我們透過對S&P500指數報酬率之波動度建模，利用總體經濟變數因子與情緒因子的代理變數將波動度解構成兩組不同力量驅使的波動度。再藉著比較理性因子與情緒因子所驅動出來的波動度大小之交疊，實證Lo (2004) 提出之適應性市場假說再美國股票市場的適用 性。我們不僅發現S&P500指數報酬率的波動度確實會隨著市場氛圍的轉變而由不同力量驅使，市場情緒高張的時期還會隨時間推移越來越多。在依據市場效率無法藉由新資訊獲得超額報酬的假設，我們建立適應性市場投資組合，當市場呈現理性有效率時，持有三個月到期之政府公債；反之當市場情緒高張時，買進S&P500指數以賺取報酬。根據這樣的策略，長期來看可以獲得超過買進持有指數3.5倍的累積報酬。不僅如此，我們還發現這樣的超額報酬來不被Fama and French (1992) 的三因子模型的公司規模與帳市值比所解釋。我們參考 Lo (2004) 從宏觀經濟的角度實證AMH，並且應用市場適應性建構出投資組合以賺取報酬。;To decompose the volatility of S&P500 index return, we used macroeconomic factors as rational power and sentimental factors as sentimental power. By comparing the crossing of both powers, we documented the fact that the adaptive market hypothesis (AMH), proposed by Lo (2004), is suitable for the US stock market.We found out that the atmosphere of the stock market not only changes periodically, but also that the sentimental power conquers the market more frequently over time. Based on the assumption of e cient market hypothesis (EMH), we formed an investing strategy to take advantages from the adaptive phenomena of stock market. We invested in 3-month T-Bills when the market appeared rational, inversely, we invested in S&P500 index when the market appeared sentimental. The cumulative return of that strategy is 3.5 times higher than the cumulative return of the buy-and-hold index strategy. Moreover, we found out that these outstanding returns could not be explained by the size e ect and the BM ratio e ect proposed by Fama and French (1992). The main contribution of this paper is that we proved AMH macro-economically by decomposing volatility and proposed a useful strategy based on our decomposition.
