農曆新年與反處置效應

The Chinese Lunar New Year and Reverse Disposition Effect

農曆新年在中華文化具有一元復始，萬象更新的象徵。本文假設投資人賣出股票的投資決策，亦受這文化象徵影響。根據Ingersoll & Jin (2013) 建立的反處置效應模式，本文假設投資人在年前可能賣出較多帳面虧損的股票，卻賣出較少帳面獲利的股票─一種反處置效應行為。如此，虧損股票在年後可以重新建立部分，好處是新的參考價格可能較舊部位的參考價格為低。於是，增加未來賣出部位時的獲利可能性。也就是在年前認賠，俾股票投資績效在年後展開新局面。本文實證結果和預期相同，即發現購買帳面獲利股票及同時賣出帳面虧損股票的投資策略在年後平均將導致顯著損失。相對而言，該投資策略在年前僅有不顯著的損失。最重要的，該投資策略年後的損失主要來自帳面虧損股票的價格反轉。可見年後帳面虧損股票上漲的幅度較帳面獲利股票為高，同時隱含年前前者較後者面臨較大賣壓。鑑於實務上觀察發現，農曆年前有資金需求強烈的現象，本文假設前期M1B減少的愈嚴重，代表短期資金供給愈不足，則上述反處置效應行為愈強烈。再者，考慮共同基金可能在年底有增加持股(portfolio pumping)的行為，以推升其持股較重股票的股價。此買盤可能會抵銷處置股票的賣盤，故本文假設共同基金持股水準愈高，上述反處置效應愈弱。上述二假設都獲得證實，同時相關結果無法由Fama & French (1993) 的風險三因子解釋之。

The Chinese Lunar New Year (CLNY) traditionally symbolizes a fresh start, including for investment performance. Therefore, we assume the presence of a reverse disposition effect prior to the CLNY in the Taiwanese stock market, that is, investors are more willing to sell losers than winners to reset the reference prices of losers so as to realize gains in future as predicted by Ingersoll & Jin (2013). Consistent with the predictions, we find that the arbitrage portfolios of longing paper-winners and shorting paper-losers generate significantly negative returns in the post-CLNY period but the portfolios yield only weak negative returns in the pre-CLNY period. More importantly, this strong negativity is primarily attributable to a price reversal of losers. Furthermore, we use the lagged change of M1B to inversely proxy for the necessity to sell stocks to satisfy the customary liquidity demands prior to the CLNY and thereby hypothesize an inverse relationship between the money supply and the reverse disposition effect. Moreover, we assume that mutual fund holdings are negatively associated with the effect because (due to year-end portfolio pumping behaviors) mutual funds provide a buying force against the aforementioned selling forces. Consistent with the assumptions, the evidence indicates that a significant negative arbitrage return is observed during the post- CLNY period when there is a decreased lagged change in M1B and for stocks with low or medium levels of mutual fund holdings. The conditional negative arbitrage returns survive tests of the three risk factors identified by Fama & French (1993).