CityU Institutional Repository >
CityU Electronic Theses and Dissertations >
ETD - Dept. of Management Sciences >
MS - Master of Philosophy >
Please use this identifier to cite or link to this item:
|Title: ||Range-based forecasting models for stock and energy prices|
|Other Titles: ||Ji yu gu piao ji neng yuan jia ge ji cha de yu ce mo xing|
|Authors: ||Kwok, Tsz Kin (郭子健)|
|Department: ||Department of Management Sciences|
|Degree: ||Master of Philosophy|
|Issue Date: ||2009|
|Publisher: ||City University of Hong Kong|
|Subjects: ||Stock price forecasting.|
Petroleum products -- Prices -- Forecasting.
Prices -- Forecasting.
|Notes: ||CityU Call Number: HG4637 .K85 2009|
vii, 91 leaves : ill. 30 cm.
Thesis (M.Phil.)--City University of Hong Kong, 2009.
Includes bibliographical references (leaves 85-91)
|Abstract: ||Price range has been conceived of as a more efficient volatility estimator than the traditional daily close to close stock returns data. To further investigate the information contained in the price range, this thesis is divided into two separate but related studies concerning trading strategies in the stock market, and a study of the range and corresponding high and low extreme values in the crude oil market.
The first study examines the trading strategy recently proposed by Cheung, Cheung, He and Wan (2007), which is based on price range and the features of Callable Bull and Bear Contracts (CBBC). The result of their empirical study is very encouraging when the market sampling is conducted during a “strong bull” period. In this study, we replicate the above proposed trading strategy in a strong bull and bear market period using a higher explanatory power model, an augmented vector error correction model, to conduct forecasts. We further divide the out-of-sample trading period into two, and form three different scenarios: the bull and bear market period, the bull market period, and the bear market period. Using data from the Hang Seng Index exchange-traded fund (ETF2833) as the trading instrument, it is shown that the trading strategy is only profitable in the strong bull market period and might incur enormous losses in a fluctuating market. Thus, we do not recommend using the proposed trading strategy.
The second study extends the use of the price range and its corresponding high and low in forecasting the West Texas Intermediate (WTI) spot crude oil price. A transfer function (TF) model is suggested as an alternative model to a vector error correction model (VECM), which is a conventional modeling technique to forecast the daily highs and lows, when the highs and lows are cointegrated. The in-sample analysis shows that the dynamic relationship between the highs/lows and the ranges can explain a significant portion of the high/low oil price movements. In addition we evaluate the range forecasts obtained from the TF models by comparing them with those derived from other models in the out-of-sample contest, i.e., the random walk model (RWM), the autoregressive integrated moving average (ARIMA) model, the VECM, and the VECM with (1,-1) restriction on co-integrating vector (VECM-R). Our finding is that the range forecasts from the TF model conspicuously outperform the ARIMA model and the RW walk model, and has the same forecasting ability as the VECM and the VECM-R in all time-horizons that are examined.|
|Online Catalog Link: ||http://lib.cityu.edu.hk/record=b2374869|
|Appears in Collections:||MS - Master of Philosophy |
Items in CityU IR are protected by copyright, with all rights reserved, unless otherwise indicated.