CHANGE LANGUAGE |
Home >
Doc >
Continuous overreaction, insiders trading
activities and momentum strategies >
Insider trading refined momentum strategies

Continuous overreaction, insiders trading activities and momentum strategies

Empirical findings and discussions

Insider trading refined momentum strategies

In this section, we introduce the insider trading information in the processes of momentum portfolio forming to study the influence of insider trading activities on naı¨ve momentum effects. The insider trading refined momentum strategies are formed as the following processes. First, five quintiles are formed based on the past return performance order in the same procedure described in Section 3.1.

Each stock is then assigned a dummy variable, 1, −1 or 0, based on their AITR in past F months. Six portfolios are then formed based on past return performance quintiles and insider trading dummy variables. Winner and insider-buy (WB) contains those stocks are both winners and insider-buy. The WZ, WS, LB, LZ, and LS are defined in the same way. These positions are then held for H months. Again, 1-month gap is left between forming and holding period. The zero-positions are then formed by:

(3)

where WB(LS) is the number of stocks in the WB(LS) portfolio; R ^{WB}_{t }( R ^{LS}_{t}) is the
average equally weighted return of portfolio WB(LS); R ^{WBLS}_{t }is the zero-position
return.
The zero-position, WBLS, is our refined momentum strategy. All other zero-positions
are defined in the same way.
Table 3 reports the average/expected profits earned by these portfolios and
positions and their corresponding t-statistics. The third to ninth columns report the
portfolio returns of WB, WZ, WS, LB, LZ and LS, respectively.

The last two columns report the profits earned by WBLS, insiders trading refined momentum strategies, and WZLZ, the momentum strategies among stocks that contain no insiders trading information. Compare the profits earned by the refined momentum strategy with the profits earned by nai¨ve momentum strategy and insider trading based momentum strategy, the refined strategy has a much higher profit earning ability than the other two.

The 6/6 strategy as an example, the insiders trading refined momentum strategy can earn an average profit of 1.66% per month, while the nai¨ve momentum strategy can only earn 0.93% per month and insider trading based momentum strategy can earn 0.37% per month. Secondly, let's compare the results column by column in Table 3. As stated above, we separate both the winner and loser quintile into three portfolios based on insiders trading activities, insiderbuy, insider-zero and insider-sell. The insider-buy portfolios earn highest return in both winner and loser group. Here, again we take the 6/6 strategy as an example.

The insider-buy portfolio in the winner-group of the 6/6 strategy can earn 1.85% per month, while the insider-zero and insider-sell in the winner-group of the same forming and holding period can only earn 1.24 and 1.41% respectively. A similar thing happened in the loser-group, the per month returns earned by the 6/6 insider-buy, insider-zero and insider-sell portfolios are 0.77, 0.45 and 0.19% respectively. The insider-sell in the loser-group earns the lowest return among the six portfolios in all different time-horizon strategies.

The analysis of the returns of the zero-positions or the spread between the return of these portfolios can draw a clearer picture for us. The WBLS, insiders trading refined momentum strategy, earns higher return than WZLZ. The 6/6 WBLS earns 1.66% per month, compared with WZLZ’s 0.78%. As designed, the WBLS captures the effects of both the naı¨ve return momentum and insiders trading based momentum. The WZLZ captures the effects of the return momentum among those stocks that don’t have insider trading information.

The WBLS can earn as much as two times the profits earned by the naı¨ve momentum strategies and five times as much as the profits earned by insiders trading based momentum strategies. We also compared the WZLZ, the momentum strategy among stocks that don’t have clear insider trading information, with the naı¨ve momentum strategies. We find that the WZLZ’s returns are always less than the naı¨ve momentum strategies in any time horizon. It is reasonable to believe that part of profits earned by naı¨ve momentum strategies is captured by WBLS, the refined momentum strategies.

We would like to interpret our findings as follows. The investors overreact on the stock when they observe insiders are buying (selling) a stock. If the public information can confirm their findings, they will overreact continuously. The continuous overreaction is the cause of momentum effects. Based on the above argument, investor will overreact on WB and LS continuously and underreact on WZ and LZ. The theory will predict that: the WB earns the highest return, while the LS earn the lowest return. The results in Table 4 confirmed the prediction. However, we still cannot reject the possibility that these portfolios capture different risk groups. The WB may be composed by higher risk stocks, while the LS may be composed by the lower risk stocks.

The insider may intend to buy small growth company to obtain abnormal returns and to sell bigger matured value companies for liquidating or other non-abnormal reasons. The profits earned by these WBLS reflect the return spread among different risk level stocks. To study whether the risk dispersion is the source of the profits earned by these zero-positions, we will study the performance of these strategies by adjusting risk factors. Table 4 reports the average number of stocks in each of portfolios formed above. The table shows that there is at least 100 more firms in each portfolios we studied.

Jihong Xiang, Jia He , Min Cao

Continuous overreaction, insiders trading activities and momentum strategies

The paper investigates the influence and explanatory power of aggregate insiders trading activities on momentum trading strategies. We find that insiders trading activities can predict cross-sectional returns and can strengthen the naı¨ve momentum effects.

Jihong Xiang, Jia He , Min Cao