FEER model is a method based on macroeconomic equilibrium. The core of macroeconomic equilibrium method is that current account equals capital account. FEER method focuses on the decision-making of current account. Generally speaking, the current account can be interpreted as a function of domestic total output (or total demand), foreign total output (or total demand) and the real effective exchange rate. In many applications of FEER method, the medium-term capital account equilibrium can be estimated according to relevant economic factors. Therefore, the real effective exchange rate can be obtained according to domestic total output (or total demand), foreign total output (or total demand) and capital account. The actual effective exchange rate obtained here is the exchange rate suitable for macroeconomic equilibrium, that is, the basic element equilibrium exchange rate (FEER) in 1983 as Williamson said. It can be seen that when the parameters of the current account model are given, especially when the current account flow is sensitive to the actual effective exchange rate, the exogenous net sustainable capital flow can be used to calculate FEER. It must be pointed out that FEER is only a method to calculate the equilibrium exchange rate, not a theory of exchange rate determination. Because the calculation implicitly assumes that the real effective exchange rate will gradually converge to FEER, the exchange rate determination theory embodied by FEER method is the current account determination theory of exchange rate.
The above analysis clearly shows that FEER calculation requires considerable parameter estimation and judgment, including: (1) current account model; (2) an estimate of the potential output of the country and its major trading partners; (3) Estimate or judge the equilibrium value of capital account. The first two aspects have been the subject of extensive theoretical and empirical analysis, and their concepts and calculation methods have been very clear. However, the meaning and calculation method of capital account are still worth studying. For example, Williamson believes that the basic principles for choosing a current account are: (1) check the past imbalances and their relationship with the level of effective savings and investment to see if they reflect rational economic behavior, which is contrary to misleading government policies; (2) Check whether the seemingly rational balance is sustainable, and if not, we must reduce the result to make it sustainable; (3) Check whether the results are internationally coordinated, and if not, revise the targets that are too high or too low for all countries until they are coordinated. From the perspective of international coordination, it is relatively reasonable that the proportion of general current account to GNP reaches 1% ~ 2%.
Williamson once inferred the current account target value of 14 countries and regions from the factors such as investment demand determined by debt cycle, the influence of population age structure on savings behavior and the judgment of sustainability and consistency. When Bayoomi (1994) calculated the equilibrium exchange rate of major industrialized countries at 1970, it was assumed that the target current account was equal to 1% of GDP. This ratio comes from the current account revenue and expenditure targets of countries put forward by the United States during the Smithsonian Agreement when discussing the appropriate exchange rate parity of major industrialized countries. Of course, the subjectivity of the above two methods is obvious. In order to solve this contradiction, Isard and Faruqee and others (1998) regard the current account balance as the difference between savings and investment under the condition of full employment. Savings and investment can be obtained according to the function of variables such as actual output and potential output gap and fiscal deficit under the condition of full employment. This method basically does not need to rely on subjective judgment.
Because FEER method uses a set of special economic variables to calculate exchange rate, it abstracts short-term cyclical and temporary factors and focuses on basic economic factors, so these basic factors are regarded as variables that may have medium-term influence. These variables may not necessarily happen in the future, in fact, they may just be a will assumption that has never been realized. In this sense, the FEER method measures the normative concept. In fact, Williamson (1994) has summarized the characteristics of FEER as the equilibrium exchange rate when it is consistent with the ideal economic conditions. This normative method itself should not be criticized, because it only reflects the method of measuring exchange rate under clearly defined economic conditions. Of course, people can choose different conditions to calculate the exchange rate.
From the above analysis, we can see that the exchange rate level estimated by FEER model has the characteristics of moving and changing. Because, first of all, the growth rate of different countries is different, and the difference in productivity requires the exchange rate appreciation of countries with faster growth. Second, deficit countries will accumulate foreign debts. In order to maintain the balance of current account, it is necessary to devalue the real exchange rate, so as to increase trade income to compensate for the increase in debt interest payment 13. Similarly, countries with persistent surpluses need real appreciation to support the absorption of relatively increased output. Finally, if the income elasticity of import demand and the domestic growth rate exceed the income elasticity of export demand and the growth rate of foreign countries respectively, it will have a negative impact on the current account in the long run, which also needs continuous depreciation to eliminate this negative impact. Even if they are equal, similar results will be produced if the initial levels of imports and exports are very different.
FEER model abstracts short-term cyclical factors and temporary factors, focuses on the analysis of the influence of basic economic factors on the equilibrium exchange rate, and reveals the essence of equilibrium exchange rate changes. This model provides a concise and systematic method to estimate the equilibrium exchange rate by analyzing the current account, and provides a basis for policy makers to evaluate the exchange rate. However, the local equilibrium method of FEER model also has obvious shortcomings:
(1) The calculation of this model implies two assumptions: first, the method assumes that the future output and the exogenous input of capital accumulation are consistent in two directions; Secondly, this method implicitly assumes that these inputs are independent of the actual exchange rate, in other words, it assumes that the model has a recursive structure, so that asset accumulation and output can be obtained before or independently of the exchange rate results. In fact, there are many reasons to prove that the above assumption is not valid. In terms of output supply, it depends on FEER, because the real exchange rate affects the actual consumption income or asset cost. In addition, when all industrial countries import an important proportion of capital goods, the real exchange rate will also affect the cost of capital. There is also a direct relationship between the real exchange rate and asset accumulation, but this is conducted by the real interest rate.
(2)FEER only analyzes the flow equilibrium and does not consider the stock equilibrium. But in fact, even when studying the medium-term equilibrium exchange rate, we should include stock indicators such as debt balance, because it will continue to affect the risk return.
(3) The exchange rate calculated by 3)FEER model is the equilibrium exchange rate when it is consistent with the ideal economic conditions. However, the estimation of the potential output of China and its related trading partners, the estimation or judgment of the equilibrium value of the capital account, etc. Both involve the problem of value judgment, so it has normative requirements.
(4) Although the model focuses on the analysis of the influence of basic economic factors on the equilibrium exchange rate, in actual analysis, due to the difficulties in data processing, how to screen short-term cyclical factors and temporary factors from real data is still a problem that needs to be further solved. In addition, in many cases, the variable effect that affects the actual behavior of exchange rate is not reflected in the actual calculation. Behavioral equilibrium exchange rates theory was put forward and applied by Peter B.Clark and Mcdonald (1998) and others in relevant literature, mainly aiming at the deficiency that the basic element equilibrium exchange rate model (FEER) does not reflect the variable effect that affects the actual exchange rate behavior. At present, the main research results in this field have been compiled and published by Sarquiss (1997).
In the previous analysis, we know that the concept of equilibrium exchange rate calculated by FEER method is the exchange rate when the current account is consistent with the sustainable capital flow under full employment. However, in many cases, the calculation does not reflect the variable effect that affects the actual behavior of the exchange rate. Under this method, as long as the internal and external equilibrium positions are not disturbed, the exchange rate will remain unchanged. However, in the behavioral sense, it is not clear whether the exchange rate is in equilibrium, that is, whether it reflects the effect of factors that determine the exchange rate in the medium term is not clear. BEER method tries to overcome this limitation, which includes direct econometric analysis of real effective exchange rate behavior. From the perspective of methodology, BEER method is a model strategy, trying to explain the actual behavior of exchange rate under the conditions of relevant economic variables, which is why Mcdonald and others call it behavioral equilibrium exchange rate. In Bill's method, the concept of equilibrium is given by a set of appropriate explanatory variables (generally estimated by their actual values), instead of using macroeconomic equilibrium as the concept of equilibrium to evaluate the real exchange rate as in FEER's method. Therefore, the feature of BEER method is to embed some variables related to the system behavior between exchange rate and its determinants in the model.
BEER method is characterized by using simplified model instead of FEER to estimate the equilibrium exchange rate. The core of this simplified model is to interpret the real effective exchange rate as a function of the economic basic factor vector with long-term lasting effect, the economic basic factor vector that affects the real exchange rate in the medium term, the temporary factor vector that affects the real exchange rate in the short term and the random disturbance term. Therefore, in any period, the total exchange rate imbalance can be decomposed into three aspects: short-term temporary factor effect, random disturbance effect and basic economic factors deviating from its sustainable level effect. It can be seen that the behavioral equilibrium exchange rate method can be used not only to calculate the equilibrium exchange rate, but also to explain the periodic changes of the real exchange rate in principle.
In the actual econometric analysis, the long-term exchange rate model starts from the similar risk-adjusted interest rate parity condition, that is, the realistic equilibrium exchange rate is determined by the expectation of the real exchange rate, the difference of the real interest rate at home and abroad and the risk premium. Assuming that the time difference of risk premium is a function of the relative supply of government debt at home and abroad, the obvious increase of the relative supply of domestic debt relative to foreign debt will increase the domestic risk premium, thus requiring the realistic balanced depreciation of the real exchange rate. Introduce a long-term equilibrium exchange rate and assume that it is equal to the expectation of uncertain exchange rate. According to Clark and MacDonald's analysis (1998), the long-term equilibrium exchange rate is mainly a function of three variables: terms of trade, Balassa-Samuelson effect and foreign net assets. Therefore, based on the above analysis, it can be concluded that the behavioral equilibrium exchange rate (BEER) is a simplified model of variable functions such as the difference of real interest rate at home and abroad, the relative supply of government debt at home and abroad, the terms of trade, the ballas-Samuelson effect, and foreign net assets. Clark and MacDonald (1998)~ 1] used the simplified model mentioned above to make an empirical analysis of the real effective exchange rates of the US dollar, the German mark and the Japanese yen. The results show that BEER method has good explanatory power.
Beer method does not directly consider internal and external balance. But in theory, the value of basic economic factors can be adjusted to the level of full employment and low inflation. That is to say, it can be adjusted to a level consistent with internal balance. Relatively speaking, there is no obvious relative adjustment method for external balance. There are two reasons for this phenomenon: first, the model is based on the assumption of uncompensated interest rate parity, so it can not effectively limit external unbalanced financing; Secondly, the model embodies such an adjustment mechanism, that is, it produces a balanced change in the real exchange rate that is compatible with the level of government debt and net foreign assets, thus achieving external balance, at least in the long run. This is the shortcoming of the beer method. BEER method includes direct econometric analysis of the real effective exchange rate behavior, trying to explain the real exchange rate behavior under the conditions of related economic variables, thus providing a better calculation method and explanation examples. From the results of empirical analysis, BEER method has a good explanatory power. NATREX (natural real exchange rates in English) was put forward by Jerome L.Stein in 1994. Its basic meaning is: without considering cyclical factors, speculative capital flows and changes in international reserves, the medium-term actual equilibrium exchange rate, which is determined by actual basic economic factors and can balance international payments. This theory mainly explains empirically the long-term change of the real exchange rate when the basic elements such as thrift and productivity are determined by real variables.
The basic assumptions of the natural equilibrium exchange rate model are: the price is clear from the market, and the output at this time has reached the inherent potential level; Adjust the real exchange rate to a realistic equilibrium level; Money demand and money supply are equal; The central monetary authority does not interfere in foreign exchange; Assume currency neutrality, regardless of monetary factors (such as nominal money supply, nominal price and nominal exchange rate system, etc.). ); Assume that long-term capital flows are relatively high. Under these assumptions, the natural equilibrium exchange rate model can be expressed by an equation similar to the national income account, that is, the difference between savings and investment is equal to the current account difference. It can be seen that the core content of the natural equilibrium exchange rate model is reflected in investment, savings and net capital flow (that is, the difference between savings and investment). When the real exchange rate depreciates and the net capital inflow (that is, the difference between investment and savings) increases, the real stock of physical capital, wealth (that is, the difference between the real stock of physical capital and net foreign debt) and net foreign debt change respectively, which in turn changes the required balance between investment, savings and current account, thus requiring a new equilibrium exchange rate level. Only when the economy reaches a long-term equilibrium state, that is, the basic economic factors and the actual asset stock remain unchanged, can the natural equilibrium exchange rate remain unchanged.
Exogenous basic economic factors, such as domestic and international frugality and productivity changes (including exogenous terms of trade and international real interest rate of small countries), will affect the natural equilibrium exchange rate in two ways: (1) First, it will affect the required investment, savings and current account, thus inducing corresponding changes in the medium-term natural equilibrium exchange rate; (2) By changing the actual stock of physical capital, the accumulation rate of wealth and the net foreign debt, we can change the trajectory of the natural equilibrium exchange rate to a new long-term equilibrium level. As long as it is not at the long-term equilibrium level, the natural equilibrium exchange rate is a function of exogenous and endogenous actual basic economic factors at any other level. A complete natural equilibrium exchange rate model can determine the medium-term actual equilibrium exchange rate (natural equilibrium exchange rate), the change track of natural equilibrium exchange rate and the long-term equilibrium exchange rate (stable equilibrium exchange rate). It is only a function of exogenous actual basic economic factors).
The basic viewpoints of this theory are: (1) The changing trend of real exchange rate can be explained by such basic factors as productivity and frugality (including exogenous terms of trade of small countries and international real interest rate); (2) Natural equilibrium exchange rate is a moving equilibrium real exchange rate consistent with the continuous changes of exogenous and endogenous basic economic factors, and the real real exchange rate constantly adjusts the moving equilibrium real exchange rate; (3) The long-term impact of borrowing foreign debt on the natural equilibrium exchange rate depends on whether borrowing foreign debt is used for consumption or net investment. If foreign debt is borrowed for consumption, the real exchange rate will change from appreciation to gradual depreciation. If foreign debt is used for productive investment, the real exchange rate will go from appreciation to gradual depreciation and then appreciation (if the country becomes a net creditor). (4) The most effective way to improve the current account is to change the difference between investment and savings, rather than changing the demand and supply of tradable goods through monetary and trade policies.
NATREX method is a model family, and the specific NATREX model depends on the following characteristics: the economic scale of tradable goods and assets relative to trading partner countries, the elasticity of supply and demand of foreign goods and assets, and the substitutability of goods, assets and countries. According to these characteristics, NATREX model can be subdivided into two categories: asset market type and money market type. NATREX model has different characteristics in different countries. But their similarities are mainly concentrated in: medium and long-term equilibrium exchange rate; Changes in investment, savings and long-term net capital flows, as well as changes in the stock of material capital, wealth and net foreign debt caused by all these changes; The influence of the above changes on the moving equilibrium real exchange rate.
The core of NATREX model is a set of general equilibrium exchange rate models, whose rationality and optimal behavior determine the equilibrium real exchange rate. These models provide logical economic judgments for empirical research. Before NATREX method was put forward, it was difficult for general exchange rate determination theory to explain the fluctuation of nominal exchange rate, especially the reason why nominal exchange rate and real exchange rate of US dollar appreciated first and then depreciated in 1980s. NATREX model has a good explanation for this. Stan inspected the exchange rates of the United States and G- 10 countries, and the changes were consistent with the changes of basic economic factors, and the predictions made were also consistent with the actual situation. Guay Lin and Stein studied Australia, a relatively small country, and the actual basic factors better explained the changes in the real exchange rate. Liliane Cmuhy—Veyrac and Michele Saint Marc studied medium-sized economies such as Germany and France, which can not affect international interest rates, but can affect trade prices and trade structure. The results show that the change of real effective exchange rate is consistent with the change of basic elements, but its adjustment speed is slower than that under the condition of floating exchange rate.
Finally, it should be pointed out that NATREX is an empirical rather than normative concept. It discusses the exchange rate determined by actual basic factors on the basis of existing economic policies, and does not involve social welfare issues. This is the difference between Natrex and FEER. In FEER model, the actual equilibrium exchange rate is the exchange rate when the current account measured by potential output is consistent with the willing capital flow, in which the willing capital flow is not distorted by public policy. This normative requirement is the main difference between the two. Of course, when necessary, the optimal strategy can also be reflected in the NATREX model. In addition, Natrex model has the characteristics needed to consider the condition of stock balance, which is also different from FEER model. From the empirical analysis results, NATREX model has good explanatory power. The theory of equilibrium real exchange rate (ERER) was first put forward by Sebastian Edwards in 1989, and later revised by Edwards (1994) and Ibrahim A 'Elbadawi (1992). According to Edwards' definition (1989), the equilibrium real exchange rate (ERER) is the relative price of nontradable goods and tradable goods. If the values of other related variables (such as taxes, international terms of trade, business policies, capital flows and technology, etc.). ) is sustainable, then the internal and external equilibrium will be realized. When the non-tradable goods market is clear now and in the future, realize internal equilibrium; When the current account balance between reality and the future is consistent with the long-term sustainable capital flow, the external balance will be realized.
ERER is different from the traditional definition of purchasing power parity ERER is not only affected by the actual value of basic factors, but also by the expected value. Abadawi (1992) thinks that a successful ERER model should include at least three elements: (1) It should develop ERER into a long-term function of a basic element; (2) Allow the real exchange rate to be adjusted flexibly and dynamically according to ERER; (3) Short-term and medium-term macroeconomic and exchange rate policies should be allowed to have an impact on ERER.
ERER model originated from salter-Swan model, which is a small country or dependent economic model. Such a country is too small to affect its own terms of trade (salter,1959; Swan, 1960). Williamson once called ERER the definition of "Chicago". Therefore, ERER model is mainly aimed at the equilibrium exchange rate of developing countries.
Because the economy of developing countries has the characteristics of foreign exchange control, trade barriers and parallel exchange rates, Edwards made the following assumptions about the model: (1) Consider a small open economy with three commodities in its market: export, import and non-trade, which are produced in China and consumed by import and non-trade; (2) There is a dual exchange rate, the fixed nominal exchange rate is suitable for commodity trading and the freely floating nominal exchange rate is suitable for financial trading; (3) Local residents hold both local currency and foreign currency, and the private sector has also accumulated a certain amount of foreign currency; (4) Government revenue comes from undistorted tax revenue and domestic credit creation, government consumption imports and nontradable goods; (5) The government and private individuals cannot borrow from abroad, and there is no domestic public debt; (6) there are import tariffs, and the tariff revenue is transmitted back to the public without distortion; (7) The export price expressed in foreign currency is fixed and equal to1; (8) Initially, it was assumed that there were effective capital controls and no international capital flows. Later, it was relaxed to the point that the government disapproved of capital control and there was capital in and out; (9) Have complete expectations.
Under the above assumptions, Edwards constructed the 16 equation including five parts: asset determination, demand department, supply department, government department and external department. When the non-tradable goods market is cleared, the external sector is balanced (that is, the changes of international reserves, current account balance and money stock are equal to or equal to zero respectively), the fiscal policy is sustainable (that is, government expenditure is equal to undistorted tax revenue), and the asset portfolio is balanced, the economy is in a stable state. At this time, the exchange rate has reached a long-term sustainable equilibrium. Using these conditions and combining the equations, we can get a long-term equilibrium real exchange rate model. According to the derived model, Edwards found that the long-term equilibrium real exchange rate is a function of basic economic factors such as terms of trade, capital flow, tariff level, labor productivity and government consumption. In the short term, the change of currency variables will also affect the change of real exchange rate. Edwards (1989) specifically analyzed the influence of actual disturbance on the equilibrium real exchange rate: (1) terms of trade and the equilibrium real exchange rate. Generally speaking, the improvement of terms of trade will produce two effects: income and substitution. Generally speaking, the improvement of terms of trade will lead to the appreciation of the equilibrium real exchange rate, while the deterioration of terms of trade will lead to the depreciation of the equilibrium real exchange rate. However, due to the actual situation in different countries, the opposite situation may also occur. (2) Tariff and equilibrium real exchange rate. The main function of tariff is to regulate imported goods by influencing their prices. Assuming that other conditions remain unchanged, if Marshall-Lerner condition holds, then reducing tariffs will lead to the depreciation of the equilibrium real exchange rate. Of course, different types of tariff cuts have different effects on the equilibrium real exchange rate. If it is a short-term tariff reduction, it will deepen the degree of depreciation. If it is a long-term tariff reduction, the impact on the equilibrium real exchange rate will be more stable than the short-term tariff reduction. If we expect tariff reduction, the opposite result may occur. (3) Foreign capital inflow and equilibrium real exchange rate.
Generally speaking, the inflow of foreign capital will lead to the appreciation of the equilibrium real exchange rate in the short term. In the long run, if foreign capital flows in the form of loans and all of them enter the consumption sector without creating any productivity, it will lead to the depreciation of the equilibrium real exchange rate; If it enters the investment sector, it will lead to the appreciation of the equilibrium real exchange rate. If foreign capital flows in the form of industrial investment, especially in export-oriented sectors, it will lead to the appreciation of the equilibrium real exchange rate. (4) Technological progress and balanced real exchange rate. Due to the Balassa—Samulson effect, the technological progress in the trade sector is relatively faster than that in the non-trade sector, so the relative prices of tradable goods and non-tradable goods will gradually decrease over time, so in general, technological progress will lead to the appreciation of the equilibrium real exchange rate.
On the basis of the long-term equilibrium of the real exchange rate model, Edwards (1994) also constructed the structural dynamic equation of the real exchange rate movement. The structural dynamic equation shows that the change of the real exchange rate is mainly influenced by four forces: one is the automatic adjustment mechanism of the exchange rate, that is, the real exchange rate is gradually adjusted to the equilibrium real exchange rate when other conditions remain unchanged; Second, the policy adjustment mechanism, that is, macroeconomic policies such as monetary policy and fiscal policy are gradually adjusted to their sustainable levels; Third, the lag adjustment mechanism of nominal exchange rate actually reflects the nominal depreciation rate; The fourth is the adjustment mechanism of the gradual narrowing of exchange rate differences between different foreign exchange markets.
Since the equilibrium real exchange rate is mainly influenced by the basic economic factors such as terms of trade, capital flow, tariff level, labor productivity and government consumption, it can be concluded that the change of the real exchange rate is a function of variables such as terms of trade, the proportion of non-tradable goods consumed by the government to GDP, tariff level, technological progress, capital flow and the proportion of investment to GDP. Edwards used 12 data of developing countries to verify the most important meaning of the model: (1) In the short term, the real exchange rate change is a reflection of both the actual disturbance and the currency disturbance; (2) In the long run, the change of the equilibrium real exchange rate only depends on the real variables; (3) In the short term, discontinuous expansionary macroeconomic policies will lead to the imbalance (overvaluation) of the real exchange rate; (4) If the depreciation of the nominal exchange rate is out of balance with the real exchange rate and accompanied by appropriate macroeconomic policies, then the depreciation of the nominal exchange rate will have a lasting impact on the balance of the real exchange rate. The results of empirical analysis well support the meaning of the model. In addition, Iobadawi constructs an ERER model with long-term expectation on the basis of Edwards model, and makes an empirical analysis with the data of Chile, Ghana and India. The result is more reasonable than Edwards model.
ERER model fully considers the characteristics of transitional economies of developing countries, so it is more suitable for calculating the equilibrium exchange rate of developing countries and evaluating the real exchange rate. However, there are some insurmountable contradictions in the practical application of this model. First of all, the data of some variables in the model cannot be obtained and must be replaced by other variables. Secondly, some variables that reflect the characteristics of transitional economies in developing countries may not be obvious in the regression results. In addition, the equilibrium real exchange rate obtained by the model cannot be directly observed, and how to measure its accuracy is worth studying.