美國公司盈餘分配以發放股利的方式有現金股利和股票股利二種，每種分配方法在學術上都有其深入的探討，但鮮少文獻將二種股利發放方式之選擇作一綜合性討論。此外，我們並不了解什麼因子驅動現金股利與股票股利之選擇及投資人從公司管理者之股利選擇中可以獲得什麼資訊。因此，本文中我們分析了在美國市場上NYSE、AMEX、NASDAQ上市上櫃公司，其現金股利與股票股利發放之關係。我們先從CRSP資料庫中找出現金股利及股票股利宣告日之分類碼，將這些宣告日年化以整合COMPUSTAT資料庫中的公司財務資料。樣本期間從1984到2003年。本文發現公司規模、投資機會和獲利性在解釋現金股利和股票股利之相對支付方式有顯著的解釋能力。在公司特徵分析和多項式邏輯模型所求之結果一致。公司特徵分析中提出了現金股利發放公司及股票股利發放公司的差異，現金股利支付公司的規模較大且有較多的保留盈餘、營運現金流量，並有較高的歷史股利發放率。反觀股票股利，股票股利支付公司有較多的資本支出及未來的投資機會。在本文最後，我們用多項式邏輯模型來預測可能的支付方式（現金股利與股票股利之選擇性），結果和公司特徵分析一致，並支持了在財務彈性上，股票股利仍為公司股利政策考量上不容忽視的支付選擇方式，尤其是高投資機會的公司。 Corporations use cash dividends and stock dividends as the methods to distribute earnings to shareholders. While each method has received considerable attention in the academic literature, fewer studies examine the choice between cash dividends and stock dividends. In particular, we do not know what factors drive the choice between cash dividends and stock dividends and what, if any, information investors infer from this managerial choice. In this paper we analyze the relationship between the cash dividends and stock dividends on the NYSE, AMEX and NASDAQ. We start with a complete listing of announcements of dividends distribution code supplied by CRSP database. From those announcements, we annualize the dividends to integrate with financial variables supplied by COMPUSTAT database. The sample period contains two decades through 1984 to 2003. This study shows that firm size, investment opportunities and profitability have significant power in explaining relative payouts of cash dividends versus stock dividends. The univariate results tell a similar story to the multivariate results. The univariate comparisons suggest that cash dividend paying firms and stock dividend paying firms are noticeably different. Cash dividend paying firms have more retained earnings and operating cash flows with large firm size and historically high payout ratios. Unlike cash dividends, firms with paying stock dividends have higher capital expenditures and Tobin’s q which means firms have investment opportunities in the future. Finally, we use multinomial logit model to predict payout policy. Our multivariate analysis is consistent with the univariate comparisons discussed earlier and generally support the hypothesis that the financial flexibility inherent in the stock dividends distribution is also a consideration in the choice of payout methods, especially for higher capital expenditure firms.
