policy instruments aimed at mitigating exchange rate swings and their macro-financial repercussions.
Intervention in foreign exchange markets is a tool that can help shape more favourable trade-offs from exchange rate swings. EME central banks have extensively relied on this instrument over the past two decades, as reflected in a significant increase in their FX reserves. And, validating this practice, there is empirical evidence indicating that sterilised FX purchases in EMEs exert a statistically and economically significant depreciating effect on exchange rates, at least temporarily.9
FX intervention helps address the challenges from exchange rate swings in two main ways.
First, through its effect on the exchange rate, it can directly counteract exchange rate swings that would have undesired effects on the inflation rate and on the real economy. In doing so, it takes some of the burden off conventional monetary policy conducted through interest rates and adds a degree of freedom for monetary policy.
Second, the accumulation of reserves has macroprudential-like features. For one, it provides self-insurance against large devaluations in the future and, in doing so, it represents an integral element of the global financial safety net. Indeed, there are indications that FX reserve buffers helped mitigate the impact of recent episodes of global financial stress on EME exchange rates, including during the GFC. For this purpose, the reserve accumulation itself does not even need to have an influence on the exchange rate. In fact, when building up reserves with this objective in mind, some central banks seek to have as little impact as possible on the external value of the currency. This objective was quite common after the experience of the currency crises of the 1990s.
At the same time, sterilised FX intervention counteracts the mutually reinforcing feedback loop between exchange rate appreciation and capital inflows that fuels domestic credit creation.10 In other words, FX reserve buffers do not just clean, once capital flows reverse and stress arises, but their accumulation also "leans" against the build-up of financial imbalances in the first place, reducing the risk, or at least the amplitude, of a possible reversal.
The analogy with macroprudential instruments such as minimum loan-to-value ratios or the countercyclical capital buffer is obvious.11 Indeed, there are good arguments for including both macroprudential measures and FX intervention as part of an integrated macro-financial stability framework.12
However, holding FX reserves is costly, in particular in periods of very low interest rates in reserve currencies, as is the case currently, and especially in countries with high domestic interest rates. The extent of precautionary reserve accumulation and of the use of intervention as a stabilisation tool will depend on the assessment of the net benefits, taking into account all perceived benefits and costs, which will vary across countries and over time.
Thus, while national foreign reserves are an important element of the global financial safety net, they are quite costly and, also for that reason, will always be limited. In times of large stock adjustment by global investors, outsize capital outflows can overstretch the central bank's FX reserve buffer. The mere possibility of such a scenario already makes speculative runs more likely. In order to mitigate this risk, sound policy frameworks and FX reserve buffers at the country level must be complemented by adequately equipped global lender of last resort facilities at the IMF.
Targeted policy measures
Besides FX intervention, EME central banks have also resorted to traditional macroprudential tools and (although less frequently) to non-orthodox balance sheet policies to deal with the challenges from exchange rate swings.
The targeted nature of macroprudential tools has the advantage of reduced collateral damage if imbalances and vulnerabilities are concentrated in a specific segment of the financial sector. At the same time, this makes them susceptible to circumvention, which might reduce their effectiveness. Overall, the experiences of the past two decades suggests that such targeted measures can be effective and can help alleviate the trade-offs faced by monetary policy in the face of exchange rate swings. They can enhance the resilience of the economy and mitigate the build-up of vulnerabilities. That said, our knowledge about the effectiveness of macroprudential tools is still somewhat limited, and more research is needed to enhance our understanding in this respect. Moreover, most macroprudential instruments are bank-oriented and hence unsuitable to deal with imbalances and vulnerabilities that may arise in capital markets.
In such situations, EME central banks may instead resort to non-orthodox balance sheet policies. One such policy that has been implemented by several EME central banks is to offer foreign exchange protection to investors without affecting the level of international reserves. This is achieved by auctioning non-deliverable forwards (NDFs), thereby compensating the holders for domestic currency depreciation. The central bank has a natural hedge for this derivative exposure, precisely through its international reserves. Thus, the offering of exchange note protection through NDFs is equivalent to adjusting the currency denomination of the central bank's balance sheet.
Central banks could also prevent capital flows, and thus exchange rate pressures, by facilitating the duration adjustment of portfolios in times of stress. Specifically, when a large amount of foreign capital has been channelled to long-duration public debt and threatens to flow out quickly, the central bank can use its balance sheet to stabilise markets by offering shorter-duration instruments. For example, in 2013 the Bank of Mexico swapped long-term securities for short-term securities via auctions. The reason was that long-term instruments were not in the hands of strong investors and there was market demand for short-term securities in order to address stock adjustment by global investors. This policy stabilised conditions in peso-denominated bond markets.
EME monetary policy frameworks: which way forward?
What do all these considerations mean for monetary policy frameworks in EMEs? The textbook version of the inflation targeting framework, which prescribes pursuing inflation stability with floating exchange rates through adjustments of a short-term interest rate, is obviously too narrow for EME central banks. In particular, the financial channel of the exchange rate gives rise to difficult trade-offs for monetary policy, while at the same time complicating the conduct of monetary policy by weakening its transmission.
EME central banks have risen to this challenge through their innovative use of additional policy instruments. They have turned to FX intervention to deal directly with the financial channel or insure against undesired exchange rate swings, and to other non-orthodox balance sheet policies as well as macroprudential tools to deal with specific imbalances or vulnerabilities in a targeted way. EME central banks' policy reaction function in the pursuit of price stability can therefore be described as a multi-instrument reaction function responding to multiple-indicator variables, including the exchange rate. Interest rates, FX intervention and targeted measures can be seen as forming the corners of a policy triangle, which authorities rely on in the pursuit of price stability. The calibration of the multi-dimensional instrument strategy will depend on country-specific characteristics and the underlying factors driving exchange rate and macro-financial dynamics.
Going forward, EME central banks will need to further develop their toolbox for dealing with the challenges of exchange rate and capital flow gyrations. In particular, in a time of large and internationally mobile stocks of financial capital and low interest rates, search for yield and risk-taking acquire greater prominence in global capital flows and can expose EMEs to disruptive stock adjustments by global investors. In order to deal with this challenge, EME central banks may need to consider whether to further develop non-orthodox balance sheet policies to deal with stock adjustments, such as asset purchases or asset swaps similar in nature to the measures launched by major AE central banks to bring down long-term interest rates once short rates hit the lower bound.
EME central banks also need to contend with how to address the intertemporal trade-offs that exchange rate fluctuations have given rise to. As discussed before, through the presence of powerful financial channels, the implications of exchange rate gyrations for inflation in the short and medium run can be very different. For instance, exchange rate appreciation can be disinflationary in the short term, but may foster the build-up of financial imbalances, raising the risk of large capital outflows and exchange rate depreciation in the future, thus creating risks for medium-term price stability. In order to address this challenge, besides drawing on an extended set of policy instruments, EME central banks need to incorporate sufficient flexibility and sufficiently long horizons in the interpretation of their price stability mandates. That way, the longer-run risks to price stability posed by exchange rate-driven financial imbalances could be incorporated into the decision-making process, and short-term policy activism be avoided.
Finally, EME policymakers need to consider measures to reduce the vulnerability of their countries to exchange rate gyrations. This could be achieved in particular through structural reforms boosting growth potential and thus enhancing debt sustainability. And, as always, there is just no substitute for strong economic fundamentals.
Thank you very much.
责编 胡晓涛 金天
海外之声 | BIS总裁：货币与支付系统的未来
海外之声 | 欧央行执行委员会成员：欧洲与欧洲金融市场的竞争力
海外之声 | 国际清算银行金融稳定研究所主席谈监管在市场一体化中的作用
海外之声 | 美联储监管副主席：逆周期缓冲资本框架设计
海外之声 | 拉加德：一带一路——实现金融互联互通的关键渠道
We only share the most valuable financial insights.