| Summary: | 碩士 === 雲林科技大學 === 財務金融系碩士班 === 96 === This paper uses Granger causality test, vector autoregression, impulse response function and forecast error variance decomposition to test lead-lag relationship and price discover among the finance insurance index, finance future and finance ETF. We use daily closing in prices and return data during 2/1/2007~29/2/2008. The conclusion can be summarized as follows: (1) According to Granger causality model, there is a feedback effect between finance index with finance future and finance future with finance ETF. (2) Base on the VAR model, finance index had better ability about price discover.(3) In the aspect of the impulse response function, finance index would be affected by itself and not by other market. it could also be found in the variance decomposition, finance index had the better account of forecast error variance decomposition. In other words, finance index was the leading indicator of price discover.
|
|---|
