Out-of-Sample Exchange Rate Forecasting - Essay Example

?International Finance Table of Contents Summary 3 Appraisal 6 Reference 9 9 9 Summary Out-of-sample exchange rate forecasting when used with real-time data helps to determine the exchange rate of the country. Numerous economists used different models to determine the exchange rates which is not claimed to be appropriate due to the unavailability of real-time data which are available with the market participants. However, Taylor rule indicates that the central bank regulates the short-run nominal interest rate in accordance with the changes in output gap and inflation level. During the financial crisis in the year 2008–2009, Taylor rule of out-of-sample exchange rate forecasting had been used for predicting the exchange rate of euro or Dollar with other countries currency. Furthermore, the models have been estimated with Organization for Economic Co-operation and Development (OECD) estimates of the output gap in relation to the unemployment gap and use these two figures of the same period along with one-year-ahead forecasts for determining the inflation level in order to compare with the real economic activity. Thus, the article provides an assessment of four models with specified Taylor rule. Moreover, the performances of conventional monetary, Purchasing Power Parity (PPP), and interest rate differential model have been evaluated for comparing each of the four models (Molodtsova & Papell, 2010). Due to the unavailability of euro or dollar exchange rates till 1999 when the euro had been introduced, rolling regressions methods have been used to predict the exchange rate changing from the year 1999 along with 26 observations for each regression. However, the number of observations has been kept constant while deriving the results from the year ending 2007 with 37 predictions, all the way through to 2010, with 45 predictions. The results have further been represented through three test statistics, namely, ‘ratio of the mean squared prediction errors (MSPE) of the linear and random walk models, DMW test of Diebold and Marino (1995) along with West (1996) and the CW test of Clark and West (2006)’ with the significant values of McCracken (Molodtsova & Papell, 2010). The results derived from the Taylor rule fundamentals model reveal an attractive apparent pattern. It has been observed that along with the variables of that period and rising inflation, MSPE of the Taylor rule model is lesser compared to MSPE of the random walk model. Furthermore, utilizing the CW and DMW tests at 5% or higher level for the early predictions of year ended 2007, it has been noted that the random walk null can be discarded in favor of the Taylor rule model. Therefore, following the number of predictions increase, the strength of the rejections increased which hit the highest level in 2008. However, from the next quarter of 2008, the strength of the rejections started to decline and thus, climbed to the extreme level of financial crisis thereby, favoring the Taylor rule specifications sharply. In 2009, the ‘phoenix’ Taylor rule forecasting evolved which was discarded in favor of Taylor rule models at 1% implication level for all specifications between 2009 and 2010. Although the pattern of results has similarities with the inflation level in the beginning of 2008, the strength of the rejections is weaker. Furthermore, the results derived from forecasted variables are weaker than actual variables from that period (Molodtsova & Papell, 2010). Mark (1995) theories focused on obtaining a set of long-run fundamentals from the different models which helped in evaluating out-of-sample forecasts on the basis of the difference between the present exchange rate and its value in the long run. Another part of the theory uses the interest rate obscured by Taylor rule whereas, Molodtsova and Papell (2009) uses the variables that are utilized in Taylor rules to estimate the exchange rate predictions. Furthermore, the Taylor rule fundamentals model helps to evaluate the correlation between the exchange rate and a number of variables that occur when central banks set the interest rates on the basis of Taylor rule. In order to obtain the Taylor-rule-based forecasting equation, the interest rate differential is created by deducting the interest rate reaction function of the Euro Area from the figure determined for the United States. Moreover, on the basis of experimental research on the forward premium and deferred overshooting puzzles proposed by other economists, illustrates that an increase in the interest rate can cause constant exchange rate appreciation only if the investors either thoroughly misjudge the determination of interest rate shocks or take occasional portfolio decisions (Molodtsova & Papell, 2010). Interest rate differential model derives that the estimated change in the log exchange rate is equal to the nominal interest rate differential which can be used in the forecasting equation. An alternative Taylor rule based model has been proposed by Engel, Mark and West (2008), named Taylor rule differentials model in order to differentiate from the interest rate differentials model and Taylor rule fundamentals model. However, the most commonly used approach to evaluate the exchange rate models from the given sample is to signify a change in the nominal exchange rate as a function of its divergence from the fundamental value (Molodtsova & Papell, 2010). Thus, the Taylor rule models for forecasting out-of-sample exchange rate in the late 2008 did not prove to be effective, it was assumed that the phoenix or random walk model evolved in 2009 will be used for forecasting exchange rate in the long run. However, the Taylor rule models in 2009 and 2010 outperformed the random walk model. Furthermore, the Taylor rule models have proved to be effective at describing and recommending interest rate setting at central banks which have been evaluated in the article with the help of Taylor rule fundamentals and differentials. Moreover, the models help to accurately predict the euro/dollar exchange rate with greater support from the out-of-sample predictability than random walk benchmark. None of the models could forecast the appreciation of Dollar during the financial crisis in 2008, however the Taylor rule models gained success in predicting the exchange rate in 2009 and 2010 (Molodtsova & Papell, 2010). Appraisal The study has been focused to evaluate the effectiveness of Taylor rule models which along with other conventional models could not forecast the exchange rates of euro or dollar during the financial crisis in 2008. The study further helped to assess the appropriate model by comparing with different conventional models with Taylor models. Thus, it can be stated that it was important to conduct the research as it will help financial institutions and employers to identify the appropriate model and forecast the appreciation or depreciation of euro or dollar which has a potential effect on the business. The research further suggested that none of the other conventional models of forecasting used real-time data available with the market participants and without it the results cannot be claimed as accurate from the out-of-sample forecasting. Furthermore, to prove the suggestion, the research illustrated an example where certain authors conducted a survey using the monetary model concerning five countries for forecasting the US Dollar exchange rates using the real-time data. The research has been conducted in order to identify the appropriate model for forecasting the euro/dollar exchange rate in order to enable Fed and other central banks decide on the interest rate settlement. Since the research has been conducted to analyze the appropriate models amongst the conventional models for forecasting exchange rates, the design primarily involved collecting the data regarding exchange rate changes from the introduction year of the Euro which have been forecasted through rolling regressions. Furthermore, results have been derived from a large number of forecasts from 2007 through to 2010 in order to effectively analyze the appropriateness of forecasting models. The research has been based on evaluating the effectiveness of Taylor model which was assumed to be ineffective after the financial crisis that occurred in 2008. Therefore, to avoid potential sources of biasness, the results were derived through statistical equations in order to eliminate the chances of random errors that can influence the overall findings to other directions. Although most of the researches that are based on empirical studies use hypothetical testing to justify the findings in relation to the questions, this research did not use hypothetical questions since it is based on the evaluation of the appropriate forecasting model from the comparative analysis. However, the prime question of the research has been justified through statistical interpretation. In order to fulfill the core objective of the research, all the data collected has been calculated through statistical equation for deriving appropriate findings. Significant books and journals were reviewed for performing the statistical analysis correctly. Furthermore, electronic mediums have been used to avoid any miscalculations. Following the statistical equations, each of the variables has been clearly denoted and explained. Furthermore, detailed explanations for each of the equations have been provided for enhancing the knowledge of the reader. Last but not the least, graphical representations have been prepared for illustrating the findings more prominently. Moreover, the comparisons of each forecasting models has been demonstrated apparently to analyze the appropriate model which can be effective in forecasting the exchange rates of dollar or euro in the long run. The results that have been derived through the statistical equations are well justified in the conclusion part. The aim of the research was to prove the effectiveness of Taylor rule model in comparison to other conventional forecasting models. Therefore, the graphical representation will provide an apparent distinction between Taylor rule model and other forecasting models. Furthermore, the research indicated that during the time of financial crisis when all other conventional forecasting models could not predict the exchange rates of euro or dollar, Taylor rule model provided an approximate idea for the exchange rates to the central banks which was used to set the interest rates. Since the predictions did not yield exact results, assumptions were made about the demise of the model. However, with the findings and data interpretation analysis it has been concluded that Taylor rule model is the appropriate forecasting model for predicting the exchange rate of euro or dollar. Since the research is not based on field surveys that involve participants’ views or notions, probability of arising conflicts of interests were minimal. Generally, conflicts of interest occur when huge number of participants is involved in providing their personal perceptions regarding a particular issue. Furthermore, chances of misinterpretation or manipulation are higher in researches which collect data from the perceptions of individuals. Thus, it can be stated that as the research has been conducted with the actual data from various reliable sources, genuineness and authenticity has been maintained. Moreover, the interpretations of the data have been made through statistical equations which ensure error free results and analysis. Reference Molodtsova, T. & Papell, D. H., 2010. Phoenix Taylor Rule Exchange Rate Forecasting During the Financial Crisis. University of Houston. [Online] Available at: http://www.google.co.in/url?sa=t&rct=j&q=&esrc=s&source=web&cd=1&ved=0CCAQFjAA&url=http%3A%2F%2Fwww.aeaweb.org%2Faea%2F2011conference%2Fprogram%2Fretrieve.php%3Fpdfid%3D261&ei=o8FuUOaCC4vNrQeAooCYAw&usg=AFQjCNEXp9RZvlZK9svEBytbkyM0TIFw6g&sig2=nyUTQgwdsA_C6O9YhvJeUQ [Accessed October 05, 2012]. Read More