#### DMCA

## The role of asymmetries and regime shifts in the term structure of interest rates (2006)

Citations: | 22 - 2 self |

### Citations

2573 | Statistical Analysis of Cointegrating Vectors - Johansen - 1988 |

2521 |
Large Sample Properties of Generalized Method of Moments Estimators
- Hansen
- 1982
(Show Context)
Citation Context ...sified observations to the total number of observations used in the logit estimation. Estimates are obtained by generalized method of moments calculated by two-step nonlinear two-stage least squares (=-=Hansen 1982-=-). The optimal weighting matrix is obtained from the first step two-stage least-squares parameter estimates; the instrument set includes 12 lags of each of inflation and the output gap. Standard error... |

1711 | A simple positive, semi-definite, heteroskedasticity and autocorrelation consistent covariance matrix - Newey, West - 1987 |

1291 | Comparing Predictive Accuracy - Diebold, Mariano - 1995 |

1156 | Heteroskedasticity and autocorrelation consistent covariance matrix estimation - Andrews - 1991 |

826 |
Discretion versus policy rules in practice
- Taylor
- 1993
(Show Context)
Citation Context ...the term structure appear to be intimately related to the business cycle. Building on these findings, and on the literature relating monetary policy to indicators of the business cycle and inflation (=-=Taylor 1993-=-, 1999; Clarida, Gali, and Gertler 1998, 1999, 2000), we estimate logit models relating the probability of being in either the high– or the low–interest rate regime to appropriate economic indicators.... |

624 | Modelling Nonlinear Economic Relationships - Granger, Terasvirta - 1993 |

595 |
Maximum likelihood estimation from incomplete data via the EM algorithm
- Dempster, Laird, et al.
- 1977
(Show Context)
Citation Context ...1526 .1686 .1706 .1215 .1373 .1520 .1545 .1488 .688 .118 .275 ˜˜Pp ; yp[ ] [ ] .312 .882 .725 r(A)p .22510 Tildes denote estimated values obtained using the EM algorithm for maximum likelihood (=-=Dempster et al. 1977-=-). Figures in parentheses are asymptotic standard errors. Symbols are defined as in eq. (13). P and y denote the transition matrix andM#M the M-dimensional ergodic probabilities vector, respectively. ... |

579 | Specification Analysis of Affine Term Structure Models - Dai, Singleton - 2000 |

524 | Cointegration and tests of present value models - Campbell, Shiller - 1987 |

475 | Critical values for cointegration tests
- MacKinnon
- 1991
(Show Context)
Citation Context ...e-year eurorates, respectively;i i i i i0,t 4,t 13,t 26,t 52,t and D is the first-difference operator. The critical value at the 1% (5%) significance level is 3.446 (2.868) to three decimal places (=-=MacKinnon 1991-=-). period from February 7, 1982, to December 31, 2000, a total of 987 observations for each series.7 In our empirical work, we carried out our estimations over the period February 1982–December 1991, ... |

433 | Evaluating Density Forecasts with Applications to Financial Risk Management - Diebold, Gunther, et al. - 1998 |

422 | A Note with Quantiles of the Asymptotic Distribution of the Maximum Likelihood Cointegration Rank Test Statistics - Osterwald-Lenum - 1992 |

365 |
Modeling the Conditional Distribution of Interest Rates as a Regime-Switching Process
- Gray
- 1996
(Show Context)
Citation Context ... characterized by a nonlinear equilibrium correction model due to factors such as, for example, nonzero or asymmetric transactions costs or infrequent trading or the existence of regime shifts (e.g., =-=Gray 1996-=-; Anderson 1997; Enders and Granger 1998; Bansal and Zhou 2002; Sarno and Thornton 2003).1 In addition to this growing amount of statistical evidence, there are sound economic reasons to believe that ... |

359 |
Nonlinear Statistical Models
- Gallant
- 1987
(Show Context)
Citation Context ...imator2ĵ is given by the Andrews rule, , where is a function of and . The1/5 2ˆ ˆ ˆˆAp 2.6614[z(2)T] z(2) r j Parzen window has been used because it minimizes the mean square error of the estimator (=-=Gallant 1987-=-, 534). 1210 Journal of Business unknown. Therefore, the marginal significance levels reported below should be interpreted with caution.21 Table 7 gives detailed results of the accuracy of the forecas... |

324 | Specification Searches: Ad Hoc Inference with Non-experimental Data - Leamer - 1978 |

319 |
Exchange rates and fundamentals: evidence on long horizon predictability
- Mark
- 1995
(Show Context)
Citation Context ...t frequency zero. Diebold and Mariano show that the DM statistic is distributed as standard normal under the null hypothesis of equal forecast accuracy. Consistent with a large literature (see, e.g., =-=Mark 1995-=-), the loss differential functions we consider are either the difference between the APFE for the two models or the difference between the SPFEs. A consistent estimate of the spectral density at frequ... |

313 | Monetary policy rules in practice: some international evidence - Clarida, Gali, et al. - 1997 |

275 |
State-Space Models with Regime Switching
- Kim, Nelson
- 1999
(Show Context)
Citation Context ...ntation of an expectation-maximization (EM) algorithm for maximum likelihood estimation that yields estimates of the remaining parameters of the model (Dempster, Laird, and Rubin 1977; Hamilton 1993; =-=Kim and Nelson 1999-=-; Krolzig 1999). We now turn to a brief discussion of our data set and then to our empirical analysis. IV. Empirical Results A. Data, Unit Root Tests, and Cointegration Analysis Our data set comprises... |

211 |
Unit-root tests and asymmetric adjustment with an example using the term structure of interest rates
- Enders, Granger
- 1998
(Show Context)
Citation Context ...ear equilibrium correction model due to factors such as, for example, nonzero or asymmetric transactions costs or infrequent trading or the existence of regime shifts (e.g., Gray 1996; Anderson 1997; =-=Enders and Granger 1998-=-; Bansal and Zhou 2002; Sarno and Thornton 2003).1 In addition to this growing amount of statistical evidence, there are sound economic reasons to believe that regime shifts and asymmetries improve ou... |

176 | Regime switches in interest rates
- Ang, Bekaert, et al.
(Show Context)
Citation Context ...m p1 Dy p n(z ) GDy Py u , (12)t t i ti t1 t ip1 where and . For parsimony considerations andu ∼ NIID(0, S (z )) z p 1, 2t u t t consistent with previous research in this context (Gray 1996; =-=Ang and Bekaert 2002-=-; Bansal and Zhou 2002), we limited ourselves to discriminating between linear models and Markov-switching models allowing for only two regimes in the VECM. We applied the conventional “bottom-up” pro... |

137 | Tests of equal forecast accuracy and encompassing for nested models - Clark, McCracken - 2001 |

134 | Exchange Rates and Monetary Fundamentals: What Do We Learn From Long-Horizon Regressions - Kilian - 1999 |

132 | Some tests for parameter constancy in cointegrated VAR-models - Hansen, Johansen - 1999 |

128 | An Equilibrium Characterisation of the Term Structure - Vasicek - 1977 |

124 | Non-Linear Time Series Models in Empirical Finance - Franses, Dijk - 2000 |

122 | Term structure of interest rates with regime shifts. Journal of Finance 57:1997–2044
- Bansal, H
- 2002
(Show Context)
Citation Context ...n model due to factors such as, for example, nonzero or asymmetric transactions costs or infrequent trading or the existence of regime shifts (e.g., Gray 1996; Anderson 1997; Enders and Granger 1998; =-=Bansal and Zhou 2002-=-; Sarno and Thornton 2003).1 In addition to this growing amount of statistical evidence, there are sound economic reasons to believe that regime shifts and asymmetries improve our understanding of the... |

118 | Quadratic term structure models: theory and evidence - Ahn, Dittmar, et al. - 2002 |

113 | A Cointegration Analysis of Treasury Bill Yields", The Review of Economics and Statistics , 74
- Hall, Anderson, et al.
- 1992
(Show Context)
Citation Context ...a to be time-varyinggk,t and requires only that they be realizations of stationary stochastic processes. Even this very weak form of the EH, however, has important and clear statistical implications (=-=Hall et al. 1992-=-). To see this, note that we can rewrite equation (2) as follows: k1 m1 i i p E Di g , (3)( )k,t 1,t t 1,tj k,tk mp1 jp1 where D is the first-difference operator. Under the assumption that the... |

104 | Density Forecasting: A Survey
- Tay, Wallis
- 2000
(Show Context)
Citation Context ...rtance of evaluating the forecast accuracy of economic models on the basis of density—as opposed to point—forecasting performance (see, e.g., Diebold, Gunther, and Tay 1998; Granger and Pesaran 1999; =-=Tay and Wallis 2000-=-; Timmerman 2000). Especially when evaluating nonlinear models, which are capable of producing highly nonnormal forecast densities, it would seem appropriate to consider a model’s density forecasting ... |

103 |
Rational expectations econometric analysis of changes in regime: An investigation of the term structure of interest rates
- Hamilton
- 1988
(Show Context)
Citation Context ...he dynamic adjustment of the term structure in response to deviations from equilibrium may in fact be asymmetric (Enders and Granger 1998; Sarno and Thornton 2003) and characterized by regime shifts (=-=Hamilton 1988-=-; Gray 1996; Bansal and Zhou 2002). In this paper, we therefore develop a VECM approach that is capable of allowing for all these possibilities simultaneously. III. Asymmetric Markov-Switching Equilib... |

89 | A new measure of fit for equations with dichotomous dependent variables - Estrella - 1998 |

76 | The Out-of-Sample Success of Term Structure Models as Exchange Rate Predictors: A Step Beyond,” NBER Working Paper No - Clarida, Sarno, et al. - 2001 |

58 |
Robust Out-of-Sample Inference
- MCCRACKEN
- 2000
(Show Context)
Citation Context ...lected using the Andrews (1991) AR(1) rule.20 Several problems may arise when one uses DM statistics in small samples and takes into account parameter uncertainty (West 1996; West and McCracken 1998; =-=McCracken 2000-=-). In the present case, where we are dealing with nested competing forecasting (linear and nonlinear) models and with multi-step-ahead forecasts, the asymptotic distribution of the DM statistic is non... |

51 | A Decision Theoretic Approach to Forecast Evaluation, in (eds
- Granger, Pesaran, et al.
(Show Context)
Citation Context ...cently emphasized the importance of evaluating the forecast accuracy of economic models on the basis of density—as opposed to point—forecasting performance (see, e.g., Diebold, Gunther, and Tay 1998; =-=Granger and Pesaran 1999-=-; Tay and Wallis 2000; Timmerman 2000). Especially when evaluating nonlinear models, which are capable of producing highly nonnormal forecast densities, it would seem appropriate to consider a model’s... |

51 |
Markov-Switching Vector Autoregressions
- Krolzig
- 1997
(Show Context)
Citation Context ...procedure essentially extends Hamilton’s (1988, 1989) Markov-switching regime framework to nonstationary systems, allowing us to apply it to cointegrated vector autoregressive (VAR) and VECM systems (=-=Krolzig 1997-=-, 1999). Consider the following M-regime pth-order Markov-switching vector autoregression (MS(M)-VAR(p)), which allows for regime shifts in the intercept term:3 p y p n(z ) P y , (4)t t i ti t ... |

50 |
Estimation and Testing of Cointegrated Systems by an AutoregressiveApproximation,EconometricTheory,8,1-27
- Saikkonen
- 1992
(Show Context)
Citation Context ...and to estimate the matrix of cointegrating parameters b. In fact, use of the conventional Johansen procedure is legitimate in the first stage without modeling the Markovian regime shifts explicitly (=-=Saikkonen 1992-=-; Saikkonen and Luukkonen 1997). The second stage then consists of the implementation of an expectation-maximization (EM) algorithm for maximum likelihood estimation that yields estimates of the remai... |

44 |
Transaction Costs and Nonlinear Adjustment Towards Equilibrium
- Anderson
- 1997
(Show Context)
Citation Context ...zed by a nonlinear equilibrium correction model due to factors such as, for example, nonzero or asymmetric transactions costs or infrequent trading or the existence of regime shifts (e.g., Gray 1996; =-=Anderson 1997-=-; Enders and Granger 1998; Bansal and Zhou 2002; Sarno and Thornton 2003).1 In addition to this growing amount of statistical evidence, there are sound economic reasons to believe that regime shifts a... |

42 | The Dynamic Relationship Between the Federal Funds Rate and the Treasury Bill Rate: An Empirical Investigation
- Sarno, Thornton
- 2003
(Show Context)
Citation Context ... such as, for example, nonzero or asymmetric transactions costs or infrequent trading or the existence of regime shifts (e.g., Gray 1996; Anderson 1997; Enders and Granger 1998; Bansal and Zhou 2002; =-=Sarno and Thornton 2003-=-).1 In addition to this growing amount of statistical evidence, there are sound economic reasons to believe that regime shifts and asymmetries improve our understanding of the behavior of the entire y... |

42 |
Asymptotic Inference about Predictive Ability” Econometrica 64: 1067-1084 Data Appendix Unless otherwise stated, we use seasonally-adjusted quarterly data from the IMF International Financial Statistics ranging from the second quarter of 1973 to the last
- West
- 1996
(Show Context)
Citation Context ...e optimal truncation lag has been selected using the Andrews (1991) AR(1) rule.20 Several problems may arise when one uses DM statistics in small samples and takes into account parameter uncertainty (=-=West 1996-=-; West and McCracken 1998; McCracken 2000). In the present case, where we are dealing with nested competing forecasting (linear and nonlinear) models and with multi-step-ahead forecasts, the asymptoti... |

41 | Intercept corrections and structural change - Clements, Hendry - 1996 |

34 | Some difficulties of interpretation encountered in the application of the chi-square test - Berkson - 1938 |

30 |
Testing Cointegration in Infinite Order Vector Autoregressive Processes
- Saikkonen, Luukkonen
- 1997
(Show Context)
Citation Context ...the matrix of cointegrating parameters b. In fact, use of the conventional Johansen procedure is legitimate in the first stage without modeling the Markovian regime shifts explicitly (Saikkonen 1992; =-=Saikkonen and Luukkonen 1997-=-). The second stage then consists of the implementation of an expectation-maximization (EM) algorithm for maximum likelihood estimation that yields estimates of the remaining parameters of the model (... |

27 | Modeling and Forecasting Stock Returns: Exploiting - Sarno, Valente |

26 | 2003, An evaluation of multi-factor CIR models using LIBOR, swap rates, and cap and swaption prices - Jagannathan, Kaplin, et al. |

22 | The Term Structure of Euromarket Interest Rates - Campbell, Clarida - 1987 |

16 | Nonlinear Dynamic Structures. Econometrica 61: 871–907 - Gallant, Rossi, et al. - 1993 |

14 | A Class of Non-Linear Time Series Models - Karlsen - 1990 |

13 |
Regime shifts, risk premiums in the term structure, and the business cycle
- Bansal, Tauchen, et al.
- 2004
(Show Context)
Citation Context ...an level.19 As we have noted, the explanatory variables we have used are in line not only with recent empirical research on the regime-shifting behavior of interest rates (e.g., Bansal and Zhou 2002; =-=Bansal et al. 2004-=-) but also with the very large literature on monetary policy rules—so-called Taylor rules (see, e.g., Taylor [1993, 1999] and Clarida, Gali, and Gertler [1998, 2000] and the references therein). The T... |

12 |
Nonlinear permanent–temporary decompositions in macroeconomics and finance
- Clarida, Taylor
- 2003
(Show Context)
Citation Context ...and .′Pp ab u ∼ NIID(0, S (z )) z {1, … , M}t u t t In order to take into account the empirical evidence indicating that interest rates display asymmetric adjustment (e.g., Enders and Granger 1998; =-=Clarida and Taylor 2003-=-; Sarno and Thornton 2003), we allow the MSIH-VECM (6) 5. There is a slight shift in notation here from that employed in Sec. II in that representsi0,t the spot-next rate, which is an overnight rate r... |

10 |
Co-integration and Equilibrium Correction Representation
- Engle, Granger
- 1987
(Show Context)
Citation Context ...ing H yields, each given by the stationary spread for . Moreover, given that cointegration between a seti i kp 2, … , Hk,t 1,t of variables implies, according to the Granger representation theorem (=-=Engle and Granger 1987-=-), the existence of a statistical representation for the yields in the form of a vector equilibrium correction model (VECM), this provides a rationale for modeling the dynamic interrelationship betwee... |

7 | Density Forecasting in Economics and Finance: Editorial - Timmermann - 2000 |

6 |
Cointegration, long-run movements, and long-horizon forecasting
- Stock
- 1997
(Show Context)
Citation Context ...We follow the empirical literature because very persistent series with a root at least very close (if not equal) to unity are better approximated by I(1) processes than by stationary ones (see, e.g., =-=Stock 1997-=-). Asymmetries and Regime Shifts 1201 We then employed the Johansen (1988, 1991) maximum likelihood procedure in a VAR for and an unrestricted constant′y p [i , i , i , i , i ]t 0,t 4,t 13,t 26,t 52,t... |

4 |
Residual Based Stochastic Predictors and Estimation in Nonlinear Models
- Brown, Mariano
- 1984
(Show Context)
Citation Context ... use a large number of simulations, by the law of large numbers this procedure should produce results virtually identical to those that would result from calculating the exact forecasts analytically (=-=Brown and Mariano 1984-=-, 1989; Gallant, Rossi, and Tauchen 1993). Forecast accuracy is evaluated using absolute and square error criteria (see Bansal and Zhou 2002), specifically, the average absolute cross-sectional pricin... |

3 |
Developments in the study of cointegrated variables. Oxford Bulletin of Economics and Statistics 48:213–28
- Granger
- 1986
(Show Context)
Citation Context ... 5 _ …p p p M1 M2 MM where for .…p p 1 p p i {1, … , M}iM i1 i,M1 1198 Journal of Business (the deviation from the long-run equilibrium) is measured by the stationary stochastic process (=-=Granger 1986-=-; Engle and Granger 1987). If in-′h p b yt t deed there is cointegration, the cointegrated MS-VAR (4) implies a Markovswitching vector equilibrium correction model or MS-VECM of the form p1 Dy p n(z ... |

1 | Regime Shifts 1223 In Statistics and finance: An interface - Asymmetries |

1 |
Modeling the yield curve. Economic Journal 102:524–37
- Taylor
- 1992
(Show Context)
Citation Context ...m structure and on building equilibrium correction models of the dynamic interaction between interest rates at different maturities (e.g., Campbell and Shiller 1991; Hall, Anderson, and Granger 1992; =-=Taylor 1992-=-). Recently, an interesting strand of this literature has developed that allows for asymmetric or nonlinear adjustment toward equilibrium in modeling interest rate movements. In this work, researchers... |