本研究分別以總風險、系統風險和獨有風險當作衡量風險的變數，根據風險變數分類產業，產業風險的計算方式分別採簡單平均加權和市值加權，發現台灣產業間有低波動異常現象，低風險產業的報酬高於高風險產業的報酬，再使用一個月觀察期與一個月持有期建構產業低波動投資組合，發現產業低波動投組報酬的標準差有市場預警的能力，並且使用簡單平均加權計算產業風險的市場預警能力優於使用市值加權計算產業風險的市場預警能力，最後使用此指標建構市場擇時策略，策略報酬優於買進持有大盤。 ;In the financial theory, we know that there is a positive trade-off relationship between return and risk, which is called "high risk, high return". But some researches have found that return of low-volatility portfolio are higher than return of high-volatility portfolio in some markets. This phenomenon is called low volatility anomaly. Subsequent researches also found that return of low-volatility portfolio constructed by low-volatility anomalies have market warning effects. This thesis uses total risk, system risk and idiosyncratic risk as risk variables, and sorts industries by risk variables. The calculation method of industrial risk are equal weighted and capitalization weighted. The result shows that there is industrial low volatility abnormal phenomenon in Taiwan. The return of low-risk industries is higher than return of high-risk industries. This thesis uses one-month formation period and holding period to construct industrial low volatility portfolio. The result shows that the standard deviation of portfolio return have market warning effect. The market warning effect of using equal weighted to calculate industrial risk is better than using capitalization weighted to calculate industrial risk. Finally, this thesis uses the indicator to construct a market timing strategy. The return of timing strategy is better buy and hold index.
