臺灣50 ETF之期貨避險績效

The Futures Hedging Performance of Taiwan Top 50 ETF

自2003年6月推出元大寶來臺灣卓越50證券投資信託基金（Yuanta／P-shares Taiwan Top 50 ETF，簡稱臺灣50 ETF）以來，臺灣指數型股票基金（Exchange Traded Fund，簡稱ETF）市場呈現蓬勃發展。在2008年臺灣證券交易所藉由推行ETF流動量提供者制度（Liquidity Provider Program）來確保交易價格可以更貼近基金之淨值。在此制度下，如何針對ETF一籃子股票部位進行市場風險管理勢將為參與證券商必須面臨的重要議題。本文藉由臺灣50ETF標的之期貨避險分析，探討採用各種不同期貨合約情況下之避險績效。雖然期貨合約能協助ETF參與證券商管理一籃子股票所面臨的價格風險，但採用何種避險模式可以產生較佳的避險效果亦將為參與證券商所關心。除了傳統的靜態避險比率外，本文亦採用不同動態避險模式來估計最小變異避險比率，藉此提供市場參與者從事風險管理時之參考依據。實證結果顯示，臺灣50 ETF與期貨部位之風險及相關性均呈現與時變動（Time-Varying）特徵；特別的是當市場呈現巨幅變動的情況下，本研究發現採用一般化正交GARCH（Generalized Orthogonal GARCH，GO-GARCH）模型可以提供較佳的樣本外避險績效。

This paper empirically investigates the hedging performance of the Yuanta/P-shares Taiwan Top 50 ETF (Taiwan Top 50 ETF) listed in Taiwan Stock Exchange, with the uses of highly correlated futures contracts listed in both Taiwan Futures Exchange and Singapore Exchange. Due to high costs for ETF creation/redemption, a futures hedge provides the market makers a cheaper way to reduce the market risk of their basket inventories. Numerous econometric hedging models, including the static regression and multivariate generalized autoregressive conditionally heteroskedastic (GARCH), are adopted for estimating the minimum-variance hedge ratios, and the performance of the corresponding hedged portfolios by the different models are then compared in terms of risk reduction and expected utility standpoints. Using daily data running from June 2003 until December 2013, the empirical performance of the hedges are compared for both in-sample and out-of-sample periods. Since the data sets cover the recent Subprime Mortgage Crisis, the ETF hedging performance during that period is also examined. The results suggest that the joint distributions of Taiwan Top 50 ETF and futures are time-varying; and it is noteworthy that the GO-GARCH (generalized orthogonal GARCH) model provides a better out-of-sample hedging performance.