City University of Hong Kong
DSpace
 

CityU Institutional Repository >
3_CityU Electronic Theses and Dissertations >
ETD - Dept. of Economics and Finance  >
EF - Doctor of Philosophy  >

Please use this identifier to cite or link to this item: http://hdl.handle.net/2031/6482

Title: Three essays on market-based forecasting models
Other Titles: San pian ji yu shi chang yu ce mo xing de lun wen
三篇基於市場預測模型的論文
Authors: Wu, Liang ( 武亮)
Department: Department of Economics and Finance
Degree: Doctor of Philosophy
Issue Date: 2011
Publisher: City University of Hong Kong
Subjects: Economic forecasting.
Notes: CityU Call Number: HB3730 .W8 2011
v, 153 leaves : ill. 30 cm.
Thesis (Ph.D.)--City University of Hong Kong, 2011.
Includes bibliographical references (leaves 134-153)
Type: thesis
Abstract: This thesis consists of three essays, all of which attempt to explore the forecast ability of market expectations on future macroeconomic variables. To obtain pure market view of the future economy, innovative methods are employed to filter out unwanted noises embedded in the observed data. Conventional models including time series models and theory-based economic models are used as benchmarks for out-of-sample comparison. In the first essay, we analyze the ability of Federal funds futures in predicting the monetary policy stance of the FOMC. Gürkaynak et al. (2007) showed that Federal funds futures provide superior forecasts than other financial instruments and sophisticated time series models out to six months, but the expected monetary policy rate in their analysis is contaminated by both the FOMC scheduled meeting date and risk premia embedded in Federal funds futures. We provide two approaches to clean the noise. The first one is to correct for the impacts of the FOMC scheduled meeting date while the second approach is to correct for risk premia in futures contracts. We show that our adjusted expectation of the Federal funds rate does a better job in predicting the future monetary policy rate than directly using Fed funds futures. Our approaches also beat the Taylor rule model (1993) with real time data. Our analysis suggests that participants in the Federal Funds futures market make accurate predictions of Federal funds rate. In essay two, we attempt to identify the business cycle peak turning point of the 2007-2008 U.S. recession using Google Trends data. Official recession announcements by the NBER always come with substantial lags after recessions occur. Econometric models such as the multivariate Markov-Switching model by Chauvet and Hamilton (2005) or the Harding-Pagan Algorithm (2002) made significant improvements over the NBER announcements in terms of timing, but the signals are still 5-6 months later than the actual turning point. We found that the "recession" search query index in Google experienced a significant jump in December 2007 at the onset (the turning point) of the 2007-2008 U.S. recession. Also, we found that the "recession" and "mortgage" search query index have abnormally high correlation between December 2007 and February 2008, whereas this correlation was low in other periods. The "recession" search volume is made up of online activities of web surfers who are also participants of the economy. Our study suggests that a large group of people had already started to worry about the economy at the time recession occurred. Their discernment is more prompt than econometric models and the NBER. In essay three, we attempt to predict China's Copper and Aluminum net imports using the commodity futures contracts traded in SHFE and LME, respectively. Copper and Aluminum are two most important based metals for industrial production and they have special meanings for China economy due to the economic growth structure of China. We show that the monthly average futures spreads for Copper and Aluminum have significant predictive power for net imports of these two commodities in the subsequent two months. The predictability can be attributed to the commodity arbitrageurs taking advantages of profitable opportunities when the futures prices spread in SHFE and LME exceeds the no-arbitrage band. SOE related parties are particularly active in carrying out commodity arbitrage as it requires credit approval from commercial banks and the possession of large commodity inventory in stock, both conditions are extremely difficult to be met by private retail investors in China.
Online Catalog Link: http://lib.cityu.edu.hk/record=b4085716
Appears in Collections:EF - Doctor of Philosophy

Files in This Item:

File Description SizeFormat
abstract.html132 BHTMLView/Open
fulltext.html132 BHTMLView/Open

Items in CityU IR are protected by copyright, with all rights reserved, unless otherwise indicated.

 

Valid XHTML 1.0!
DSpace Software © 2013 CityU Library - Send feedback to Library Systems
Privacy Policy · Copyright · Disclaimer