Management of Employee Stock Option Pricing Model Input Assumptions

員工認股選擇權訂價模型參數假設之管理

本文探討台灣高科技公司是否藉由管理選擇權訂價模型參數假設來低估員工認股選擇權價值。我們選取台灣310家高科技公司為樣本，期間涵蓋2004至2009。我們估算員工認股選擇權每權價值低估約新台幣1.2元。另外，低估現象主要是發生在早期。隨後，該價值低估程度呈逐年遞減趨勢。此外，選擇權價值低估主要是預期股價波動率低估所致。再者，證據顯示股價波動率低估的現象是系統性低估而非隨機造成。同時我們也發現選擇權存續年限亦有低估的情形，但其低估對選擇權每權價值的影響數不具經濟意義。我們總結如下，在2008年之前，認股權發行公司確有投機性低估員工認股權價值的現象。但是，或因新經驗的累積，或因第39號公報自2008年起生效，因而所揭露的參數假設將受更嚴格外部審計查核之故，公司在後期較能審慎地估計選擇權模型參數，因而所估計的認股權價值較過去精確。

This study examines whether managers in Taiwan are motivated to systematically understate their employee stock option (ESO) values by managing the option-pricing model input assumptions. Using 310 high-tech firms in Taiwan from 2004-2009, we find that the dollar amount of the discretionary ESO value per option, i.e., the difference between per option value using reported inputs and that using historical inputs, is NT $-1.20. Further, we find that the magnitude of the discretionary ESO value per option is more salient for the period prior to 2008 and exhibits a decreasing trend over time. Additionally, we find the understated ESO value is mainly caused by the stock return volatility, which apparently is determined in an opportunistic manner. We also find that firms tend to report a lower than expected option life; however, the corresponding economic implication is negligible. Taken together, we conclude that the economic consequence of the management of inputs mainly happened in the early stage of the sample period. Managers subsequently become more sophisticated and cautious, probably because firms gain more experience from the past or because disclosed inputs become subject to more rigorous scrutiny by external auditors due to the mandated recognition of ESO expenses starting from 2008.