Reducing the U.S. trade deficit : an analysis of the effect from the Chinese renminbi on the U.S. trade deficit

(1)

__________________________________________________________________________________

Reducing the U.S. trade deficit:

An analysis of the effect from the Chinese renminbi on the U.S. trade deficit.

Thesis BSc. Economics A.N. Mijnen1 University of Amsterdam

University supervisor: Ms. S. Chan MSc Internationale Monetaire Betrekkingen

University of Amsterdam

Abstract This paper analyzes the relation between the Chinese renminbi and the U.S. trade deficit. Literature and panel data approach is used to analyze to what extent a revaluation of the Chinese renminbi would have effect on the U.S. trade deficit. The analysis shows that the Chinese renminbi is not being kept undervalued and that a revaluation would not necessary reduce the imbalance of the U.S. trade deficit.

Keywords U.S. trade deficit, Chinese renminbi, Revaluation, Undervalued, Chinese Central Bank, Bank policy

_____________________

1

A.N. Mijnen is a Bachelor student at the University of Amsterdam |Faculty of Economics and Business | Specialization w.r.t. Economics and Finance | Penningkruid 8, 1911 PW Uitgeest, The Netherlands | Student-ID: 10000208

(2)

2

Introduction

According to the National Bureau of Statistics of China, the Chinese Economy has grown with an average of 9.5% in the past ten years (NBSC, 2014). With a BNP of 8.2 trillion in 2012, China has become a more important player in the world economy. Since 2003, China is facing high pressure from criticisms of the U.S. and Japan, demanding an appreciation of the Chinese renminbi. They accuse the Chinese Central Bank of ‘manipulating’ the exchange rate and keeping the value of the RMB low. They claim that this is the cause of the U.S. balance of trade deficit (Bowles and Wang, 2006). China refuses to accept that this is caused by the value of the Chinese currency. Therefore it was interesting for me to investigate the two parts of this accusation, namely if the Chinese Central Bank is keeping its currency undervalued and whether a revaluation of the Chinese currency would reduce the U.S. trade deficit.

First I start with reviewing the literature about the undervaluation of the RMB. I found some interesting ways to determine undervaluation. One of them is the relative PPP approach, which will be used in this research to validate the value of the renminbi.

The relative PPP approach will be explained in detail and is useful for determining whether a currency is being kept under- or overvalued. After validating the value of the renminbi, I start

searching for literature about trade deficits. Some interesting studies are mentioned where they also researched the effect of a revaluation on the U.S. trade deficit. It is important to know what factors affect a trade deficit and what model is useful to determine the effect of a revaluation on a trade deficit.

After the literature research, an empirical research will be done where the PPP approach and the MC model will be applied to determine the undervaluation and the effect on the U.S. trade deficit. The MC model of Dr. R.C. Fair will be described which will be used whether a revaluation would reduce the U.S. trade deficit or not. The results are given in Appendix 1&2 and are briefly analyzed in this paper. At last, a conclusion is taken based on the literature- and the empirical research.

(4)

4

Undervaluation renminbi

To value the renminbi, several methods have been developed. One is the PPP approach, which states that the exchange rate can be determined by price levels. In 1986, a simpler version of the PPP approach was introduced by The Economist. It implied that prices of a Big Mac hamburger were compared between two countries, which should indicate the currency value differences between the countries (Clements, Lan and Seah, 2012). The current Big Mac index states that the Chinese currency is undervalued by 40,7% (The Economist, 2014). The actual exchange rate (USD/RMB) is at 6,05 while the exchange rate according to the Big Mac Index should be at 3,59. The undervaluation of the renminbi, measured by the Big Mac index, could be a result of the nontradable nature of the product and the wage differences. This is because China’s labor and rental costs are significantly lower than those in the United States (Yuan, 2004). The Big Mac Index is not perfect, but provides plausible insights into the operation of currency markets (Clements, Lan and Seah, 2012). They conclude that the index is a biased predictor of currency values. The main reason for this is that the BMI relies on absolute PPP, which ignores barriers to international trade. On the other hand, their research showed that the relative version of the PPP approach is proven to useful for validating the value of a currency. The research of Lu & Chang (2011) showed evidence that a long-run PPP relationship in China holds. They tested this by using the threshold cointegration test where they used the U.S. and Japan as base countries. Therefore a relative PPP approach will be used to determine the under- or overvaluation of the Chinese currency.

The PPP approach has 2 forms: absolute PPP and relative PPP (Pilbeam, 2013). The absolute PPP states that the exchange rate between 2 countries (let’s say U.S. and China) should be identical to the ratio of price levels between those countries. This is derived from a concept called ‘the law of one price’, which states that the real prices of goods should be the same across countries. This can be expressed in the following formula:

Where S is the exchange rate (S = spot rate) and P stands for the price level for the home country. The asterisk is used to indicate the foreign price level. Some conditions must be met for this

relationship to hold: goods must be freely tradable on the international market, the price levels used should be based on the same goods and all prices should be indexed in the same year.

The research of Woo (2008) tested the absolute PPP approach. He used the CPI of China and the United States as price levels. Next he defined the goods arbitrage and the relationship between the

(5)

5

price levels of tradable and non-tradable products. After defining that, the relationship between the PPP adjusted- and the actual exchange rate was derived. He concluded that the actual exchange rate always would indicate undervaluation compared to the absolute PPP exchange rate. He states that the ‘market-clearing exchange rate’ is the correct exchange rate, as there is no intervention of any central bank.

The relative PPP approach determines the real exchange rate by adjusting the nominal exchange rate with the inflation differences between the selected countries. It is determined by the change in prices as indicator of a change in inflation and therefor produces an indication of a change in the spot rate. The theory states that a country which experiences a higher inflation rate compared to another country, should devalue its currency towards this country to keep the goods prices for both countries equal. This can be expressed in a formula:

Where indicates the forward rate and is the current exchange rate. So the left part of the formula shows the change in the exchange rate. The stands for the inflation rate in the foreign country and stands for the inflation rate in the domestic country (Pilbeam, 2013).

The PPP approach for this research will be based on an earlier research where the renminbi was tested for under- or overvaluation versus the dollar, Euro and Yen (Nair and Sinnakkannu, 2010). They applied the relative-PPP approach to determine the deviation of the exchange rate from the adjustment for inflation. Rearranging the formula above by multiplying both sides with gives:

Where is the nominal exchange rate and is the PPP-adjusted exchange rate at time t. To indicate the inflation for the foreign and home country, the changes of the price levels will be taken. The is an indicator for the inflation rate in the foreign country, whereas stands for the inflation rate in the home country.

Chang and Shao (2004) also used a regression form based in the PPP theory to determine whether the Chinese currency was undervalued or not. They adjusted the exchange rate based on the

(6)

6

absolute PPP theory. After that they did a regression with the real exchange rate, GDP per capita and an error term. They found that it was undervalued by 22,5%.

Another research that used the PPP approach was from Frankel (2006). He used the absolute PPP approach and did a regression (including the real income per capita variable) on a cross-section of 118 countries, where he found an undervaluation of 22.4%.

Goldstein and Lardy (2007) also tested whether the Chinese currency was undervalued or not, but used a more exchange-rate related approach. They used the ‘underlying asset’ approach which uses the real effective exchange rate to determine at what level it would produce an equilibrium, where the ‘underlying’ current account position is equal to net capital flows. The result was that the renminbi was undervalued by a range of 20-40%.

Another exchange rate-related research had been done by Zhang and Pan (2004). They used the exchange rate to check whether its path in the long-run deviates based on the inflation differential, relative GDP growth and government intervention. They concluded that the Chinese currency was kept undervalued by a range of 15-22%, which comes close to the results of the authors mentioned before.

Factors that influence the trading balance

To find the factors that can influence a trade balance for a country, different theories are examined. This paper uses the ‘CA’ notation as an indicator of the trade deficit (Current Account), which is calculated by subtracting the import from the export (X-M). One of the theories states that ‘Absorption’ seems to have a significant impact on a trade balance. Absorption is the total sum of expenditures in a country. This includes consumption, investments and government expenditure. To reduce the current account deficit, a country should reduce its absorption relative to income, and the opposite has to be implemented abroad (McKinnon, 2007). Given the following accounting identity (which only applies for a two country model):

Where Y represents the GDP (output), A is absorption (total spending: C + I + G) where C is

consumption, I represents investments and G is government expenditure. The asterisk on a variable indicates a foreign counterpart. According to McKinnon a reduction of absorption (A) should increase the current account, but at the same time an improvement of the foreign absorption (A*) should occur symmetrically because that will reduce the trade imbalance of a large country like the United States. Otherwise, unilateral absorption adjustment by either side to right the trade imbalance will always be frustrated. For example, a demand for an increase in consumption in China or Japan requires an equal reduction in consumption within the United States. This implies that ∆A = -∆A* (McKinnon, 2007).

(7)

7

The absorption theory also states that if real income falls faster than absorption, it will have two consequences. The export will decline relative to imports, which implies that the current account (hence, trade balance) worsens. Another effect is that the domestic currency will depreciate. The opposite will happen when absorption declines faster than real income (Daniels & VanHoose, 1998). The effect of an increase in absorption will lead to a decrease of the current account, as the import will rise when the domestic spending increases. The real exchange rate ( ) has a positive effect on the current account (which will be explained later in the elasticity approach). This is summarized in the Marshall-Lerner condition:

( )

Another theory is the elasticity approach. This approach focuses on the relationship between the (real) exchange rate and the flow of goods and services as measured by the current account. First, the real exchange rate will be determined by using the following formula:

( )

Hence, S stands for the ‘Spot rate’, measured as ‘home currency divided by the foreign currency’. For example when the home country is Europe and the foreign country is the U.S., then the spot rate would be €/$. This will be multiplied by the price ratio of the foreign country ( ) and the home country ( ). And in our example, this would be a $/€ ratio. Therefore it is a relative price, namely the price of American goods relative to European goods. The real exchange rate will be used in this theory for the current account in the following formula:

( ) ( ) ( )

The first effect in the equation ( ( )) is positive because if the real exchange rate rises, American goods become more expensive and the demand for European export goods will rise. This will also reduce the demand for American export and therefor the import for Europe will be lower. So the second effect in the equation is ( ( )) is negative. The equation shows how the current account depends on the real exchange rate (Daniels & VanHoose, 1998). But as you can see, the real exchange rate formula looks similar with the absolute PPP approach formula for the real exchange rate mentioned before. As this paper uses the relative PPP approach, the effects might be similar but they would not exactly be the same.

China and the U.S. trade deficit

China has the second largest economy in the world and with an average export of 1841.88 (x

hundred million $) per month in 2013, it is the largest exporting country in the world (GAC, 2014). As mentioned before, the United States accuses the Chinese Central Bank of keeping its currency undervalued. This would be the main reason for the trade deficit problem according to the United States. The research of Bowles and Wang (2006) concluded however that the academic side of that

(8)

8

debate showed no consensus that the Chinese currency is undervalued, or that the exchange rate is being ‘manipulated’. They also showed that the debate concluded that a revaluation of the RMB wouldn’t drastically reduce the U.S. current account deficit. The main reason for this is that the U.S. doesn’t produce the products that it imports from China anymore. So this means that the U.S. would have to go to higher cost suppliers.

Zhang and Sato (2012) did a structural VAR approach to determine whether the renminbi exchange rate is the main cause of the trade surplus of China. They found that the trade balance of China is only affected by a world demand shock or a trade balance shock and that an exchange rate shock has an undetermined pattern, but has little effect on the trade balance. They concluded that the trade imbalance between the U.S. and China is mainly caused by the US output shock during the post unification period and even before, while the exchange rate effect is small.

This of course raises the question whether a revaluation of the Chinese currency would reduce the U.S. trade deficit, when another approach is used then the one’s mentioned above. A model that is useful for predicting and forecasting when experimenting with policy changes was found by Dr. R. Fair. This model is called the Multicountry (MC) model of Fair (2004) and it has proven to be useful for predicting output outcomes for countries when policy changes occur (Villaverde, 2008). It is composed for two sections: The U.S. model and the rest of the world model. The U.S. model has six sectors: financial, foreign, federal government, households, firms and state and local governments. It uses thirty stochastic equations and adds 101 accounting identities. The rest of the world model uses thirty-eight countries for which structural equations are estimated. Another twenty-five countries are added but only the trade share equations are estimated. Then the two sections are formed together with other equations to determine trade shares and world prices. It uses a 2SLS technique for the model’s equations (Villaverde, 2008). The theory behind the model is that households choose their relevant future variables and make expectations, to maximize their utility (Fair, 2004). Fair states that the main variables chosen are the expenditures and the labor supply. Firms also make expectations and maximize expected profits. The main variables for firms are prices, wages,

production, investment, employment and dividends. The model assumes that firms are behaving in a monopolistic competitive environment. The model also assumes that the expectations that are made by households are not rational, because they don’t have all the information about the complete model. The last assumption is an econometric assumption, namely that the variables are stationary around their deterministic trend. Because when this assumption is violated, the estimates of the error terms won’t be accurate.

(9)

9

The model

Information flows in one direction: first the government, then to the firms and at last to the households. The decisions are made before transactions occur at the beginning of the period, and transactions occur in the rest of the period. The steps taken can be presented as follows:

1. Government decides monetary and fiscal policy and therefor also determines the interest rate.

2. The firms receive the tax- and interest rates and solve their maximization problem for the next periods. Here the price, wage rate and amount of labor are determined.

3. The households receive information on the tax- and interest rate and also on the price, wage and amount of labor that firms choose. Then they solve their maximization problem, where labor supply and consumption are determined.

4. After the household decisions are made, transactions take place.

Zhang (2012) applied the MC model to forecast what will happen to the U.S. trade deficit, when a revaluation of the renminbi would occur. From his research he rejects that a significant appreciation of the RMB (one time 25%) would reduce the trade imbalance between the U.S. and China.

According to Dr. Zhang this is due to a low overall price elasticity in combination with a high domestic price elasticity for tradable goods, and because the price pass-through in the US market is limited. The effect of a revaluation would have an adverse effect on the Chinese economy and a little effect on the U.S. trade deficit.

Another approach has been done by Groenewold and He (2007), they used the two-country model of Rosen and Yellen, only added a third country representing the rest of the world. The model consists of demand and supply equations for import and export, which are determined by the real exchange rate and the GDP of both countries. Next, they defined the trade deficit between the U.S. and China as the ratio of U.S. exports and imports and taking the log of the variables used for the demand and supply equations. Their conclusion was that an appreciation would also have little effect on the trade imbalance between China and the United States. Even with a large revaluation (say 50%), the effect on the trade deficit will be small under the assumption of proportionate import and export

contributions.

Woo (2008) researched whether China should implement a revaluation to reduce the U.S. trade deficit. In his research he did a regression for the private saving rate on the productivity growth, the aged working population ratio, the young working population ratio, the length of retirement and the interest rate. He found that the backward financial system in China made the non-government saving rate higher than the one in the United States. This would be the cause of the CA surplus of China. He

(10)

10

concluded that China shouldn’t do a revaluation versus the dollar, but instead do multilateral adjustments (U.S. reduces its budget deficit while China reduces its import trade barriers) and a cooperation to make the Doha Round (negotiations of the WTO to reduce trade barriers between countries) successful.

Another research had been done by Bhattarai and Mallick (2013). They used the Ricardian

comparative static and dynamic general equilibrium models to test whether the undervaluation of the renminbi affects the U.S. trade deficit. They tested with quarterly data of China’s relative wage cost, interest rate differentials and the real exchange rate. He concluded that the undervaluation of the renminbi is a cause of the rising trade deficit of the United States. He found that rising relative prices in China would lower the benefits to China and this would gradually reduce the U.S. trade deficit. He states that higher production cost and lower welfare of households, together with a rise of the relative GDP in China would reduce the imbalance of trade for the United States.

Empirical research

PPP approach

One of the main issues of this study is to determine if the renminbi is being kept over- or

undervalued. To validate this, a relative PPP approach will be applied here. Therefore an assumption had to be made, namely that the PPP in both the U.S. and in China holds.

First the methodology for the PPP approach will be discussed and after that the results will be analyzed. After determining the under- or overvaluation of the renminbi, the MC model of Fair will be discussed and applied to see whether a revaluation of the renminbi would have effect on the trade deficit of the United States. This will be done for a one-time 25%-, 40%- and a linear 25% revaluation.

The methodology that was used for the PPP approach will now be discussed.

The first indicator that will be used for this approach is the renminbi’s monthly spot exchange rate against the USD (from April 2011 until April 2014). For simplicity, the average (taken per month) spot rates will be taken. The second indicator that will be used is the consumer price index (CPI) of China and the U.S., which are shown in Appendix 1. The CPI data will be used to determine the price levels of the countries. After that the changes in the price levels will be calculated to determine the inflation rates of both countries.

After the PPP-adjusted exchange rate is determined (the way of calculating is mentioned before in the theoretical part), it will be used to calculate the difference with the nominal exchange rate at time t. This will be done by subtracting the nominal exchange rate from the PPP-adjusted exchange

(11)

11

rate. When the output value is negative, the currency is being kept undervalued. If the outcome is positive, the currency is being kept overvalued. This value will be divided by the nominal exchange rate and multiplied by 100 to get a percentage of over- or undervaluation. The following formula will be used to do this:

(( ))

Where stands for the under- or overvaluation as a percentage at time t and where and are the nominal- and the PPP-adjusted exchange rate at time t respectively.

In the graph below, the results from the relative PPP approach are presented from the period April 2011 until April 2014. Figure 1 is based on the calculated values from the table in Appendix 1. The summarized table shows by how much the Chinese currency is over- or undervalued, along with its standard deviation. The average values and standard deviations are calculated based on the values in Appendix 1, they all represent the period that’s indicated under ‘Year’. The PPP theory states that a country with a low inflation rate should devalue its currency against a country with a high inflation Figure 1: The under- or overvaluation of the RMB in percentages.

Based on the values calculated in appendix 1. Period April 2011 – April 2014.

rate in the long run (Coakley, Flood, Fuertes and Taylor, 2004). From Figure 1 it’s clear that the renminbi is not being kept undervalued for the last three years. The graph presents the outcome of the relative PPP approach, which shows that the under- or overvaluation stays within a +/- 1,5% margin. With an average of 0,085% and a standard deviation of 0,68 (taken over three years), the Chinese currency is not considered to be kept undervalued against the USD. The low standard deviation of the undervaluation indicates the low volatility of the exchange rate.

The general relative PPP states that even in the short run, the nominal exchange rate should be directly proportional to the relative price level (Coakley, Flood, Fuertes and Taylor, 2004). In Figure 2

-2 -1,5 -1 -0,5 0 0,5 1 1,5 2011 2011 2011 2011 2011 2012 2012 2012 2012 2012 2012 2013 2013 2013 2013 2013 2013 2014

Under- or overvaluation RMB in %

(12)

12

it shows that the nominal exchange rate is close to the PPP-adjusted exchange rate throughout the period, which doesn’t indicate undervaluation. All data of the exchange rate and the undervaluation percentages can be found in Appendix 1. Although the renminbi appreciates against the USD in this period, the adjusted exchange rate follows a same pattern, and is not being constantly undervalued.

Figure 2: The Nominal and PPP-adjusted exchange rate (USD/RMB)

Between April 2011 and April 2014. PPP-adjusted exchange rate calculations can be found in Appendix 1.

To conclude this section, the PPP approach showed that the Chinese Central Bank doesn’t artificially keep the renminbi undervalued for the past three years. The exchange rate of the renminbi is close to the adjusted exchange rate according to the PPP approach, where it fluctuates between over- and undervaluation of 1.5%. This rejects the first claim that the United States makes against China, namely that the CCB keeps the renminbi undervalued. Of course my methodology can be wrong and therefor it is interesting to investigate the second part of the problem, namely the claim that a revaluation of the renminbi would reduce the U.S. trade deficit. This will be investigated in the next section, with the use of the MC model.

Technical aspects of the MC model of Fair

Brief model explanation

The MC model uses four types of agents: households (h), firms (f), banks (b) and governments (g). I will describe for each agent how their decisions are made and in what order they occur in the MC model (Fair, 2013).

Households

When wages go up or prices go down, the household starts working more and consuming more. If wealth raises, the household works less and consumes more. When interest rates increase, the household saves at the beginning and dissaves at the end. An increase in tax rate causes the household to work less and to consume less. If a decrease in transfer payments occurs, the

household starts working more and consumes less. A labor constraint binds the household to work 5,7 5,8 5,9 6 6,1 6,2 6,3 6,4 6,5 6,6 20 11 20 11 20 11 20 12 20 12 20 12 20 12 20 13 20 13 20 13 20 13 20 14 20 14

Nominal exchange rate PPP-adjusted exchange rate

(13)

13

less therefor consume less. The household spends more time keeping money balance low when the wage rate is low or when the interest rate is high.

Firms

If a firm changes its prices or wages, other firms will follow the same pattern. Excess labor within a firm leads to a fall in employment, excess capital leads to a fall in investment. Because if a company decides to save capital (and thus gain excess capital), it will reduce its investment. An increase in interest rates leads to a substitution away from less labor-intensive machines and lowers investment expenditures, and vice versa. The interest rate can also influence opportunity costs of holding inventory, and so it may influence price and production decisions. If aggregate demand goes up, the firm prices go up and they start contracting. If demand goes down, the firm prices go down and they start expanding.

Banks

Banks play a passive role in this model where they don’t maximize a utility function. Each bank receives deposits from firms and households. They do need to keep a proportion of reserves in their bank. It can give loans and borrow money from a monetary authority. The before-tax profit is the difference between the interest revenue of its loans and the interest payments on the borrowed money. A bank is assumed to pay all of its after tax profits in dividends. This will keep the savings level zero and produces the following budget constraint:

Where is the amount of loans, is the level of deposits, is the bank reserves and is the borrowing from the monetary authority.

Government

The government is assumed to be both the fiscal authority and the monetary authority. So it receives taxes from the households, firms and banks and receives interest on the loans provided as monetary authority to the bank. The government pays interest on the borrowings and has wage costs and costs for goods purchased. Therefore it has the following budget constraint:

∑ ∑

Where is the level of savings, are the bank reserves, are the loans towards the bank and is the value of the net assets for the government. This equation states that every nonzero level of savings should result in a change in non-borrowed reserves or government borrowing. Fiscal

(14)

14

policy is taken to be expansionary. So the government can either increase its spending or lower its taxes to increase aggregate expenditures and aggregate demand.

Between two countries

The complete model uses 27 steps within a country and uses then trade equations between countries to complete the model simulation. The first 27 steps within a country can be found in Appendix 3, as the equations between countries are more relevant to discuss here (with the exchange rate involved). All the information in this chapter is obtained from the workbook of Fair (2013). Note: the subscripts that are being used here are government (g), firm (f), bank (b) and household (h).

After the equations had been done within a country, countries start to trade with each other. An example between two countries will be explained here. The model uses 17 equations per country and one redundant equation. First, the demand for both goods by the private sector of country 1 is determined:

( ), given that [1]

( ), given that [2] Where stands for the ’th derivative of function . The variable is the real interest rate, determined by the following formula: ( ), where is the expected value of for the next period based on the current information in this period. So the first equation shows the demand for the goods of country 1 ( ), the second shows the demand for the goods of country 2 ( ). Both demands are from the private sector of country 1. Next, the domestic price level is

determined. The domestic price level is affected by output and the import price level. A fall in import price levels would have a negative effect on the price level of domestic produced goods. So if a revaluation of the renminbi occurs it would have a deflationary effect on the economy. The model assumes that it is under pressure of the domestic demand ( ) and the level of import prices ( ): ( ), given that [3]

Where stands for the exchange rate (spot), when increases it means that the currency of country 1 depreciates. The level of exports is determined by a weighted average of domestic price levels and the world price level converted to the local currency. Since in the model of Fair China has some power to influence its export price levels, a revaluation would affect the export price level in both local currency and the dollar.

(15)

15

The level of imports is affected by the amount of absorption (which was mentioned before by McKinnon) and a relative price variable (which is determined by the domestic price, and import price levels). For example, a fall in the import price level relatively to the domestic price level causes the import to increase because the import demand increases. So after a revaluation it could be expected that the import level of China will increase and this could have a positive effect on the U.S. trade deficit (Zhang, 2012).

The model makes an assumption that there are no inventory investments. This means that production is equal to sales:

[4]

Where stands for the amount of goods that the government of country 1 buys from country 1 and stands for the amount of goods that country 2 buys from county 1. Of course, these values are important for this research as they determine the export and import levels and so the trade balance. The exchange rate will influence the import and export price levels and this will influence the

domestic price level. The total sales are affected by price levels and so the exchange rate can influence the import and export levels (and therefor the trade balance) through a price channel. Taxes paid to the government are determined next:

[5]

After that, the model determines the demand for real balances: ( ), given that [6]

Equation 7 is to determine the borrowings of the bank from the monetary authority: ( ), given that [7]

The private sector is assumed to be the only sector holding money:

[8]

This means that all money is held by the bank. Since banks are assumed to have no excess reserves, it means that the bank reserves should be equal to the total amount of money times the reserve requirements:

(16)

16

The demand for country 2’s bonds is dependent on the expected return on the bonds and the interest rate:

( ), given that [10]

But if Uncovered Interest Parity holds (but the model doesn’t assume that), this equation will fall out because the interest rate will then be equal across countries. After this, some financial equations are added where the amount of saving for each sector will be determined:

[11] [12]

[13]

After that, for each sector the budget constraint will be determined:

[14]

( ) [15] ( ) [16]

The variable stands for the international reserve. After these constraints have been determined, another constraint will be made for all sectors which states that someone’s asset is someone else’s liability w.r.t. the bond of country 1:

[17]

And at last, the redundant equation is added to the simulation which states that the change in reserves is zero across countries. But because this is already done by equations 11-17, the equation is redundant:

Drawbacks of the MC model

The major drawback that this model has is that during the simulation part, the model makes a lot of assumptions. For example: during step 5 & 6 (in the ‘within a country’ section in Appendix 3) it assumes that prices and wages are equal in the entire country, which is not realistic. Also during step 19 it is assumed that the government determines the new interest rate before the household makes

(17)

17

its decision, but in reality there is no guarantee that that will happen. Therefore the credibility of the model becomes less, because the assumptions make it less realistic.

Another drawback is that the model assumes that there are only two companies in each country, namely f and k. This limits the interaction that would have occurred between companies when there would have been, for example, 2.000 companies per country. Firms could start competing on prices or quantity/quality, invest in marketing etcetera. This would make it more realistic.

The model also takes the government policy as exogenous. This means that the government

purchases of goods ( ), the tax rate ( ), the discount rate ( ) and the reserve requirement rate ( ) are taken to be exogenous. This makes the model more static as the government always uses the same policy, which is not realistic.

Applying the MC model of Fair

At first, I used the MC model to make a forecast for the period 2014.1 – 2021.4 with nothing changed in the current situation (so no revaluation). After that I changed the dollar/renminbi exchange rate by 25% and 40%, so that I could compare the values with the values when no revaluation had occurred. So that means an appreciation of the RMB compared to the dollar by 25% and 40%. I chose to do a revaluation upwards because the U.S. claims that the currency is undervalued and that it should be revaluated to a higher value. Therefore it is interesting for this research to see what happens if such a revaluation will occur. I chose to do a 25% revaluation based on the studies before, which

determined the over- or undervaluation of the Chinese currency. The forecasted period of 25% will be first discussed and then the 40% revaluation prediction will be used to see what a larger

revaluation would do to the U.S. trade deficit. After the revaluations it was interesting to investigate whether a different revaluation policy would cause a different outcome. So instead of revaluating the currency with 25% immediately, I adjusted the exchange rate slightly every quarter until in 2021.4 the revaluation would be at 25%. My expectation is that the revaluation would reduce the U.S. trade deficit as the U.S. imports from China will be negatively affected by the revaluation. I expect that after the revaluation, products would become more expensive for the U.S. to import from China. This would reduce the import level for the U.S. relative to the export level, hence: a reduction in the U.S. trade deficit.

Using the model with no revaluation (current situation)

At first, it was interesting to see what the variables will do across time when no revaluation had occurred. This information would be useful to compare with the information that comes out the model when the revaluations occur. I made a forecast for the period begin 2014 until the end of

(18)

18

2021. Table 2 shows the outcomes of the MC model for the United States. PMus indicates the import

prices and PXus shows the export prices. The volume of export is indicated as Xus and the volume of

import as Mus. The domestic price level (PYus) and the level of export from the U.S. to China (Xus,ch) are

determined. At last, the current account of the United States was determined by subtracting the volume of import from the volume of export, given in the table as CAus. In table YY the results from

the MC model for China are given. The same variables as mentioned above apply for this table; only the us is now changed in ch as it is for China. For both Table 2 and 3 the period of measurement is

given in the first column, where Q stands for quarter (period of three months).

First the price levels for the U.S. are given in Graph 3 below, the values of the graph can be found in Table 2 in Appendix 2.

Graph 3: The price levels of the United States when no revaluation had occurred, based on the values of Table XX in Appendix 2.

The import (PMus) and export (PXus) price for the U.S. follow the same pattern and keep on rising in a linear way, as time passes. The domestic price level (PYus) is at a lower absolute value but follows the same pattern as the import and export prices. According to this model they tend to grow in a linear way.

The graph below shows the export (Xus), import (Mus), trade deficit (CAus) and the export level from the United States to China (Xus,ch). There are some interesting patterns here. The export level of the United States seems to be relatively constant, compared to the import level. According to this model, the import level of the United States keeps on growing which results in a larger gap between the export- and import level. The Current Account is calculated as export minus imports, therefor the CA-line keeps on declining as the import level relatively keeps rising compared to the export level.

0,000000 0,200000 0,400000 0,600000 0,800000 1,000000 1,200000 1,400000 1,600000 1,800000 20 14Q1 20 14Q3 20 15Q1 20 15Q3 20 16Q1 20 16Q3 20 17 Q1 20 17Q3 20 18Q1 20 18Q3 20 19Q1 20 19Q3 20 20Q1 20 20Q3 20 21Q1 20 21Q3 PMus PXus PYus

(19)

19

Graph 4: The import- and export levels of the United States when no revaluation had occurred, based on the values of Table 2 in Appendix 2.

Graph 5 below shows the price levels of China. The import- (PMch), export- (PXch) and domestic price levels show a similar growth as the United States level. The difference here is that the export price level of China grows relatively more than the import price level of China, which explains the relatively higher import volume in $ compared to the export level of the United States. The domestic price level is above the export- and import price levels, which differs from the situation of the United States where the domestic price level was below the export- and import price level. One remarkable thing is the triangle-shaped point in the import price line. The sudden raise and drop can be seen from the raw data in Appendix 2. It looks like dr. Fair made a typing error in the Eviews data source as the price level goes from 1,013 to 1,216 to 1,029. If this is changed to 1,013-1,016-1,029 it would make more sense as the price level would then follow the same linear pattern as the other price levels. So the 1,216 should be changed to 1,016 if we want to eliminate the triangle-shaped point. I chose to leave it this way because I don’t want to bias the model by adjusting things by myself, but it does explain the sudden raise and drop of the import price level for China.

-600.000 -400.000 -200.000 0 200.000 400.000 600.000 800.000 1.000.000 20 14Q1 20 14Q3 20 15Q1 20 15Q3 20 16Q1 20 16Q3 20 17Q1 20 17Q3 20 18Q1 20 18Q3 20 19Q1 20 19Q3 20 20Q1 20 20Q3 20 21Q1 20 21Q3 Mus Xus CAus Xus,ch

(20)

20

Graph 6: The price levels of China when no revaluation had occurred, based on the values of Table 3 in Appendix 2.

The graph below shows the export (Xch), import (Mch), trade balance (CAch) and export level from China to the U.S. (Xch,us). We can see that the level of exports from China to the U.S. remains relatively constant compared to the other levels within the graph. The export level of China shows a slight increase as time passes while the import level keeps on increasing until it passes the export level, which explains the drop in the trade deficit of China over time. So if no revaluation had been done, the model predicts that the trade deficit of the United States will keep on growing as time passes.

Graph 7: The import- and export levels of China when no revaluation had occurred, based on the values of Table 3 in Appendix 2. 0 0,5 1 1,5 2 2,5 20 14Q1 20 14Q3 20 15Q1 20 15Q3 20 16Q1 20 16Q3 20 17 Q1 20 17Q3 20 18Q1 20 18Q3 20 19Q1 20 19Q3 20 20Q1 20 20Q3 20 21Q1 20 21Q3 PMch PXch PYch -1.000.000 -500.000 0 500.000 1.000.000 1.500.000 2.000.000 2.500.000 3.000.000 20 14Q1 20 14Q3 20 15Q1 20 15Q3 20 16Q1 20 16Q3 20 17Q1 20 17Q3 20 18Q1 20 18Q3 20 19Q1 20 19Q3 20 20Q1 20 20Q3 20 21Q1 20 21Q3 Mch Xch CAch Xch,us

(21)

21

Using the model for a 25% revaluation

The model was made for Eviews, where all of the equations will be solved by the program. Before I made a simulation for the period 2014.1-2021.4 an adjustment of the exchange rate had to be made to obtain the results after a revaluation. So for example: the exchange rate in 2014.1 was at 6.275, but after an adjustment of 25% it was changed to 4.706. This was done for every quarter of the period mentioned before. After the new forecasted results were obtained (the variables mentioned in the appendix), I subtracted the new values from the old values and putted the change into

percentages of the old value. The obtained results can be seen in Appendix 2: Table 4 (U.S) and Table 5 (China).

Graph 8: Predicted values for the U.S. by the MC model, based on the values in Table 4 in Appendix 2, after a 25% revaluation.

From Graph 8 we can see that there is a large decline in the U.S. trade deficit in the first year. The import prices for the U.S. on average rose because the price of the Chinese products became 25% more expensive after the revaluation. Therefore when the companies tend to maximize their utility function according to this model, the price increases causes a drop in imports of the United States. This can be seen with the purple line which immediately declines after the revaluation. From Graph 9 we can see that the export level from China to the United States drops after the revaluation, which confirms the drop of imports from the United States. This effect in combination with a slight increase in the exports of the U.S. can explain the short-term drop in the current account.

The depreciation of the dollar w.r.t. the renminbi raises the import price level in local currency. This increase has two effects: first the demand for import falls and then the domestic price level will rise.

-10,00 -8,00 -6,00 -4,00 -2,00 0,00 2,00 4,00 6,00 PMus PXus Xus Mus CAus PYus Xus,ch

(22)

22

The depreciation also reduces the price of exports in dollars unless the country is small and thus a price taker (which is not the case here). This leads to an increase demand in exports.

The Current Account line (CAus) seems to have a ‘J-curve’ pattern (Pilbeam, 2013). The CA initially

reduces after the revaluation as the imports become more expensive and exports become cheaper. But since the export prices are lower, the competitiveness increases and therefor the exports start to increase. Local households will also tend to switch to local goods (because the domestic price level PYus lowers) instead of importing goods from another country. This will slightly increase the current

account level toward its original level, or even above it. This is known as the ‘J-curve effect’.

The revaluation seems to have a negative effect on the Chinese economy. Although products (there is no product differentiation in this model) become cheaper for Chinese companies to import, the import level still drops among with the export level. From Graph 10 we can see that the import prices for China dropped immediately 25% after the revaluation. Chinese companies who already imported from the U.S. will start importing more from the U.S. as the products became cheaper. This can be seen from the export level line from the U.S. to China, which increases over time. This explains the slight increase of the export level for the United States. The export level of China decreases as products became more expensive for the United States (See Graph 8, PMus). Therefore companies in China will make fewer profits, which results in a drop in the average price level within China (PYch) as they tend to boost sales for maximizing their utility function according to this model. The loss in sales also causes a drop in imports for China because the companies have lesser money available to invest. The overall effect of the loss in sales on imports dominates the effect that imports from the U.S. become cheaper. Therefore a decrease in the Chinese import level can be seen in Graph 10. Graph 9: The absolute values for the CA of china, based on the values in Table 6 in Appendix 2.

-700.000 -600.000 -500.000 -400.000 -300.000 -200.000 -100.000 0 100.000 200.000 300.000 400.000 20 14Q1 20 14Q4 20 15Q3 20 16Q2 20 17Q1 20 17Q4 20 18Q3 20 19Q2 20 20Q1 20 20Q4 20 21Q3 CA with no revaluation CA after 25% revaluation

(23)

23

The current account of China shows that it fluctuates between positive and negative effects (see Table 5, Appendix 2) and therefor it is difficult to determine what will happen to it. The large

percentage differences in Table 5 in Appendix 2 can be explained from the absolute values in Table 6, which are graphed below. As you can see, the CA with no revaluation is close to zero at period 2016.1, where the CA after the revaluation is at a way lower absolute value. This explains the huge percentage difference in Table 5.

What we can see in Graph 8 is that the effect on the current account of the United States is small in the long run, as the line slowly tends to return to the predicted values of the MC model when no revaluation had occurred. Perhaps companies tend to search for other, cheaper alternatives to import instead of China. This can be seen in the graphs as the imports of the United States (Graph 8) slowly returns to the original value while in Graph 10 it can be seen that the export from China to the United States keeps on dropping. Therefore a revaluation of 25% would have little effect on the U.S. trade deficit in the long run. One remarkable thing is that the import price of China shows a sudden drop in the third quarter of 2015. This happened for every simulation I did, but after two quarters it is returned to its original value. This can be explained from the typing error mentioned before in the import price level of China.

Graph 10: Predicted values for China by the MC model, based on the values in Table 5 in Appendix 2 with a 25% revaluation.

These findings are in line with the research of Zhang (2012). He found that a revaluation of the renminbi would have little effect on the economic variables of the United States, when he used a 25% revaluation. This can be seen from Graph 8 where the variables stay within a deviation range of [-8%, 4%]. Therefore his conclusion was that it would have little effect on the U.S. trade deficit. He also finds a drop in the price level of China and this would be one of the reasons for the fall in

-40,00 -35,00 -30,00 -25,00 -20,00 -15,00 -10,00 -5,00 0,00 20 14Q1 20 14Q3 20 15Q1 20 15Q3 20 16Q1 20 16Q3 20 17Q1 20 17Q3 20 18Q1 20 18 Q3 20 19Q1 20 19Q3 20 20Q1 20 20Q3 20 21Q1 20 21Q3 PMch PXch Xch Mch PYch Xch,us

(24)

24

imports for China, which is in line with my results. He used the elasticity theory mentioned before in combination with the MC model. He finds that the Chinese imports are mainly dominated by the income effect, rather than the price effect. This would confirm one of the findings in my research, that the overall effect of the total sale loss (and thus lesser spending), dominates the fact that products become cheaper.

Using the model for a 40% revaluation

It was interesting to test whether a large revaluation of the Chinese currency would have effect on the trade deficit, as the 25% revaluation proved that it would have little effect. After the changing the exchange rate one time by 40%, that means that for example in 2014.1 (when the exchange rate was at 6.275) the value was changed to 3.765, another forecast for the period 2014.1-2021.4 has been made. Then I subtracted the new values from the old values and putted them into percentages. The results are shown in the tables in Appendix 2: Table 7 (U.S.) and Table 8 (China).

All calculated values for the graph below can be found in Table 7 in Appendix 2. From Graph 11 it can be seen that on short-term the price level of imports are affected by the revaluation. They will rise with an average of 6,55% throughout the year 2014. But as we move further towards 2021 the change lowers until it’s below a change of 1%. So the effect on the long-run import prices of the U.S. is small.

Graph 11: The results from the MC model for the U.S. with a 40% revaluation based on the values calculated in appendix 2.

The export prices for the U.S. are affected throughout the period 2014-2021 with an average of 1.5%. The Current Account of the U.S. will drop significantly on the short term, with an average of 10.03% in 2014 and an average of 12.42%. This follows from the small increase in export volume (X) and the decrease in import volume (M). The export volume has increased because after the revaluation it

-15,00 -10,00 -5,00 0,00 5,00 10,00 15,00 PMus PXus Xus Mus CAus PYus Xusch

(25)

25

becomes cheaper for China to import from the United States, as can be seen in Table 8 where the price level of imports decreased with about 40%, which is caused by the 40% revaluation. The export level from the U.S. to China increases (Xus,ch) which confirms that companies from China start

importing more because of the lower import price level. The level of imports decreased because the products imported from China became more expensive for the U.S. after the revaluation. This can be seen in Table 7 in Appendix 2 where the price level of imports (PM) rises in the short term. The graph seems to follow the same ‘J-curve’ pattern which was explained in the 25% revaluation part.

Graph 12: The values predicted by the MC model for China with a revaluation of 40%. All values are based on calculations which are given in Table 8 in appendix 2.

What can be seen from Graph 12 is that the export volume of China will be significantly lower for the next 5/6 years and their import will slowly decrease as time goes further. The explanation for this is given in the last column, where it can be seen that the export volume from China to the U.S. keeps getting lower until it almost hits -50% in 2021Q4. The effect on the CA of China is unpredictable. It can either go up or down as can be seen in the 5th column of Table 8 in Appendix 2. The price drop of the import and export products of China cause the domestic price level to drop as well, as can be seen in column 6. The fall in export causes the GDP of China to fall and therefor the imports will fall as well (Pilbeam, 2013). The import price level again makes a drop in the third quarter of 2015 and returns to its original value two quarters later. Only this time the export price drops as well. Although in the short-run the CA changes for the U.S. are big (see Table 7), in the long term the changes are little. It follows the same pattern as we saw earlier in the experiment where a 25% revaluation had been done. With an average expected change of 0,94% in 2021 it can be shown that a revaluation of 40% would also have little effect on the U.S. trade deficit in the long run.

-60,00 -50,00 -40,00 -30,00 -20,00 -10,00 0,00 10,00 20,00 20 14Q1 20 14Q3 20 15Q1 20 15Q3 20 16Q1 20 16Q3 20 17Q1 20 17Q3 20 18Q1 20 18 Q3 20 19Q1 20 19Q3 20 20Q1 20 20Q3 20 21Q1 20 21Q3 PMch PXch Xch Mch PYch Xch,us

(26)

26

Using the model for a linear 25% revaluation

It was interesting to investigate whether a different revaluation policy would give different results. I chose to adjust the exchange rate every quarter until it reaches a revaluation of 25% precisely in period 2021.4. So therefore an adjustment of 25/32 % (≈0,0078%) had to be made as the period 2014.1-2021.4 consists of 32 quarters. So for every quarter the exchange rate was adjusted in a linear way. The results are shown in Table 9 (U.S.) and Table 10 (China) in Appendix 2.

Graph 13: The results from the MC model for the U.S. with a linear revaluation of 25%, based on the values calculated in appendix 2.

Graph 13 above shows some differences with the graphs of an immediate 25- or 40% revaluation. The margin of the United States graph lies between -2% and 1.5%, so that means that all variables remain close to the original values. There are some differences. The import of the U.S. slowly moves towards a lower level while in Graph 8 and 11 the import first declines heavily but then returns to the original value. This can be explained by the PMus line, which shows that the import prices slowly rise because the exchange rate is slightly adjusted every period. Therefore the import level will gradually become lower as products slowly become more expensive for companies to buy (Pilbeam, 2013). The decrease in import combined with a slight increase in export explains the drop in the U.S. current account. From graph 14 it can be seen that the import prices for China sharply decline (PMch). Companies from China can now import more for the same money after the revaluation, this effect slightly happens as can be seen from the export level from the U.S. to China line (Xus,ch). To be

competitive, firms need to lower their price as they become more expensive for the U.S. after the revaluation (Pilbeam, 2013). This can be seen with the red line (PXch). This will decrease their profits and therefor they can import less than normal. This can be seen with the purple line (Mch). The export from China to the U.S. show the same pattern as with the 25- and 40% revaluation as it keeps

-2,50 -2,00 -1,50 -1,00 -0,50 0,00 0,50 1,00 1,50 PMus Pxus Xus Mus CAus PYus Xus,ch

(27)

27

dropping, but the total effect is lower compared to the other 25% revaluation. Again the drop in the import price level is at the point where the typing error had been made, as we’ve seen earlier. Graph 14: The results from the MC model for China based on the values calculated in appendix 2, with a linear 25% revaluation.

The results show that on the long run the effect on price levels and import/export volumes are lower than when an immediate revaluation of 25% would occur. This policy can be used when banks choose for more economic stability, as the variables are less affected in the long run with a linear revaluation. But as the margin of the graph is low, in the long run the effect on the trade deficit will be small. This would be a deviation of ≈1.5% from the value of the trade deficit when no revaluation had occurred. -25,00 -20,00 -15,00 -10,00 -5,00 0,00 20 14Q1 20 14Q3 20 15Q1 20 15Q3 20 16Q1 20 16Q3 20 17Q1 20 17Q3 20 18Q1 20 18 Q3 20 19Q1 20 19Q3 20 20Q1 20 20Q3 20 21Q1 20 21Q3 PMch PXch Xch Mch PYch Xch,us

(28)

28

Combining the results

All experiments seem to have the same result, namely that the effect on the U.S. trade deficit is small. This result shows that the effect on the trade deficit will be the same regardless the volume of the revaluation or the policy that will be chosen. Graph 15 shows all the results for the U.S. trade deficit in absolute values for every experiment that had been done.

Graph 15: The absolute values of the trade deficit for the U.S. for every revaluation experiment. All the values used for this graph can be found in Table 11.

As can be seen from Graph 15, the absolute values for the trade deficit are close to each other. As we move further in time, the values tend to move closer to each other. The next graph shows the

percentage deviation from the absolute values of no revaluation, done for each revaluation experiment.

Graph 16: The deviations in of the trade deficit values for the U.S. for every revaluation experiment, given in %. All the values used for this graph can be found in Table 12.

From the graph we can see that the impact of a direct revaluation causes a quick drop on the short term, but in the long-run every experiment moves back close to the original value of the trade deficit.

-600.000 -500.000 -400.000 -300.000 -200.000 -100.000 0 20 14Q1 20 14Q3 20 15Q1 20 15Q3 20 16Q1 20 16Q3 20 17Q1 20 17Q3 20 18Q1 20 18Q3 20 19 Q1 20 19Q3 20 20Q1 20 20Q3 20 21Q1 20 21Q3 No revaluation 25% revaluation 25% revaluation linear 40% revaluation -15,00 -10,00 -5,00 0,00 25% revaluation 25% revaluation linear 40% revaluation

(29)

29

Conclusion

Earlier researches about the undervaluation of the RMB showed that it was being kept undervalued for a range of 15-40% based on the researches of Goldstein and Lardy (2006), Chang and Shao (2004), Zhang and Pan (2004) and Frankel (2006). The result of the PPP approach shows that the RMB is not being kept undervalued for the last three years, based on the relative PPP theory. The MC model showed that a revaluation of 25% would have little effect on the U.S. trade deficit in the long run, where in the short run the trade deficit is indeed affected. The revaluation of 40% showed a same pattern and also showed that it would have little effect on the U.S. trade deficit in the long run. This is in line with earlier researches done by Zhang (2012), Zhang and Sato (2012) and Groenewold and He (2007). Revaluating the currency with 25% in a linear way causes the effects on price levels and import/export volumes to be lower in the long run. But for the U.S. trade deficit, the effect remains small.

Of course, the results are predicted by the MC model and a drawback of the MC model is that it makes a lot of assumptions. As a result, the credibility of the predictions decline. There is also an unexplainable drop in import prices in the third quarter of 2015. Therefore, more research is encouraged.

References

Bhattarai, K & Mallick, S (2013). Impact of China's currency valuation and labour cost on the US in a trade and exchange rate model. North American Journal of Economics and Finance, Vol. 25, Special issue, pp. 40-59.

Bowles, P. & Wang, BT. (2006). 'Flowers and criticism': The political economy of the renminbi debate. Review of International Political Economy, Vol. 13, No. 2, pp. 233-257.

Chang, G.H. and Shao, Q. (2004). How much is the Chinese currency undervalued? A quantitative estimation. China economic review, Vol. 15, No. 3, pp. 366–371.

Clements, KW. & Lan, YH. & Seah, SP. (2012). The Big Mac Index two decades on: an evaluation of burgernomics. International Journal of Finance & Economics, Vol. 17, No. 1, pp. 31-60.

Coakley, J. & Flood, RP. & Fuertes, AM. & Taylor, M. (2004). Purchasing power parity and the theory of general relativity: the ﬁrst tests. Journal of international Money and Finance, Vol. 24, No. 2, pp. 293-316.

(30)

30

Daniels, J. & Van Hoose, D. (1998). International Monetary and Financial Economics, ISBN-13: 978-0-132-46186-3, South-Western College Publishing.

Fair, RC. (2004). Estimating How the Macroeconomy Works. Used on 11-06-2014 from <<http://fairmodel.econ.yale.edu/rayfair/pdf/2003APUB.PDF>>.

Fair, R.C (2013), Macroeconometric modeling. Paper used on 28-06-2014 from website <<http://fairmodel.econ.yale.edu/mmm2/mm.pdf>>.

Frankel, J. (2006). On the Yuan: the choice between adjustment under a ﬁxed exchange rate and adjustment under a ﬂexible rate. CESifo economic studies, Vol. 52, No. 2, pp. 246–275.

General Administration of Customs (2014), China exports, used on 3-6-2014 from << http://www.tradingeconomics.com/china/exports >>.

Goldstein, M. and Lardy, N. (2006). China’s exchange rate dilemma. American Economic Review, Vol. 96, No. 2, pp 422–426.

Groenewold, N. & He, L. (2007). The US-China trade imbalance: Will revaluing the RMB help (much)?. Economics Letters, Vol. 96, No. 1, pp.127-132.

Lu, YCR & Chang, TY (2011) Long-run purchasing power parity with asymmetric adjustment: further evidence from China. Applied Economics Letters, Vol. 18, No. 9, pp. 881-886.

McKinnon, R. (2007). The transfer problem in reducing the US current account deficit. Journal of policy modelling, Vol. 29, No. 5, pp. 669-675.

Nair, MKV. & Sinnakkannu, J. (2010). De-pegging of the Chinese renminbi against the US dollar: analysis of its effects on China's international trade competitiveness. Journal of the Asia Pacific Economy, Vol. 15, No. 4, pp. 470-489.

National Bureau of Statistics of China (2014), China GDP Annual Growth Rate, used on 22-04-2014 from << http://www.tradingeconomics.com/china/gdp-growth-annual>>.

Pilbeam, K. (2013), International Finance, custom edition for the UvA - fourth edition, ISBN 978-1-137-34657-5, Basingstoke (UK).

The Economist (2014). The Big Mac Index, used on 3-6-2014 from << http://www.economist.com/content/big-mac-index>>.

(31)

31

Villaverde, JF. (2008). Horizons of Understanding: A Review of RAy Fair's Estimating How the Macroeconomy Works. Journal of Economic Literature, Vol. 46, No. 3, pp. 658-703.

Woo, WT (2008). Understanding the Sources of Friction in US-China Trade Relations: The Exchange Rate Debate Diverts Attention from Optimum Adjustment. Asian Economic Papers, Vol 7, No. 3, pp. 61-95.

Yang, JW. (2004). Nontradables and the valuation of RMB - An evaluation of the Big Mac index. China Economic Review, Vol. 15, No. 3, pp. 353-359.

Zhang, F. and Pan, Z. (2004). Determination of China’s long-run nominal exchange rate and ofﬁcial intervention. China economic review, Vol. 15, No. 3, pp. 360–365.

Zhang, JXH. (2012). Will RMB appreciation reduce trade deficit in the US? Journal of the Asia pacific economy, Vol. 17, No. 1, pp. 171-187.

Zhang, ZY. & Sato, K. (2012). Should Chinese Renminbi be Blamed for Its Trade Surplus? A Structural VAR Approach. World Economy, Vol. 35, No. 5, pp. 632-650.

DATA SOURCES

CPI China: http://www.tradingeconomics.com/china/consumer-price-index-cpi CPI USA: http://www.rateinflation.com/consumer-price-index/usa-cpi

(32)

32

Appendix 1

Table 1: The over- (+) or undervaluation (-) and the standard deviation Year Under- or overvaluation

Average in % Std deviation Q2 2011 0,14 0,60 2012 -0,19 0,74 2013 0,11 0,59 Q1 2014 -0,82 0,53 TOTAL AVERAGE: -0,085 0,68

Where Q2 = the last three quarters of 2011 and Q1 = the first quarter of 2014

USD/CNY PPP-adjusted Difference in

Under- or overvaluation exchange rate exchange rate exchange rates in %

Apr 2011 6,52 -6,52 May 2011 6,49 6,50 0,01 0,17 June 2011 6,47 6,55 0,08 1,26 July 2011 6,45 6,47 0,02 0,36 Aug 2011 6,39 6,41 0,03 0,40 Sept 2011 6,38 6,37 -0,01 -0,10 Oct 2011 6,36 6,35 -0,01 -0,18 Nov 2011 6,35 6,29 -0,05 -0,85 Dec 2011 6,35 6,36 0,01 0,09 Jan 2012 6,30 6,35 0,05 0,72 Feb 2012 6,28 6,19 -0,09 -1,43 Mar 2012 6,31 6,26 -0,04 -0,71 Apr 2012 6,30 6,28 -0,03 -0,41 May 2012 6,31 6,28 -0,02 -0,39 June 2012 6,31 6,26 -0,05 -0,78 July 2012 6,31 6,30 0,00 -0,07 Aug 2012 6,33 6,29 -0,04 -0,62 Sept 2012 6,33 6,29 -0,04 -0,57 Oct 2012 6,30 6,32 0,02 0,27 Nov 2012 6,28 6,35 0,07 1,13 Dec 2012 6,29 6,33 0,04 0,62 Jan 2013 6,28 6,24 -0,04 -0,64 Feb 2013 6,28 6,30 0,02 0,26 Mar 2013 6,27 6,20 -0,07 -1,12 Apr 2013 6,24 6,30 0,05 0,86 May 2013 6,19 6,21 0,02 0,31 June 2013 6,17 6,21 0,04 0,68 July 2013 6,17 6,17 0,00 -0,05 Aug 2013 6,16 6,16 0,00 -0,01 Sept 2013 6,15 6,18 0,03 0,56

(33)

33 Oct 2013 6,13 6,17 0,04 0,73 Nov 2013 6,13 6,13 0,00 -0,03 Dec 2013 6,12 6,10 -0,02 -0,26 Jan 2014 6,10 6,09 -0,01 -0,10 Feb 2014 6,11 6,05 -0,06 -0,99 Mar 2014 6,14 6,09 -0,05 -0,81 Apr 2014 6,17 6,09 -0,08 -1,37

All relevant variables calculated to be used for the relative PPP approach.

Appendix 2

Table 2: the data for the United States, predicted by the MC model with no revaluation. Original case values in absolute form

USA

Prices Import- and Export levels

PMus PXus PYus Mus Xus CAus Xus,ch

2014Q1 1,282741 1,273031 1,079134 411.184 316.856 -94.328 29.253 2014Q2 1,285801 1,278843 1,085353 417.844 313.824 -104.020 28.426 2014Q3 1,289213 1,284713 1,091558 426.296 310.937 -115.359 27.686 2014Q4 1,293012 1,291075 1,098162 436.390 308.210 -128.180 27.016 2015Q1 1,308339 1,299233 1,105433 447.818 316.091 -131.727 30.335 2015Q2 1,312604 1,306452 1,112927 460.534 313.473 -147.061 29.722 2015Q3 1,317097 1,313995 1,120695 474.310 311.003 -163.307 29.152 2015Q4 1,321795 1,321820 1,128706 488.907 308.693 -180.214 28.618 2016Q1 1,339991 1,331684 1,137276 503.878 317.519 -186.358 31.924 2016Q2 1,344900 1,340040 1,145893 519.270 315.236 -204.034 31.419 2016Q3 1,349971 1,348497 1,154588 534.921 313.114 -221.807 30.932 2016Q4 1,355196 1,357011 1,163321 550.699 311.166 -239.533 30.463 2017Q1 1,374018 1,367349 1,172353 566.310 322.327 -243.982 33.821 2017Q2 1,379322 1,375971 1,181276 581.898 320.382 -261.516 33.361 2017Q3 1,384773 1,384543 1,190129 597.413 318.614 -278.799 32.908 2017Q4 1,390365 1,393050 1,198900 612.826 317.033 -295.793 32.463 2018Q1 1,408939 1,403157 1,207802 627.995 330.903 -297.092 35.980 2018Q2 1,414545 1,411579 1,216556 643.070 329.222 -313.848 35.533 2018Q3 1,420299 1,419919 1,225204 658.058 327.738 -330.321 35.090 2018Q4 1,426188 1,428178 1,233750 672.974 326.453 -346.521 34.652 2019Q1 1,444266 1,437859 1,242331 687.804 343.410 -344.394 38.425 2019Q2 1,450114 1,445971 1,250790 702.644 341.911 -360.732 37.975 2019Q3 1,456111 1,454015 1,259150 717.526 340.627 -376.899 37.526 2019Q4 1,462240 1,461997 1,267425 732.478 339.559 -392.919 37.080 2020Q1 1,479750 1,471248 1,275656 747.602 359.982 -387.621 41.196 2020Q2 1,485789 1,479073 1,283826 762.878 358.576 -404.302 40.728 2020Q3 1,491982 1,486862 1,291923 778.323 357.406 -420.917 40.259 2020Q4 1,498304 1,494615 1,299952 793.955 356.471 -437.484 39.789

(34)

34

2021Q1 1,514956 1,503436 1,307834 809.994 380.772 -429.222 44.320 2021Q2 1,521022 1,510974 1,315707 826.294 379.366 -446.928 43.811 2021Q3 1,527149 1,518467 1,323513 842.882 378.221 -464.661 43.299 2021Q4 1,533300 1,525911 1,331258 859.780 377.332 -482.448 42.785

Table 3: the data for China, predicted by the MC model with no revaluation. China

Prices Import/Export

PMch PXch Pych Mch Xch CAch Xch,us

2014Q1 0,970391 1,025467 1,470866 1.391.828 1.658.449 266.621 77.435 2014Q2 0,976932 1,027149 1,469550 1.391.006 1.652.186 261.180 77.970 2014Q3 0,983761 1,029237 1,468807 1.390.542 1.648.658 258.116 78.915 2014Q4 0,990849 1,031675 1,468548 1.390.380 1.647.427 257.047 80.234 2015Q1 0,998153 1,076762 1,517667 1.531.753 1.648.105 116.352 81.259 2015Q2 1,005811 1,080366 1,519132 1.530.933 1.650.521 119.588 82.617 2015Q3 1,013635 1,084313 1,521331 1.530.791 1.654.167 123.376 84.249 2015Q4 1,216070 1,088536 1,524135 1.531.206 1.658.761 127.555 86.109 2016Q1 1,029711 1,133443 1,573189 1.668.496 1.664.079 -4.417 87.425 2016Q2 1,037972 1,138482 1,577620 1.668.816 1.670.879 2.063 88.926 2016Q3 1,046341 1,143763 1,582667 1.669.946 1.678.150 8.204 90.569 2016Q4 1,054810 1,149240 1,588213 1.671.732 1.685.766 14.034 92.330 2017Q1 1,063375 1,192718 1,635641 1.807.172 1.693.638 -113.534 93.409 2017Q2 1,072016 1,198816 1,642569 1.809.445 1.702.748 -106.697 94.615 2017Q3 1,080743 1,205079 1,649930 1.812.489 1.712.048 -100.441 95.926 2017Q4 1,089556 1,211482 1,657646 1.816.156 1.721.501 -94.655 97.335 2018Q1 1,098453 1,253336 1,703329 1.954.120 1.731.088 -223.032 98.031 2018Q2 1,107404 1,260271 1,712206 1.958.762 1.742.039 -216.723 98.863 2018Q3 1,116434 1,267321 1,721371 1.964.051 1.753.159 -210.892 99.817 2018Q4 1,125544 1,274474 1,730783 1.969.867 1.764.448 -205.419 100.890 2019Q1 1,134735 1,314867 1,775060 2.114.561 1.775.917 -338.644 101.271 2019Q2 1,143971 1,322507 1,785497 2.121.761 1.789.032 -332.729 101.819 2019Q3 1,153281 1,330234 1,796137 2.129.473 1.802.383 -327.090 102.523 2019Q4 1,162669 1,338045 1,806967 2.137.613 1.815.979 -321.634 103.380 2020Q1 1,172134 1,377176 1,850144 2.292.905 1.829.830 -463.075 103.580 2020Q2 1,181613 1,385418 1,861934 2.302.799 1.845.652 -457.147 103.977 2020Q3 1,191145 1,393708 1,873889 2.313.096 1.861.762 -451.334 104.556 2020Q4 1,200718 1,402029 1,886009 2.323.747 1.878.163 -445.584 105.311 2021Q1 1,210317 1,439883 1,928269 2.493.121 1.894.853 -598.268 105.441 2021Q2 1,219448 1,446131 1,940442 2.504.972 1.908.703 -596.269 105.783 2021Q3 1,228978 1,452774 1,952771 2.517.171 1.922.958 -594.213 106.322 2021Q4 1,238936 1,459856 1,965274 2.529.708 1.937.701 -592.007 107.050

Reducing the U.S. trade deficit : an analysis of the effect from the Chinese renminbi on the U.S. trade deficit