International Financial Theory Based on Empirical Evidence - Essay Example

INTERNATIONAL FINANCIAL THEORY BASED ON EMPIRICAL EVIDENCE International Financial Theory Based on Empirical Evidence Introduction The three main theories in international finance are the purchasing power parity theory, interest rate parity theory and the international Fischer effect. In this report, the empirical verification for each of these theories is done. This report is divided into four sections. In the second section, a critical discussion on the existing theoretical and empirical evidence is done for the three theories. In the third section, original empirical data analysis for each of these theories is done .In the final section, based on both these, conclusions and recommendations are developed. 2. Critical Discussion on Existing Evidence 2.1. Purchasing Power Parity (PPP) Theory The PPP theory states that there will be price equalization of goods internationally once they are measured in the same currency due to arbitrage forces (Pilbeam, 2006). This theory is based on the law of one price, which states that identical products when sold in different markets will sell at the same price when expressed in common currency. The main assumptions for this law to hold good are the presence of a competitive market structure, absence of transport costs and other barriers to trade (Sarno and Taylor, 2002).There are two types of purchasing power parity; the absolute and relative power parity. The absolute power parity theory states that a rise in the home price level relative to the foreign price level will result in an equivalent depreciation of the home currency against the foreign currency. Relative power parity states that there will be adjustments for the exchange rate by the amount of inflation differential between two countries (Pilbeam, 2006). The main problem with the PPP theory is that it does not distinguish between traded and non-traded goods. Many studies have shown that this distinction is important for testing PPP since traded goods are determined mainly by international competition while non-traded goods are influenced by domestic supply and demand conditions (Officer, 1976, 1986). At the same time, some other authors have showed distinction between traded and non-traded goods as unclear (Sarno and Taylor, 2002).The other limitations are the assumptions underlying the PPP theory like perfect competition, absence of transport costs and barriers to trade which are questionable. Further, it is very difficult to find identical products to be compared for testing PPP in different countries (Rogoff, 1996). Thus PPP is theoretically ambiguous and remains and empirical question. Three types of empirical evidence for PPP have been done .They are based on graphical representations, simplistic data analysis and econometric evidence. The empirical studies by Frenkel and Mussa (1080) and Macdonald (1988) based on graphs obtained rejection of PPP theory for many developed and developing nations in the short run. However, the econometric studies like Lothian (1997), Sarno and Taylor (2002), obtained PPP to hold good in the long run for OECD nations. Officer (1986) distinguishing between traded and non-traded goods obtained PPP to hold good for the traded goods but not to the non-traded goods for OECD nations. Studies like Xu (2003) have shown that the results for PPP differ depending on the price index used. The study shows that PPP holds better while using tradable price index while it is less likely to hold while using CPI. Thus, the empirical evidence is also mixed and depends on factors like time horizon used in the analysis, types of goods, nature of price indices used etc. However, it is clear from the evidence that PPP holds good in the long run than in the short run. 2.2. Interest Rate Parity Theory The role of international capital movements is not covered in the PPP theory. The interest rate parity theory includes the role of capital movements and explains the relationship between interest rates and exchange rates. The interest rate parity is the equalization of the rate of return to financial assets across countries. Covered interest parity (CIP) states that the nominal or market interest differential between two currencies is equivalent to the forward premium or discount. If this is not so, arbitrage occurs and its profitability will stop only when the interest rate parity again prevails (Pilbeam, 2006). Uncovered interest rate parity states that the nominal interest rate differential between two currencies is equivalent to the domestic currency’s expected rate of depreciation (Erdemlioglu, 2007).Real interest parity implies that long-run real interest rates should be equalized across countries and measures inflation differentials by long-term nominal interest differentials. Further, the speed of convergence can be very fast, leaving monetary authorities with limited control over their real interest rate relative to other countries’ real interest rates (Frankel, 1979). The underlying assumptions for both the parity conditions are sufficient availability of funds for transactions, negligible transaction costs and free capital movement. In addition to these, for the CIP needs the assumption of forward market existence while uncovered interest parity needs the assumption of risk neutrality of investors (Pilbeam, 2006).For the real interest parity to hold good, the uncovered interest parity and PPP both need to hold good(Fuji and Chin,2001). The empirical evidence for covered interest parity shows that it holds good for short term for the developed nations while in the long term the results are mixed (Frenkel and Levich, 1975; Popper 1993; Chinn, 2007). In the case of many developing nations, forward market s not strong and hence the tests are not reliable. The empirical studies for uncovered tests, there is only weak evidence in favour of these tests in the short run (Frankel and Poonawala, 2006; Isard, 2006 etc) while some evidence is there supporting it in the long run (Chinn and Meredith, 2004; Chinn, 2007 etc). The tests for uncovered interest parity condition needs the data for market-determined interest rates, which are rarely available for developing nations, and hence this is a limitation. The tests for real interest parity conditions show that it holds good in the long run than in the short run (Cumby and Obstfeld, 1984; Singh and Banerjee, 2006; Chinn, 2007).However, this condition holds good only if both uncovered interest parity and PPP holds good which has their own limitations. 2.3. International Fischer Effect By putting together both PPP and interest parity theory, international Fischer effect is obtained. International Fischer effect states that real interest rates are equal across countries where real interest rates are risk free interest rates minus the inflation rate (Pilbeam, 2006). The international Fischer effect is based on the Fischer (1930) effect, which states that the real interest rate is not affected by the changes in the expected inflation rate because it results in equal changes in the nominal interest rate. However, the empirical evidence does not strongly support the international Fischer effect (Fama and Gibbon, 1982; Huizinga and Mishkin, 1986; Kandel et al, 1996; Hatemi-J and Irandoust, 2008 etc). The main reasons for these are suggested to be money illusion, peso problems in the market for debt, existence of a liquidity premium included in financial assets and parameter instability (Hatemi-J, 2009). 3. Empirical Analysis 3.1. Purchasing Power Parity Theory Relative PPP are examined using the following regressions to obtain the evidence in the short run. Only relative PPP is tested since it is generally agreed that absolute PPP never holds good in reality de to its unrealistic assumptions. Relative PPP: ∆ ln St = a1+ a2 (∆ln Pt- ∆Pt*) + ut..........................(1) The regression is done for different currencies using quarterly differenced data .Differenced data is used for de trending the data in order to make seasonality adjustment. Table 1 gives the estimation results for the relative PPP for the exchange rates of for different currencies against dollar. Table 1:Regression Results Exchange rate RelativePPP period a1 a2 DW Pond/dollar 73Q1-81Q4 81Q4-90Q3 91Q4-2008Q4 0.01(0.56) 0.00 (0.03) 0.02 (0.02) -0.17(-0.46) -0.21 (-0.26) -0.32 (-0.35) 1.83 1.85 1.88 Deutschmark/dollar 73Q1-81Q4 81Q4-90Q3 91Q4-2008Q4 0.00(-0.25) -0.01(-0.76) 0.02 (-0.77) 0.53 (0.89) 0.48 (0.59) 0.65 (0.65) 1.79 1.96 1.95 Yen/dollar 73Q1-81Q4 81Q4-90Q3 91Q4-2008Q4 0.00 (-0.43) 0.01 (0.93) 0.04 (0.98) 0.82 (1.82) 2.79 (3.67) 3.12(3.76) 1.94 2.01 2.13 Lira/dollar 73Q1-81Q4 81Q4-90Q3 91Q4-2008Q4 0.01 (0.74) -0.01 (0.70) 0.03 (0.65) 0.68 (2.25) 0.78 (1.67) 0.86 (2.1) 1.99 1.90 2.1 Source: IMF, 2009. The results show that relative PPP does not hold good for all exchange rates except the Lira dollar rate. For relative PPP to hold good, the coefficients a1 needs to be zero and a2 needs to be 1.In the table this holds good only for the Lira/dollar rate while in all the other cases, the coefficients are wrongly signed and statistically insignificant too. However, the regression analysis shows the estimated in the short run only. Table 2 shows the unit root test results for these currencies using the ADF tests. Table 2: ADF Test Results for the Real Exchange Rates Exchange rate With constant With constant&trend WithoutConstant&trend Inference Level Difference Level Difference Level Difference Level Difference Pond/dollar -2.2 -5.3**(1) -3.07 3.6**(4) -0.6 -3.6**(3) I(1) I(0) Deutschmark/dollar -2.79 -3.2**(4) -2.56 -5.2**(3) -1.3 -2.3**(2) I(1) I(0) Yen/dollar -2.63 -4.9**(1) -3.4 -6.5**(2) -3.5 -2.8**(0) I(1) I(0) Lira/dollar -2.82 -4.8**(1) -2.64 -6.3**(4) -1.3 -3.6**(2) I(1) I(0) The details of the test are given in appendix 1.The results show that all the rates are non-stationary and integrated of order1. Hence, table 3 gives the Johanson Jusleius(JJ) co integration results(details in appendix 2). Table 3:JJ Co integration Tests Null Hypothesis Maximum Eigen value test Trace test Alternative Statistic 95%Critical Value Alternative Statistic 95%Critical Value Pond/dollar Deutschmark/dollar Yen/dollar Lira/dollar Pond/dollar Deutschmark/dollar Yen/dollar Lira/dollar r = 0 r = 1 40.2* 36.1* 83.8* 85.6* 33.4 r  1 102.9* 95.3* 181.7* 182.4* 68.5 r  1 r = 2 27.0* 28.1* 47.5* 46.5* 27.0 r  2 62.6* 59.2* 97.8* 96.8* 47.2 r  2 r = 3 18.2* 21.1* 29.7* 29.8* 20.9 r  3 35.6* 32.8* 50.3* 50.2* 29.6 r  3 r = 4 11.0* 15.3* 18.1* 18.1* 14.0 r  4 17.4* 15.5* 20.6* 20.5* 15.4 r  4 r = 5 6..3* 4.3* 2.5 2.4 3.76 r  5 6.3* 4.3* 2.5 2.4 3.76 Note: * Denote statistical significance at 5- percent level. The co integration is fond out between the respective nominal exchange rates, domestic and foreign prices. The results show that in all cases, there is at least one co integration vector. Hence, relative PPP holds good in the long run. 3.2. Interest Parity Tests Table 4 tests for covered interest rate differentials based on monthly data collected from IMF’s International Financial Statistics in some selected developed nations. It shows that the mean differentials are positive and different from zero from 1990 to 2008 which shows that covered interest rate parity holds good in the short run for the selected developed nations in the period of analysis, Table 4: Covered Interest Rate Differentials Country Period Mean s.d. Absolute Mean s.d. UK 90M3-08M6 1.26 0.61 1.26 0.62 France 90M2-08M6 0.38 0.39 0.45 0.29 Spain 90M5-08M6 0.38 0.18 0.42 0.49 Germany 90M3-08M6 0.81 0.28 0.81 0.28 Italy 90M5-08M6 0.24 0.42 0.37 0.29 Portugal 90M3-08M6 0.03 0.67 0.56 0.37 Source: IMF (2009) 4. Conclusion and Recommendations In this report, the empirical evidence on three main theories PPP, interest rate parity and international Fischer effect have been examined. The results show that PPP does not hold good in the short run but it performs well in the long run. This poor performance in the short run can be due to the unrealistic assumption underlying the theory, which is problematic in the short run. However, these problems will be diminished in the long run and hence it holds good. The Covered interest parity holds well in the short run for developed nations. However due to data limitations, it is difficult to test it for developing nations. Uncovered interest parity holds good only in the long run while the real interest parity holds good in the long run for developed nations. The international Fischer effect does not hold good in any of the nations selected for analysis. Based on the above discussion, it can be recommended that theories like PPP, uncovered interest parity and real interest rate parity holds good only in the long run de to the problematic assumptions underlying these theories. However, these will diminish over time and finally they perform well over time. Time lags are thus an important factor determining the effects of these theories. In other words, these theories take considerable time to get their effects in real life though they can be theoretically well stated. References Chinn, Menzie D. and Guy Meredith (2004): “Monetary Policy and Long Horizon Uncovered Interest Parity,” IMF Staff Papers 51(3) (November): 409-430. Chinn M (2007): “World Economy: Interest Rate Parity”, Princeton Encyclopaedia of the World Economy. Cumby, R E. and M. Obstfeld (1984) “International Interest Rate and Price Level Linkages under Flexible Exchange Rates: A Review of Recent Evidence,” in J. F. O. Bilson and R. C. Marston (editors) Exchange rate Theory and Practice , Chicago, IL:University of Chicago Press. Dickey, D. A(1976). “Estimation and hypothesis testing in Nonstationary Time Series”, PH. D. dissertation, Ames, Iowa State University. Dickey, D.A. and W.A. Fuller (1979). “Distribution of the Estimators for Autoregressive Time Series with a Unit Root”, Journal of the American Statistical Association, 74, 427–431 Erdemlioglu, D M. (2007): “A New Test of Uncovered Interest Rate Parity: Evidence from Turkey”, Munich Personal RePEc Archive, No. 10787. Fama, E. and Gibbons, M.R. (1982), “Inflation, Real Returns, and Capital Investment”, Journal of Monetary Economics 9: 297-324. Fisher, I. (1930):”The Theory of Interest”,. New York: Macmillan Frenkel, J.A., and R.M. Levich(1975): “Covered Interest Parity: Unexploited Profits?” Journal of Political Economy 83(2): 325-338. Frankel, J. (1979): “A Monetary Approach to the Exchange Rate: Doctrinal Aspects and Empirical Evidence,” Scandinavian Journal of Economics, May, 78, 200-24 Frenkel, JA and Mussa, M (1980): “The Efficiency of Foreign Exchange Markets and Measures of Turbulence”, American Economic Review, vol.70, pp375-81. Frankel, J and Jumana Poonawala, (2006): “The Forward Market in Emerging Currencies: Less Biased Than in Major Currencies,” NBER Working Paper No. 12496 (August). Fujii, Eiji and Menzie D. Chinn (2001): “Fin de Siècle Real Interest Parity,” Journal of International Financial Markets, Institutions and Money, 11(3/4): 289-308. Hatemi-J, A. and Irandoust M. (2008), “The Fisher Effect: A Kalman Filter Approach to Detecting Structural Change”, Applied Economics Letters, 15 (8), 619-624. Hatemi-J, A (2009): “The International Fisher Effect: theory and application”, Investment Management and Financial Innovations, Volume 6, Issue 1. . Huizinga, J. and Mishkin, F.S. (1986), “Monetary Policy Regime Shifts and the Unusual Behavior of Real Interest Rates”, Carnegie-Rochester Conference Series on Public Policy 24: 231-274. IMF(2009): “International Financial Statistics”, September ,Washington Isard P (2007): “Uncovered Interest Parity”, IMF WP/06/96. . Kandel, S., Ofer, A., and Sarig, O. (1996), “Real Interest Rates and Inflation: An Ex-ante Empirical Analysis”, Journal of Finance 51: 205-225 Lothian J (1997): “Multi Country Evidence on the Behaviour of Purchasing Power Parity under the Current Float”, Journal of International Money and Finance, vol, 16, no.1, pp.19-35. MacDonald R (1988): “Floating Exchange Rates: Theory and Evidence”, London: Unwin Hyman. Officer L (1976): “The Purchasing Power Parity Theory of Exchange Rates: A Review Article”, IMF Staff Papers, vol 23 pp1-61. Officer L(1986): “The Law of One Price cannot be Rejected: Two Tests Based on the Tradable/Non Tradable Goods Dichotomy”, Journal of Macroeconomics,vol8,pp.159-82. Pilbeam K (2006): “International Finance”, United Kingdom: Palgrave Macmillan. Popper, H(1993): “Long-Term Covered Interest Parity—Evidence From Currency Swaps,” Journal of International Money and Finance, Vol. 12, No. 4, pp. 439-48. Rogoff L(1996) : “The Purchasing Power Parity Puzzle”, Journal of Economic Literature,vol.34,no.2,pp.647-68 Sarno L and Taylor M P (2002): “Purchasing Power Parity and the Real Exchange Rate”, International Monetary Fund Staff Papers”, Vol.49, pp.65-105. Singh M and A Banerjee(2006) : “Testing Real Interest Parity in Emerging Markets”, IMF WP/06/249. Xu Z (2003): “Purchasing Power Parity, Price Indices and Exchange Rate Forecasts”, Journal of International Money and Finance, vol.22, pp.105-130. Appendix 1 ADF Tests Time series data can be of two types (1) stationary and (2) non-stationary. A series is stationary if its mean, variance and covariance are invariant over time and otherwise it is said to be non-stationary. If the process is stationary, it can be represented by a simple algebraic model, whereas if it is non-stationary, it is not possible to model the process in terms of an equation with fixed coefficients, estimated from past data. Non stationarity can be due to the presence of unit root or due to the presence of a trend component. If it is the first case, then the process is said to be difference stationary and if it is the second case, then the process is said to be trend stationary. To test for stationarity, the unit roots for each of the series are examined using the Augmented Dickey Fuller (ADF) unit root tests developed by Dickey (1976); Fuller (1976); Dickey and Fuller (1979). Here the null hypothesis that the time series belongs to the difference stationary process against the alternative that it belongs to the trend stationary process has been tested using two models, one with constant and the other with constant and trend. The optimum number of lags used is determined using two model selection criteria (1) Akaike Information Criteria (AIC) and (2) Swartz Bayesian criteria. The model giving the least value of these two criteria is considered as the best model. Appendix 2 JJ Co integration Tests Johansen (1988) and Johansen and Juselius (1990) suggest two likelihood ratio tests for examining the co-integration relationships when there are more than two variables. One is the trace test, which tests the null hypothesis that there are at most r (0  r  n, where n is the number of variables) co integrating vectors. The second is the maximum Eigen value test, which tests the null hypothesis that there are r co integrating vectors against the alternative of r+1 co integrating vectors. Reimers (1992) argues that these test statistics be corrected for the number of estimated parameters to obtain satisfactory size properties in small samples. The correction is by replacing T by T-n p in the test statistic, where T is the number of observations, n is the number of variables and p is the lag length of the VAR. Our interpretation of the results is based on the test statistics with small sample correction. Read More