自觉论文读得太少，没人讨论，于是开个帖子，聊聊读的公司金融方面的论文。

第一篇：

Ferri, Fabrizio, and Nan Li, 2016, The Effect ofOption-Based Compensation on Payout Policy: Evidence from FAS123R, SSRNElectronic Journal.

下载地址：https://papers.ssrn.com/sol3/papers.cfm?abstract_id=2879218

文章运用了一个外生性冲击：

To address these problems we utilizethe 2005 adoption of Financial Accounting Standard 123R (hereinafter FAS123R)as an exogenous shock to the use of option compensation. Previously, firms were required to expense ESO’sintrinsic value at the grant date while disclosing the fair value amount infootnotes (this amount is referred to as ‘implied option expense’). As most ESOwere granted at the money and, thus, with zero intrinsic value, firms reportedno expense in their income statements. FAS123R required firms to expense ESO inthe income statement at fair value.

简言之，在这项规定之前，公司在income statement上面报的期权报酬的费用大都为零，规定之后，报的期权的费用大幅上升，导致期权报酬大幅下降。

作者引用大量文献以说明现有文献的观点是：As the managers’ option compensationincreases, managers prefer to replace dividends with repurchase becausedividend decreases the value of the executive options. The question is that whetherthe prior conclusion still holds after taking into account the endogeneityproblem.

The mainregression is reported in Table 3. The dependent variable is dividend/asset. Theregressor of interest is post*treat where treat=0 if the firms granted no CEOstock options or began expensing the options at the fair values before theregulation. The main result is that post*treat is insignificant, showing CEOstock option may not have a big impact on the dividend policy. Similarly, theresult is insignificant for repurchase. They seem to prefer first order differenceregression instead of the one using post*treat.