| 摘要: | 由於投資人是在無法準確預知未來的情況下進行股票買賣行為，如何依據現有的數據資料做出對股價未來漲跌最接近的預測是投資人所希望的。在眾多的數據資料中，計算簡便的本益比(Price-Earnings Ratio)就成為投資人在買賣股票時常用來判斷買賣股票的指標之一。本益比為評比股票價格對盈餘的比值，假定其它條件相同的情況下，低本益比的股票即代表是低價的股票，高本益比的股票則代表是高價的股票。平均而言，若能以越低的股價購入股票，而不是以高的價格購買股票，就能以越低的成本獲取比高價時購入成本相對較高的報酬。過去的研究也指出善用本益比資訊是可以得到較好的股票投資報酬，因此對投資人而言本益比是有其重要性。 本文嘗試以本益比來探討後續不同期間本益比對股票報酬的影響，瞭解期間內股票報酬是呈現慣性持續或是報酬反轉。本益比的計算則使用三種不同的盈餘數字，分別是過去、即時、未來三種盈餘，而得到三種型態的本益比。更進一步瞭解使用過去、即時、未來三種盈餘所計算之本益比在後續期間一週、二週、一個月、二個月、三個月、六個月的報酬情形。探討在前述的時間點時，股票報酬是呈現慣性持續或者是有報酬反轉的情況。本論文主要發現有，第一，使用和評價無關聯性的過去盈餘所計算之本益比缺乏對股票價格高估或低估的指標功能，本益比和隨後的報酬沒有統計上的關聯。第二，使用即時的盈餘所計算之本益比已具有對股票價格高估、低估的指示作用，結果顯示從一週後開始至三個月後的時間都持續出現了股票報酬反轉的現象，且在三個月後報酬反轉程度有減緩的現象。第三，使用具有未來性質的盈餘所計算之本益比，應具有最佳的股價高估或低估的指示功能，結果發現至第一週為止沒有報酬反轉或慣性持續的現象，第二週後則開始出現報酬反轉的現象並且一直持續到後續六個月的時間。其報酬反轉的時間是比使用即時盈餘計算之本益比結果多了一倍的修正時間。這是因為對於過高價格的股票做出了後續時間較長的持續股價修正。 另外也發現低本益比的股票在後續時期的報酬大多優於高本益比的股票，意謂著股價對盈餘而言是合理、甚至是低估的價格，沒有高估股價的情況，所以才能在後續時期獲取較高的報酬。因此建議投資人應以低本益比的股票為投資時優先考慮的目標。
Past research has found the price to earnings ratios (P/E) negatively related to subsequent stock returns. Such return reversal phenomenon suggests that P/E might be a display of stock market overreactions to changes in fundamentals. That is, high P/E stocks earn lower or even negative returns because the market initially overprices the stocks. Similarly, an underpricing of the low P/E stocks is responsible for subsequent higher returns. Therefore, P/E does not exclusively reflect differences in fundamentals but to some degree measures stock valuation errors caused by investor overreaction. This thesis attempts to examine the returns of varying P/E stocks during different lengths of subsequent time period, from one week, two weeks, one month, three months, to six months, to explore the overreaction implication of P/E. If a negative relationship is confirmed between P/E and subsequent returns, then investors may use P/E as criteria to form arbitrage portfolio and obtain excess returns. We also calculate P/E based on three types of earnings in relation to the time point of price samples: the past earnings (Ep) from the previous four quarters, the current earnings (Ec) combining the previous two quarters and the forthcoming two quarter earnings, and the future earnings (Ef) with the four subsequent quarter earnings. Since prices are forward looking, P/Ef may be the most valid measure among the three P/E s. However, Ep and Ec are currently available or easy to forecast. If P/Ep and P/Ec are also negatively related to future returns, then they can be low-cost useful portfolio selection criteria for investors. The major findings include first, P/Ep is not significantly related to future returns so that P/E based on past earnings is not useful for portfolio selections. Second, P/Ec is moderately negatively related to future stock returns and hence is a good indicator of market overreactions. Third, P/Ef is significantly negatively related to future stock returns over all six lengths of measurement period. This suggests that high (or low) P/Ef stocks with initial overpricing (or underpricing) undergo return reversal or price correction beginning the first week and continuing for as long as six months. |
