Capital Gains Tax, Supply-driven Trading and Ownership Structure: Direct Evidence of the Lock-in Effect

Dean Hanlon
Monash University
Australia

Sean Pinder`
University of Melbourne
Australia

Abstract:
Prior research suggests that capital gains tax (CGT) influences investors’ supply of shares for trading via the lock-in effect. The lock-in effect suggests that CGT creates a disincentive to sell appreciated shares, as CGT is imposed once shares are sold. A CGT rate decrease reduces the reluctance to sell appreciated shares, thereby increasing their supply available for sale. Empirical evidence exists to support this assertion (Reese, 1998; Blouin, Raedy and Shackelford, 2003; Ayers, Li and Robinson, 2006; Hanlon and Pinder, 2007; Dai, Maydew, Shackelford and Zhang, 2007). These studies, however, have only been able to infer the extent of selling pressure indirectly but the present study makes use of bid and ask quote data to directly measure the number of buyer versus seller initiated trades.

Our study tests the lock-in effect, by examining supply-side trading for a sample of 410 Australian IPOs immediately following their 12-month anniversary of listing. Doing so extends prior research in two ways. First, while prior research examines abnormal trading volume, we analyse buy-sell order imbalance. This allows us to directly test the lock-in effect by determining whether, as expected, any shift in the buy-sell order imbalance upon the 12-month anniversary of listing is supply-driven. Second, as the CGT discount extends is biased towards individual investors, we use CHESS registration data to control for the level individual investor ownership in the period surrounding the12-month anniversary of listing.

Our results confirm that there is an increase in selling pressure around the 12-month anniversary of listing. Furthermore, we find that this effect is strongest when the 12-month anniversary occurs at the beginning of the financial year (in July). These results are consistent with the selling pressure being tax motivated, as individual investors delay payment of their CGT liability until the end of the financial year. Importantly, we find that the proportion of CHESS-registered individual investors in a firm is significant in terms of explaining the cross-section variation in buy-sell order imbalance. This is the first study to directly link the three separate (but related) effects of taxes, buy-sell order imbalance and ownership structure.