An overly restrictive monetary policy

9In response to the 2008 financial crisis, the main monetary authorities (the ECB, the Fed, the Bank of Japan, and the Bank of England) reduced their policy rates to almost zero. [5] Beyond interest rates, they deployed three types of instrument in an attempt to further ease monetary policy and support economic activity:

• credit easing, which involves extending commercial banks refinancing schemes (extension of the list of eligible collateral, extension of the maturity of the loans, reduction in the haircuts applied to collateral, etc.). This easing does not necessarily involve changing the amount of money put into circulation by the central bank if the latter simultaneously releases other assets on its balance sheet to offset the volume of loans granted under these more flexible conditions (so-called ‘sterilisation’ operations);
• an increase in the size of their balance sheets by means of direct purchase of assets (quantitative easing) or by means of non-‘sterilised’ credit easing;
• forward guidance: the central bank commits to implementing a series of expansionary policies over a relatively long period of time which may or may not refer to explicit employment or inflation thresholds.

10The Fed, the ECB and the Bank of England have implemented a variety of credit easing policies since 2008. All of them have also considerably increased the size of their balance sheets, although the ECB began at a later stage than the others (cf. Figure 1). Starting in 2012, however, the ECB set itself apart from the Fed, the Bank of England and the Bank of Japan with a significant reduction in its overall balance sheet.

1

Total assets of four central banks

Total assets of four central banks

in % of GDP, base 100 in 2006Q2

11Observation 1. The ECB has implemented a number of different credit easing policies since 2008. It has also considerably increased the size of its balance sheet. Starting in fall 2012, however, the ECB set itself apart from the Fed and the Bank of England with a significant reduction in its balance sheet.

12This recent divergence reflects a major difference between the steps taken by the Fed, on the one hand, and the ECB, on the other, to increase their balance sheets. The Fed primarily engaged in direct purchasing of US Treasury bonds and mortgage-backed securities via quantitative easing operations. [6] The ECB, for its part, primarily increased its balance sheet as a result of very long-term refinancing operations (VLTROs) which provided commercial banks with cheap three-year liquidity. Two VLTROs carried out in 2011 and 2012 allocated around 1,000 billion euros to banks in the eurozone. The reduction in the ECB’s balance sheet observed in 2013 stems primarily from the early repayment by the most stable European banks of loans obtained as part of these two VLTROs, since such loans became less appealing to them as the repayment deadline approached.

13Hence, while the Fed directly and durably injected liquidity into both public and private bond markets (and primarily mortgage markets) through the direct purchasing of assets, the ECB instead made abundant liquidity temporarily available to banks in the eurozone. This difference in approach relates to the specific constraints of the ECB (mostly the absence of federal sovereign debt, and a ban on monetising government deficits), but primarily to the role of bank loans in financing businesses in the eurozone: by offering more liquidity to banks, the ECB hoped to boost bank credit to the economy. In practice, however, the zone’s banking system has made only partial and insufficient use of the available liquidity, as demonstrated by the significant amounts received in early repayments of VLTRO loans. Moreover, the available liquidity has mostly been used to invest in local sovereign debt, as in the cases of Italy and Spain, rather than to finance businesses. This outcome has made both banks and sovereigns more vulnerable to each other. [7] The bank’s reluctance to extend loans to the private sector can be partly explained by the capital constraints faced by the banks as part of the current drive to reinforce prudential regulation. [8]

14The other notable difference between the Fed and the ECB since the start of the crisis relates to their communications regarding future action plans. On July 26, 2012, the President of the ECB announced that the Central Bank would do ‘whatever it takes’ to save the eurozone. On September 6, 2012, the ECB put in place a program aimed at purchasing sovereign debt directly on the secondary market (OMT, or Outright Monetary Transactions) in the event of renewed tensions in sovereign debt markets. These announcements, accompanied by a statement on the weakness of the economic recovery, the absence of any inflationary pressure, and on the intention of the ECB of keeping key rates at a very low level for a long period of time, have helped relax medium to long-term government debt interest rates, particularly for fragile economies. Small and Medium Enterprises (SMEs), however, have continued to borrow at far higher interest rates in peripheral countries than in Germany or France. [9] Furthermore, the ECB has remained less specific than the Fed with regards to forward guidance, the latter referring to specific time frames and specific unemployment and inflation thresholds. [10]

15To summarize, our assessment is that the ECB became de facto more restrictive than the other major central banks in 2013. This contributed to the relative strength of the Euro over that period. As of early 2014, deflationary pressures are tangible; inflation at constant tax rates stood at an annual level of 1% between April and November 2013, significantly below the ECB medium-term target of 2% (cf. Figure 2), [11] the unemployment rate within the eurozone is over 12%, the credit market continues its downwards trend, the transmission of monetary policy to different member countries remains extremely fragmented (cf. supra) and the drop in German inflation makes the adjustment of relative prices and price competition benefits more difficult to achieve in peripheral countries where prices now need to fall, with all the risks associated with deflation when debt levels are elevated.

2

Inflation rates, year-on-year change

Inflation rates, year-on-year change

in %
Note: HICP = harmonised index of consumption prices.

16Observation 2. The appreciation of the euro in 2013 can be explained by market expectations of a tightening of the ECB’s monetary policy.

17Given the ECB’s mandate and the tools it currently has at its disposal, a monetary expansion could take the following three forms:

• the direct purchase by the ECB of securitised small and medium enterprises (SMEz) loans would in our view be the most effective instrument to overcome the fragmentation of eurozone financing conditions. [12] The direct purchase of securities by the central bank offers two advantages; on the one hand, it is aimed directly at SME credit, which is still very restricted in peripheral countries, and on the other hand, it eases not only the liquidity constraints of banks but also their equity constraints by removing assets with high risk weights from their balance sheets. These equity capital constraints are the main obstacle to the transmission of monetary policy in peripheral countries;
• in order to secure bank liquidity in the long term, the ECB could offer a new VLTRO-type refinancing scheme with a longer maturity, such as five years, for example. A fixed rate over five years would offer maximum visibility for borrowing banks but would significantly expose the ECB to the risk of a future rise in interest rates. The ECB could, however, limit this risk by reserving the right to review the rate, within certain limits, after three years. In order to limit the use of such funds for the purchase of government debt (which would once again reinforce the deleterious link between sovereign risk and banking risk), the eligible collateral for long-term refinancing could be limited to securities backed by credit to the private sector. Such an approach would likely lead to a significant increase in the ECB’s balance sheet whilst respecting the economy’s financing profile in the zone, which remains bank credit; [13]
• finally, with regards to forward guidance, the ECB could commit to implementing unconventional policies such as those outlined above, at least for as long as a given measurement of inflation in the eurozone remains below a given threshold.

18Recommendation 1. The economic situation in the eurozone justifies a more sustained monetary expansion, through direct interventions in the securitised corporate credit market or a new, VLTRO-type refinancing operation with longer maturities and with collateral restricted to private securities. Furthermore, the ECB could tie its hands by committing to maintain low interest rates and to pursue unconventional policies for as long as average inflation remains below an explicit threshold.

Beyond monetary policy

19With regards to the eurozone, the heralded tapering of US unconventional monetary policy represents an opportunity since it is a chance for the ECB to strengthen its expansionary policy by allowing the euro to depreciate. Yet, it also represents a potential threat in that the reflux of foreign capital that would result from such a move is likely to affect the various economies in the zone to varying degrees and to cause interest rates to increase in the long term. Indeed, recent works suggest that a flexible exchange rate does not fully protect a country from the global credit cycle, which originates in the United States and is influenced by monetary conditions in the US. [14] In this respect, sovereign issuers in the eurozone have undoubtedly benefited from the low interest rates prevailing in the United States. A contrario, a tightening of credit terms by the Fed over the coming months could lead to a global increase in long-term rates with rapid repercussions on the refinancing conditions for States and financial institutions in the eurozone. This tightening could, in particular, trigger a new vicious cycle between banking and sovereign crises in Spain and Italy.

20In light of this risk, the ECB will need to clearly announce a monetary policy that differs from that of the Fed and, if need be, to activate its OMT programme. In the longer term, it will be important to put in place a comprehensive and highly operational and coordinated variety of macro-prudential instruments, in other words, instruments designed to improve the stability of the financial system as a whole over the course of the cycle. Macro-prudential tools are intended to control the aggregate debt of the private agents within an economy either by imposing limits on the amounts that can be borrowed against a particular type of asset (by requiring greater personal investment for property loans during a credit boom and relaxing this constraint in a downturn, for example), or by imposing capital ratios on banks that vary over the credit cycle. In order to be effective, of course, macro-prudential policy must be coordinated with monetary policy. [15] We believe it to be crucial that the main macro-prudential instruments be deployed by the European banking supervisory body (the Single Supervisory Mechanism), in close coordination with monetary policy. An intimate knowledge of the situation faced by individual financial institutions is a prerequisite to an effective macro-prudential policy, and only a supranational regulator will be in a position to identify the overall risks presented by a reversal in capital flows. The banking union project provides a Single Supervisory Mechanism, SSM (directly for larger banks, indirectly for others) with a number of micro-prudential monitoring instruments (capital and liquidity ratios, dynamic provisioning for banks, treatment of systemic banks, etc.). It also entrusts them with managing counter-cyclical capital buffers for those banks considered to be ‘systemic’. Member States, however, maintain responsibility for managing these buffers in the case of smaller banks (although the SSM can decide to increase capital requirements). Furthermore, in the current system national regulators maintain responsibility for other important macro-prudential instruments, such as restrictions on property loans. With this in mind, a situation in which, owing to a credit boom, the single supervisor increases capital requirements for banks whilst the national regulator relaxes the constraints governing property loans –two contradictory measures– cannot be ruled out. We believe that macro-prudential instruments should be more widely transferred to the single supervisor, even where this would require member States to be treated differently. [16]

21Recommendation 2. It seems appropriate to enlarge the prerogatives of the single bank supervisor by entrusting it with the main macro-prudential regulation tools in order to protect the eurozone from the excessive credit ebbs and flows triggered by US monetary policy decisions.

A very French debate

22The issue of the strong euro is a recurrent theme in the French economic debate; the expression ‘euro fort’ (‘strong euro’) returns 6.5 million Google hits as opposed to only 145,000 for the expression ‘euro faible’ (‘weak euro’). The contrast with the debate in Germany, where the strong euro is far from a leitmotiv, is striking, with the expression ‘stark euro’ returning only 1.4 million hits (as opposed to 764,000 for ‘schwach euro’). [17] Where does this difference in perception stem from?

23One explanation for the differences in sensitivity to the exchange rate on either side of the Rhine relates to the performance levels achieved by both countries in the export sector, with France’s global market share slipping by 44% between 1999 and 2013, as opposed to the mere 18% drop experienced in Germany. [18] According to Artus and Fontagné (2006), [19] however, the difference in performance recorded between the two countries does not relate primarily to differing levels of sensitivity to the euro exchange rate (and to its underlying appreciation over the course of the 2000s, cf. Figure 3) but rather to a reduced responsiveness on the part of French exports to international demand.

24Differences in perception between the two countries with regards to the issue of the exchange rate might also be linked to monetary culture and the more significant preference for low inflation in Germany than in France. The strong euro is perceived there as representing a shield against inflation, and in any case, the exchange rate is the result of a monetary policy implemented by an independent central bank whose primary mandate is price stability.

3

Real effective exchange rate of the euro

Real effective exchange rate of the euro

base 100 in January 1999
Interpretation: The real effective exchange rate is the average exchange rate across a number of partners, corrected for relative consumption prices. An increase points to an appreciation of the euro.

What would be the benefits of a weaker euro?

25The nominal exchange rate of a currency determines the relative price of goods and services produced in the country at a given time, as well as the relative value of the wealth accumulated in different currencies. A nominal depreciation in the currency temporarily improves the competitiveness of exporters, who can increase their margins or reap market shares to varying extents, depending on the sector. [20] However, French exports to countries outside of the eurozone –the only kind directly affected by a depreciation in the euro– account for only 11% of the French GDP, [21] meaning that the decline in the euro cannot be the sole response to our lack of competitiveness.

26At the same time, it is important not to underestimate the impact, this time negative, of a currency depreciation on the purchasing power of households and therefore on their ability to consume goods and services; by increasing the cost of imported goods, particularly those of which households will find it difficult to reduce their short-term consumption (such as gasoline, for example), depreciation forces them to cut expenditure on local services such as leisure and personal care services. Depreciation therefore results in two types of transfer, namely from net importing companies (such as the telecommunications sector, for example) to net exporting companies (such as the aeronautics industry, for example), and from households to exporting companies.

27Ultimately, an exchange-rate depreciation induces a revaluation of assets and liabilities in foreign currencies. France, on the whole, has more assets than liabilities in foreign currencies. In late 2012, for example, France’s gross assets stood at 5,829 billion euros, as opposed to its gross liabilities of 6,259 billion euros. Our calculations indicate a net position of approximately 306 billion euros invested in dollars and 247 billion in pounds sterling. Under these conditions, a 10% appreciation in the dollar against the euro would result in a capital gain of around 1.5% of the GDP, whilst an equivalent appreciation in the pound sterling would result in a gain of around 1.2% of the GDP. [22]

The impact of the euro on exports

28We present a new estimation of the effects of the euro on French exporters based on customs data for the 1995-2010 period, [23] the method and results of which are summarised in Box 1. The estimations suggest that, all other things being equal, a 10% depreciation in the euro in relation to a partner country outside the eurozone increases the value of the average exporter’s sales to this country by around 5-6%. This increase –the major part comes into effect in the same year as the depreciation is observed– stems primarily from an increase in the volumes exported (4-5%), with the remainder (0.5-1%) resulting from an increase in the mark-up on each unit sold (by means of a slight rise in euro prices). A 10% appreciation in the euro has a symmetrical impact, with the value of exports reduced by an average of 5-6% for an exporting company.

29In aggregate terms, the impact of a 10% depreciation in the euro on the value of exports is more significant, at around 7-8%, since the depreciation not only improves the situation of existing exporters but also paves the way for new firms in export markets. [24]

30These effects are indeed significant; with French exports outside of the eurozone accounting for 11% of GDP, a 10% depreciation in the euro against the currencies of all non-Eurozone trade partners would have a positive impact on demand of around 0.7 percent of GDP. This does not mean that GDP would increase by 0.7% since we are not taking into account here the effects of the depreciation on imports (and imports of energy and raw materials, which account for around 1.5% of the GDP, in particular), purchasing power, consumption, employment, wages, etc. According to the Mésange macro-econometric model, a 10% depreciation in the euro would result in a 0.6% increase in the French GDP after one year and a 1% increase after two years. [25]

31Upon closer inspection, there is no significant difference in terms of sensitivity from one major manufacturing sector to another. The main export industries (chemistry, automotive, food processing, aeronautics, etc.) in particular are very close to the French average. French exports to OECD countries, on the other hand, are more sensitive to exchange-rate variations than those to emerging countries. Exports to the United States, for example, increase in value by 9% if the euro depreciates by 10% against the dollar, as is the case for exports to the United Kingdom. This can be explained by the fact that products exported to OECD countries are more similar to and therefore substitutable with locally produced goods (and therefore more sensitive to price differences) than exports to emerging countries.

32The most productive firms (which are also the largest exporters) are less responsive than average to exchange-rate variations. [26] Indeed, they are able to better absorb exchange-rate variations into their margins, increasing their mark-ups when the euro depreciates and reducing them when the euro appreciates. Likewise, exporting firms that import a large proportion of their intermediate goods are less responsive than average to exchange-rate variations. For these firms, the increased competitiveness that results from a depreciation in the euro is limited by the increased cost of imported inputs. Those companies that perform the best in terms of export are also those that import the most intermediate goods. Internationalisation of supply chains tends to reduce firms vulnerability to exchange-rate movements. Indeed, reducing dependence on exchange rate instability is one major reason for diversifying production sites.

33It is often said that beyond a certain threshold, an appreciation in the euro would be particularly harmful to exports. This suggests that exchange-rate variations have a non-linear effect: small when the euro is close to its equilibrium level and large when it significantly deviates from this level. In the case of French exporters, we were unable to identify such threshold effects, with a 10% appreciation in the euro reducing the average company’s exports by 5-6% irrespective of the initial exchange-rate level.

34Another argument commonly put forward is that the increase in the variety and quality of French products might insulate exports against exchange-rate variations by making them less price-sensitive. There is no doubt that improved quality makes it easier to sell more, at a given price, but it is not clear whether quality helps reduce the sensitivity of exports to exchange-rate variations. In order to establish this, we have isolated those firms with the highest average export unit value, product by product. The idea is that these companies exporting more expensive goods specialise in ‘high-end’ products. Preliminary results do not suggest that those firms exporting ‘high-end’ products are less sensitive to exchange-rate variations, although our measure of quality is too crude to enable us to reach a definitive conclusion.

35A depreciation in the euro helps reduce the prices charged by French exporters in foreign currencies. In theory, the same effect can be achieved by means of a fall in euro prices, with no exchange-rate variation. Our results confirm that a nominal depreciation in the euro has the same effect on the value of exports as a fall in prices in France relative to foreign prices. This is significant since whilst the nominal exchange rate is no longer an instrument in the French government’s economic policy, the latter can, nevertheless, influence the prices charged by exporters by means of its arsenal of economic policies that have a direct impact on business costs (social contributions, taxation, energy costs, [27] etc.). French companies can also improve their competitiveness by improving the quality of their products. In this respect, economic policy has also a significant impact in terms both of support for innovation and research and of education. Whilst a nominal depreciation has only a short to medium-term effect on competitiveness, the structural reforms that make it possible to reduce costs, together with an improvement in the quality of the goods produced, has a permanent effect on competitiveness. The advantage of a nominal depreciation is that it has a rapid effect of competitiveness but it does nothing to reduce the need for reforms that have to improve the structural competitiveness of the French economy and that encourage all exports and not just those destined for counties outside of the eurozone.

1. Determining the sensitivity of French exports outside of the Eurozone to the exchange rate

We observe the export of goods in terms of value of around 100,000 French exporters every year to all markets outside of the Eurozone between 1995-2010 (over 4 million observations). This data is recorded by the customs authorities and combines the sales of each exporter over the course of a year for a particular product category and destination. Service exports are not available. We explain variations in sales from one year to another based on a series of control variables that are specific to the destination or year in question, as well as variations in the exchange rate.
The findings outlined below correspond to the specification used to quantify the impact of a depreciation in the euro. [a] The value of exports (in logarithm) of firm f in sector s to country i over the course of year t, ln?X_fsit is regressed on the real exchange rate of the euro against country i (in logarithm), ln?EURO_it on the destination GDP (in logarithm), ln?GDP_it on price levels in the destination country (in logarithm), ln?P_it and against a series of fixed effects (firm-sector-destination, and year) that make it possible to take the specific characteristics of the exporter into account for each sector and destination:

The estimated coefficients are indicated before each explanatory variable and their standard deviation are given in brackets. The three coefficients are significant at the 1% confidence level. A 10% depreciation in the euro increases the value of firm-level French exports by around 6%. When the estimation is made by regressing the export growth rate from one year to the next on the variation in the exchange rate, we find a slightly smaller effect (5%). In aggregate terms, this results in an elasticity of 0.76. This elasticity is slightly higher than that estimated based on aggregate data for Germany, for example, which stands at 0.6. [b]

The impact of the euro on imports

36Finally, we have estimated the impact of a euro depreciation on the manufacturing imports of French companies coming from outside of the eurozone. France’s total imports outside of the eurozone account for 13% of the French GDP. We estimate the average impact of a depreciation in the euro against each supplier country on manufacturing imports from the country in question. Our estimations are a little less accurate than they are for exports. A depreciation in the euro results in a drop in the volumes imported but increases the cost of each unit imported –two contradictory effects in terms of import value. The findings would suggest that a 10% depreciation in the euro increases the prices of imported manufactured goods by around 2% for the average importer. Demand for imported goods decreases by around 0-2% according to estimations. In the short-term, therefore, the price effect appears to outweigh the volume effect, with import value increasing by 0-3%. Only after two years does the volume effect start to outweigh the value effect. In aggregate terms, import value increases by around 3.5%. This finding is significant since it shows that depreciation raises the costs incurred by French importing companies.

37Observation 3. A 10% depreciation in the euro raises export value (outside of the eurozone) by around 7-8% but also increases the cost of imports (outside of the eurozone) by around 3.5%, with no short-term drop in the volumes imported. A 10% fall in prices in France in relation to those of its partners has the same effect on exports as a 10% depreciation in the euro.

38Recommendation 3. With relative prices having just as significant an impact on exports as the exchange rate, there is a need for increased vigilance with regards to the effects of public policy (social contributions on wages, taxation, energy costs, etc.) on French costs and prices.

How does one assess the value of the euro?

39Is the euro currently too strong, and if so, based on what criterion? Box 2 summarises the different possible approaches.

40One approach is to examine the historical evolution of the real effective exchange rate. Figure 3 shows (by reconstructing the euro using the currencies of the Member States prior to 1999) no evidence of a particular trend for the euro exchange rate since 1964. Between January 2012 and November 2013, the euro stood an average 2% above its average value since 1999 and 5% above its long-term average value. Taking into account margins of error on price measurement, the euro may be considered in line with its long-term level.

41A second approach involves questioning whether the euro should not have appreciated in the long term (as a result of accumulated external surpluses) or on the contrary depreciated (as a result of the trend for lower growth in the zone compared to the rest of the world). These two contradictory effects are difficult to quantify. Existing estimations are tarnished by a significant margin of error and do not inherently change the diagnosis regarding the euro. [28]

42A third approach focuses on the short term, questioning whether, independently of the long-term considerations, the euro might be too strong in light of the difficulty the zone has been experiencing in pulling through the crisis. In order to answer this question, we might look at how the real effective exchange rate of the euro has varied in relation with the output gap observed in the zone (the difference between actual production and potential production) and compare this with what has happened in the United States. In order to have a stabilising effect, the real exchange rate (nominal exchange rate adjusted by price differentials) must vary in the same direction as the output gap, appreciating when activity is relatively high and depreciating in the opposite case. Figure 4 shows that this was the case between 1995 and 2013 in the United States but not in the Eurozone. Whilst the euro did indeed depreciate in 2012, in line with the decline in activity, this is the exception rather than the rule. In 2009, whilst activity in the zone was plummeting, the euro remained stable in real terms. In 2013, the euro appreciated relative to 2012 whilst the GDP of the zone continued to decline. Conversely, in 2000, the peak in activity in the zone coincided with a particularly weak euro.

43With this in mind, and in accordance with our diagnosis regarding monetary policy (cf. supra), it can be said (Box 2) that the adoption of a more expansionary monetary policy, accompanied by a nominal depreciation of around 10% in the euro, would both help the ECB achieve its objective of inflation and alleviate the lack of demand in the zone. Since the euro is not overvalued with regards to the various long-term standards (cf. supra), it is also important not to expect a sustained depreciation.

44Observation 4. The short-term situation in the Eurozone in early 2014 would call for a temporarily weaker euro, consistent with a more expansionary monetary policy. However, since the euro is not overvalued with regards to the various long-term standards, we cannot expect a sustained weakening of the euro.

4

Real effective exchange rate and business cycle

Real effective exchange rate and business cycle

Interpretation: A positive output gap means that GDP exceeds its potential level; a real effective exchange rate of over 100 means that the real effective exchange rate is greater than its 1995-2013 average. Sources: OCDE (2013) : Economic Outlook 2013-2, BCE and Fed.

Interventions in the foreign exchange market

45Besides monetary policy, a central bank theoretically has two tools at its disposal to influence the exchange rate: it can intervene in the foreign-exchange market, that is to say, for example, buy dollars when it wants to weaken its own currency, and also issue statements on the exchange rate in the hope of coordinating market anticipations.

46Empirical research has highlighted a number of conditions that have to be met in order for intervention to be effective; the intervention should, where possible, be coordinated between several central banks, be declared to the market (rather than kept ‘secret’), concern substantial amounts, not be too frequent and not be ‘sterilised’. [29] Furthermore, ‘oral’ intervention, that is communications issued by central banks regarding exchange rates, can prove effective if indeed they are credible, particularly if they do not contradict declarations relating to monetary policy. [30] It would undoubtedly, however, be illusory to base an exchange-rate strategy exclusively on oral interventions, the credibility of which ultimately depends on the implementation of effective initiatives relating to monetary policy or on effective intervention in the market.

47Given that the ECB’s mandate focuses on an internal objective, namely price stability, it is to be expected that its intervention policy (both effective and oral) be covered by this internal objective, meaning that the ECB would only intervene when the euro is strong and the risk of deflation very real, or when the euro is weak and the inflationary risk imminent.

48There is therefore a great temptation for governments in the Eurozone to intervene by means of statements regarding the exchange rate in the hope of coordinating expectations themselves on a different equilibrium. We have analysed this possibility by studying the impact of eleven statements made by senior French and German policy-makers between 2006 and 2013 (Box 3). The findings are incontrovertible: oral interventions on the part of policy-makers have no effect on the exchange rate.

49Observation 5. Statements made by political leaders in an attempt to curb the exchange rate are generally ineffective.

What can international coordination achieve?

50Since the Plaza Agreement, which triggered (or accompanied) the turnaround in the dollar in September 1985, and the 1987 Louvre Agreement, which put an end to its depreciation, the G7 has kept reminding us of the benefits of flexible (and therefore unmanipulated) exchange rates but with very limited success. The G20 later took over in the form of a number of carefully prepared statements that differed little from one summit to another. Both the G7 and the G20 have a mixed track record when it comes to the international coordination of monetary policy and foreign-exchange policy. On the one hand, this coordination is destined for failure as soon as the central banks become independent with a well-defined mandate in terms of internal objectives. On the other hand, coordination has proven invaluable at a time of crisis, when swap agreements between central banks have made it possible to provide banks with liquidity in different currencies.

51Above and beyond its practical feasibility, there is no consensus regarding the benefits of coordinating monetary policy at international level, in a world where governments have instruments for dealing with erratic movements of capital (adjustment of the exchange rate, macro-prudential policies) and where it is important for each zone to maintain its monetary instrument for the purposes of pursuing a policy of non-inflationary growth. With this in mind, the idea of a zero-sum game, or even negative-sum game, portrayed by the concept of a ‘currency war’, is controversial. It would appear that the G20 is more useful in dealing with financial regulation than when seeking to coordinate monetary policies.

52The concept of ‘exchange rate manipulation’, which is regularly highlighted by the American Congress with regards to the Chinese currency regime, is of little significance when it comes to monetary zones with a flexible exchange rate and perfect capital mobility, in which monetary policy is focused on achieving internal objectives. In fact, neither the United States nor the United Kingdom have highlighted the depreciation of their currencies as a key objective of their quantitative easing policies (only Japan has done so, without fully acknowledging this before the G20). A parallel can, however, be drawn between trade protection and monetary protection when the central bank intervenes heavily in the foreign-exchange market with the aim of preventing the appreciation of its currency, for example.

53Existing empirical work shows that customs duties have a far greater impact on trade than exchange rates, more than likely because, unlike exchange-rate fluctuations, they can be considered to be long-lasting. [31] The equivalent of a 10% exchange rate undervaluation is not a 10% customs duty but rather a 1% customs duty. The parallel, here, is no less justified, which raises the question of possible claims before the World Trade Organization (WTO).

54With regards to the issue of exchange rate manipulation, Article XV of the WTO relies on the judgement of the IMF, whilst accepting the principle of foreign-exchange controls. Article IV of the IMF, meanwhile, prohibits the manipulation of exchange rates in order to ‘prevent effective balance-of-payment adjustment or to gain unfair competitive advantage’. However, no sanctions have been introduced. During the 2007 and 2012 reforms, it was stipulated that a country could intervene in the foreign exchange market in response to short-term movements in exchange rates that were deemed to be excessive. An exchange rate is considered to have been ‘manipulated’ if the country in question has staged foreign exchange interventions or implemented capital controls aimed at maintaining a long-term undervaluation of the exchange rate relative to its fundamental level, and if the aim of this undervaluation is to stimulate exports. These two conditions are indeed very restrictive, particularly since it is recommended that the Member State be given the benefit of the doubt. Furthermore, the IMF has no sanctioning power in the matter. Finally, no country has ever been sanctioned for manipulating its exchange rate because coordination between the IMF and the WTO is limited, because it is difficult to obtain proof of manipulation, and because the IMF is reluctant to identify ‘manipulating’ countries, particularly when these are important members. [32]

55In light of the hesitation on the part of international organisations, there is a danger that the concept of exchange-rate manipulation will be dealt with at the level of each individual monetary zone with the risk to trigger trade wars. For this reason we believe that it would be useful to re-examine the concept and its application at multilateral level, based on greater transparency on the part of central banks with regards to foreign exchange interventions in particular.

56Recommendation 4. International coordination should focus on financial regulation and crisis management. Accusations of exchange rate ‘manipulation’ should be re-examined at multilateral level in order to prevent the situation degenerating into trade wars.