Using the information contained in the limit order book to predict volatility
| dc.contributor.author | Oz, Damla | |
| dc.contributor.examiningcommittee | Coyle, Barry (Agribusiness and Agricultural Economics) | en_US |
| dc.contributor.examiningcommittee | Arzandeh, Mehdi (Lakehead University) | en_US |
| dc.contributor.supervisor | Frank, Julieta (Agribusiness and Agricultural Economics) | en_US |
| dc.date.accessioned | 2022-02-03T15:50:52Z | |
| dc.date.available | 2022-02-03T15:50:52Z | |
| dc.date.copyright | 2022-01-20 | |
| dc.date.issued | 2022 | en_US |
| dc.date.submitted | 2022-01-20T13:40:14Z | en_US |
| dc.degree.discipline | Agribusiness and Agricultural Economics | en_US |
| dc.degree.level | Master of Science (M.Sc.) | en_US |
| dc.description.abstract | Price volatility is a research area of much interest in agricultural commodity markets. Most studies in the extant literature have used daily data for volatility estimation and forecasting, which was aligned with the traditional pit trading system. As markets switched to electronic trading, new price dynamics came into place. The limit order book (LOB) of an exchange which contains all the buy and sell limit orders, has been found to contain information that may be useful to explain volatility. However, there are few studies in agricultural commodity markets using intraday data from the LOB. The aim of this study is to investigate the effect of the price impact of buy and sell incoming orders on volatility, and to assess if those price impacts can be used to improve volatility forecasts. For this purpose, we estimate a vector error correction model with the best bid and ask quotes and three levels of depth and use impulse response functions to estimate the daily permanent price impact series. We consider two scenarios based on order placement of the incoming order in the LOB, at the market and behind the market. We then incorporate the estimated price impact series as exogenous variables into a traditional GARCH model and compare the accuracy of volatility estimations for an out-of-sample period. The forecasts obtained from the augmented GARCH models are statistically more accurate than those obtained from the traditional GARCH model. Our findings show that the LOB in the wheat market contains useful information to produce more accurate volatility forecasts, and our findings are in line with other studies on non-agricultural markets. | en_US |
| dc.description.note | February 2022 | en_US |
| dc.identifier.uri | http://hdl.handle.net/1993/36261 | |
| dc.language.iso | eng | en_US |
| dc.rights | open access | en_US |
| dc.subject | Limit Order Book | en_US |
| dc.subject | Price Volatility | en_US |
| dc.subject | GARCH | en_US |
| dc.title | Using the information contained in the limit order book to predict volatility | en_US |
| dc.type | master thesis | en_US |