Structural TAR and STR models

The structural TAR and STR models that I have employed in this book resemble their purely time series counterparts with two major differences. First, they rely on a structural variable, and that structural variable is the interest rate differential between the countries. Second, regimes are provoked not in a self-exciting manner but by an external transition vaiiable. The transition vaiiable that I have used for the structural models is the standard deviation of the exchange rates.

If I have to explain the reason why I have chosen the standard deviation of the exchange rates as the transition variable, I caimot think of a paragraph explaining my reason better than the following one that I have taken from the 78th armual report of the Bank for International Settlements (BIS). To be more precise, on 30 June 2008, BIS announced its 78th Annual Report for the financial year that began on 1 Apxil 2007 and ended on 31 March 2008. In

Table 6.4 MSPE and MAE values for the estimation errors of the structural TAR and STR models

USD/JPY

USD/AUD

MSPE TAR

5.903264

11.674726

MAE_TAR

1.648353

1.995278

MSPE STR

5.637235

12.182258

MAE_STR

1.684697

2.932346

the relevant chapter of the report regarding the foreign exchange markets, the following lines were noteworthy.[1]

Foreign exchange markets experienced a substantial increase in volatility in August 2007. This marked an important change in the factors driving market developments. Prior to August, historically low volatility and large interest rate differentials had undeipinned cross-border capital flows that put downward pressure on funding currencies . . . , and supported high- yielding currencies. [As a result of die heightened volatility] diere was a substantial reassessment of expected monetary policy actions as the dimensions of the problems in financial markets became more apparent, hi this environment, factors such as expected growth differentials, which have an important bearing on the future path of monetary policy, became more of a focal point for market sentiment than the prevailing level of interest rates.

As you must have noticed, there was an implicit assumption at the back of the mind of the analyst as these lines were written: The timing of the transitions from one regime to another is determined by the level of volatility. Upon reading these lines, I have decided to use the standard deviations of the exchange rates as a gauge of volatility in the foreign exchange rate markets. As was done in the previous section, I again used Eviews for estimating the structural TAR and STR models. The MSPE and MAE values that I have calculated are shown in Table 6.4.

  • [1] Tins excerpt appeared at the beginning of the 5th chapter about the foreign exchange ratesbeginning on the 75th page of the 7$th Annual Report of the BIS. Following URL providesthe World Wide Web lnik to these lines, https://www.bis.org/publ/arpdf/ar2008e5.pdf
 
Source
< Prev   CONTENTS   Source   Next >