Proof of a clientele effect: evidence from Taiwan

Taxes and Dividend Clientele: Evidence from Trading and Ownership Structure By Lee, Liu, Roll, and Subrahmanyam



Lee, Liu, Roll, and Subrahmanyam (LLRS) provide convincing evidence that a dividend clientele effect does exist. While previous researchers (for example Scholz-1992, Dhaliwal, Erickson, and Trezevant-1999, and Graham and Kumar-2003) have also found the existence of a clientele effect, the current paper is different in that it is based on cleaner data and does not rely on complex modeling. Rather the authors examine trading, ownership, and tax rate data from Taiwan. As they state "Taiwan offers an excellent laboratory for studying clientele because the capital gains tax is zero and share repurchases were prohibited for most of our sample period."

Before getting to the findings, some background is necessary. Taiwan does not tax capital gains, but does tax dividend income. Data include all trades as well as approximations to the traders' marginal tax rate. Additionally the rules on stock buybacks changed in September 2000 which enabled the authors to "study the behavior of firms as they became able to evade dividend taxes."

And the findings? "Individuals appear to respond in the direction predicted by the clientele hypothesis." Wealthy individuals decrease their net buying after dividend increases and increase net buying after dividend decreases. Those in lower tax brackets "do just the opposite."

"Institutions as a group display an insignificant response to dividend changes."

LLSR further examine this using regression analysis. Consistent with the above findings, "there is a strong negative relation between dividend increases and the proportion of shares held by wealthy individuals."

Further examination of the institutional ownership suggests that "Among institutional types, both tax exempts and corporations significantly prefer higher dividends per share. They also prefer higher payout ratios and are joined in this preference by financial institutions."

Finally, the authors also look at changes in behavior after the legalization of share repurchases. They find that "firms with higher concentrations of highly taxed shareholders were significantly more likely to commence repurchase programs. More than forty percent of Taiwan firms actually engaged in share repurchasing after it became possible. A significant fraction (23%) of firms that had previously been paying dividends ceased paying them entirely and 21% reduced dividends and began repurchasing. The tendency to engage in these practices is significantly related to the proportion of a firm’s shareholders in higher tax brackets."

http://www.anderson.ucla.edu/acad_unit/finance/wp/2004/5-04.pdf


I am convinced. Are you? Definitely an interesting paper and it will make it to my class notes!

BTW How can there be so many interesting articles? I just do not understand. It seems like everywhere I look there are articles that are really really good! This is no exception.