東華大學圖書館 |

語系: 繁體中文

說明(常見問題)

回圖書館首頁

手機版館藏查詢

登入

回首頁

切換: 標籤 | MARC模式 | ISBD

Econometric inference of truncated r...

Goldman, Elena.

FindBook

Google Book

Amazon

博客來

Econometric inference of truncated regression models.

紀錄類型:	書目-語言資料,印刷品 : Monograph/item
正題名/作者:	Econometric inference of truncated regression models./
作者:	Goldman, Elena.
面頁冊數:	107 p.
附註:	Director: Hiroki Tsurumi.
Contained By:	Dissertation Abstracts International63-10A.
標題:	Economics, Finance. -
電子資源:	http://pqdd.sinica.edu.tw/twdaoapp/servlet/advanced?query=3066710
ISBN:	049386105X

Econometric inference of truncated regression models.
Goldman, Elena.

Econometric inference of truncated regression models. - 107 p.

Director: Hiroki Tsurumi.

Thesis (Ph.D.)--Rutgers The State University of New Jersey - New Brunswick, 2002.

In this dissertation we study the properties of doubly-truncated time series constrained within upper and lower bounds. In Chapter 1 we describe maximum likelihood estimation and derive several asymptotic properties of doubly-truncated time series that are different from untruncated time series. We prove that the doubly-truncated random walk converges to a uniform distribution with support between the bounds of truncation. We extend this result for any unit root process with an ARMA error term following a stationary ergodic process. We show that, unlike the untruncated random walk model, the unit root coefficient estimated under double truncation converges to a normal variable. Finally, we apply a constrained maximum likelihood procedure for testing the efficient market hypothesis under the classical gold standard, 1890–1906.

ISBN: 049386105XSubjects--Topical Terms:

626650
Economics, Finance.

Econometric inference of truncated regression models.
LDR:03360nam 2200313 a 45 001 933006
005 20110505
008 110505s2002 eng d
020 $a 049386105X
035 $a (UnM)AAI3066710
035 $a AAI3066710
040 $a UnM $c UnM
100 1 $a Goldman, Elena. $3 1256746
245 1 0 $a Econometric inference of truncated regression models.
300 $a 107 p.
500 $a Director: Hiroki Tsurumi.
500 $a Source: Dissertation Abstracts International, Volume: 63-10, Section: A, page: 3645.
502 $a Thesis (Ph.D.)--Rutgers The State University of New Jersey - New Brunswick, 2002.
520 $a In this dissertation we study the properties of doubly-truncated time series constrained within upper and lower bounds. In Chapter 1 we describe maximum likelihood estimation and derive several asymptotic properties of doubly-truncated time series that are different from untruncated time series. We prove that the doubly-truncated random walk converges to a uniform distribution with support between the bounds of truncation. We extend this result for any unit root process with an ARMA error term following a stationary ergodic process. We show that, unlike the untruncated random walk model, the unit root coefficient estimated under double truncation converges to a normal variable. Finally, we apply a constrained maximum likelihood procedure for testing the efficient market hypothesis under the classical gold standard, 1890–1906.
520 $a In Chapters 2 and 3 we suggest the Bayesian approach for estimation. We propose new Markov Chain Monte Carlo (MCMC) algorithms for estimation of doubly-truncated ARMA(<italic>p, q</italic>) and ARMA(<italic>p, q</italic>)-GARCH(<italic> r, s</italic>) models. Compared to algorithms for untruncated models the key feature of our algorithms is the random-walk Markov chain. We show that the usual algorithms for untruncated models (e.g. Chib and Greenberg (1994), Nakatsuma (2000)) do not work for doubly-truncated regression models. We use new convergence diagnostics with a simultaneous test of mean and covariance stationarity of the chain.
520 $a In Chapter 4 we apply MCMC algorithms to test the efficient market hypothesis for the gold standard exchange rates, tests of purchasing power parity and interest rate parity for the period of the gold standard in Russia, the US and the UK. We account for the gold points, autocorrelation and changing volatility in estimation. We find that before 1905 Russian gold standard exchange rates are weak-form efficient, but the same result does not hold after 1905 due to the Russo-Japanese war and the 1905 revolution. We find that the US dollar-UK pound exchange rate is weak-form efficient for 1890–1906. Interest rate parity for US-UK is supported, while purchasing power parity is rejected. We show that the key estimated parameters and choice of model specification are sensitive to the bounds of truncation. We conclude that incorporating bounds in estimation is essential, while ignoring them often leads to wrong results.
590 $a School code: 0190.
650 4 $a Economics, Finance. $3 626650
650 4 $a Economics, General. $3 1017424
650 4 $a Economics, History. $3 1017418
690 $a 0501
690 $a 0508
690 $a 0509
710 2 0 $a Rutgers The State University of New Jersey - New Brunswick. $3 1017590
773 0 $t Dissertation Abstracts International $g 63-10A.
790 $a 0190
790 1 0 $a Tsurumi, Hiroki, $e advisor
791 $a Ph.D.
792 $a 2002
856 4 0 $u http://pqdd.sinica.edu.tw/twdaoapp/servlet/advanced?query=3066710