可轉換公司債回溯定價之研究

Abstract

本研究主要討論國內發行可轉換公司債時，訂定轉換價日期回溯之影響，觀察可轉債在初級發行時因轉換基準定價日與報備承銷日時間之落差，可推論發行公司有機會將定價日回溯至早先市場股價較低的時候，在這種取巧及資訊不對稱之情況下，可能對股東權益有影響。另2003年12月30日後因法令之修訂使轉換價格決定期間回溯定價可行性下降，所以用此日期作為分水嶺，利用事件研究法之市場模式進行實證分析，以可轉債轉換價格基準日為事件日，觀察事件期之股價累計異常報酬及其趨勢，分析採回溯定價方式發行之影響及其資訊內涵。本研究之實證結論如下： 1. 有回溯定價(140個樣本)：2002~2004年有回溯訂價之在定價前可轉債公司股價累計異常報酬多為負值且呈下滑的趨勢而定價後上升，轉換價格決定基準日前10、15、20個營業日為整段事件期間之相對最低點。 2. 無回溯定價(170個樣本)：2004年~2005年間無回溯訂價之在定價前可轉債公司累計異常報酬亦多為負值且呈下滑的趨勢，定價在低點但該定價時點並非為事件期間之相對最低點，定價時點反映發行公司仍擁有對目前股價及營運的優勢判斷之資訊絕對優勢，但定價後事件期間未來股價變化無絕對的掌控性；結果顯示修法後對市場公平性有正面幫助。 3.將2002年~2003年回溯定價之127個樣本，模擬若不採回溯定價方式，將報備承銷之日期作為定價之基準，與原實際發行基準價比較差異，模擬結果採用不回溯定價後2002年~2003年有97家定價上升，定價下降為30家，遠少於上升公司，而重定後整體轉換價上升平均比率7.2%，潛在利益分別為4,433百萬；模擬結果之內涵，採以較高轉換定價發行而使籌資成本較低之公司較少，大多數公司是將可轉債轉換定價和較低股價的更早日期相連結，使認購可轉債投資人立即賺得紙上利得，從另一方面而言因轉換價格下降發行公司需發行更多股本以籌得相同資金。

This research mainly discusses the impacts of backdating the conversion price base day of domestically issued corporate convertible bonds. Through observations on timing differences between the conversion price base day, the recordation date, and the bookbuilding date, we inferred that a bond issuer has an opportunity to backdate the conversion price base day to the time when the market share price was lower. This opportunistic behavior plus the existence of asymmetrical information may impact shareholders’ interests. Since 30 Dec 2003, new regulations were imposed that greatly reduced the feasibility of backdating the conversion price base day; hence we took this day as the ex-ante date for our empirical analysis using the market model of event study. We designated the conversion price base day as the event day, observed the cumulative abnormal stock returns and its trends over the event period, and analyzed the impact and implication of backdating conversion prices. The conclusions of our empirical research are as follows: 1. With backdating (140 samples): between 2002~2004, the cumulative abnormal stock returns of corporate convertible bonds with backdating clauses were mostly negative and falling before conversion prices were set, and rising after conversion prices were set. The stock prices were the lowest at 10, 15, or 20 business days before the conversion price base day. 2. Without backdating (170 samples): between 2004~2005, the cumulative abnormal stock returns of corporate convertible bonds without backdating clauses were also mostly negative and falling before conversion prices were set. The conversion prices were set low but not at the lowest among the entire event period. The timing reflected that the bond issuer still has the advantage of asymmetric information over stock prices and business performance, but has no absolute control over the changes in stock prices after conversion prices were set. 3. We have also simulated the pricing of 127 with-backdating samples between 2002~2003 as if their conversion prices were not backdated, and designated the bookbuilding date as the conversion price base day. A comparison between the simulated pricing and the actual pricing revealed that 97 samples between 2002~2003 had simulated pricing without-backdating higher than their actual pricing, the remaining 30 samples had simulated pricing lower than actual pricing; the proportion of over-priced issues was significantly less than the under-priced issues. The simulated pricing was on average 7.2% higher than the actual pricing, which translated into a potential gain of $4,433 million. The research also revealed that among the companies who issued bonds with backdating clauses, few have chosen a higher pricing that reduces the company’s funding costs while most have chosen to date the conversion price base day on days when the stock prices are cheapest, thereby producing unrecognized gains instantly for bond subscribers. The downside of the latter is that at a lower price, more shares need to be issued to raise the same amount of funds.