合併套利報酬分析

Abstract

本研究欲藉由套利操作以獲取合併與收購案中收購價和目標公司股價的差距，樣本來源為台灣上市櫃公司於2000年至2006年底間宣告以換股或是部份現金部份股份為對價之合併與收購案例，包括成功及撤銷案例。本研究除了模擬典型合併套利操作，並發展動態套利策略。依照動態操作套利策略，套利者依據交易期間目標公司股價和收購價的折價空間之放大縮小，進而再次購買或是提早獲利了結。 套利投資報酬乃模擬套利者建立合併套利投資組合所得之套利報酬的時間序列。欲檢驗套利者是否可於合併套利操作中獲取超額報酬，我們選取市場模型以及Fama and French 三因子模型做為指標。另外，根據美國學者研究結果，合併套利並非是市場中立的操作策略，合併案例失敗的機率隨著景氣變差而升高，導致市場報酬和套利報酬的策略在景氣衰退嚴重時具有正向相關。然而，澳洲的學者並未發現這樣的事實。因此，本研究將另外利用非線性模型以檢驗系統風險因子再合併套利報酬之相關性。 藉由實證結果，我們發現無論在典型操作或動態操作下合併套利皆未產生顯著超額報酬，且典型操作之套利報酬和市場報酬並無顯著相關。在市場模型中，動態操作報酬與市場有微弱的顯著負相關；在三因子模型中，此相關性在95 % 信賴水準下並不顯著。推測主因為動態操作套利有更高的絕對報酬且在選取樣本期間市場報酬表現甚至低於無風險利率。另外，合併套利的報酬並未因市場景氣而有顯著的差異，也就是實證結果顯示台灣合併套利特徵較接近澳洲，支持合併套利不因景氣衰退而有顯著正相關。

This paper is intended to be an investigation of the characteristics of returns and risks in merger arbitrage in Taiwan and compares results with results from the United States and Australia. We analyze 29 mergers including successful and failed deals which were announced between 2000 and 2006. In addition to typical investments, we introduce a dynamic arbitrage strategy which adjusts positions in the portfolio depending on the movements of the arbitrage spread in each deal. The analyses reported are based on monthly merger arbitrage returns. Monthly returns are compounded daily which is then calculated for every active deal within the month. We benchmark arbitrage portfolio returns with the linear CAPM model, Fama and French (1993) three factor model, and the piecewise linear model. The following results are obtained: first, merger arbitrage returns can not generate abnormal return whether in a typical investment or dynamic strategy, second, the static arbitrage returns is a insignificantly correlated with market returns, while there are weak negative correlation between dynamic arbitrage returns and market index returns using CAPM market model. Third, in contrast to the United States, we have the same result with Australia that correlations between arbitrage portfolio returns and market returns are the same regardless of different market returns.