ทวีตที่ปักหมุด
[RESEARCH] Do stocks outperform when insiders buy the dip?
[academic research]
Seyhun wrote a few papers + a book called “Investment Intelligence from Insider Trading” in the 90s that found that insiders are contrarian investors (buy low, sell high) and that insiders outperform. He also found that aggregate insider trading data is able to predict future stock market performance (the basis for our insider sentiment score)
Lakonishok & Lee (2001), “Are Insider Trades Informative?” reaffirmed these findings and found that insider purchases outperform even after controlling for valuation
Jenter (2003), “Market Timing and Managerial Portfolio Decisions” found that insiders buy when they believe the market is fundamentally undervaluing their company
Lasfer (2024) - "Corporate insiders' exploitation of investors' anchoring bias at the 52-week high and low" found that insiders do significantly more buying at 52-week lows and way more selling at 52-week highs, but their purchases at both the 52-week high and low significantly outperform
[ceo watcher research]
The CEO Watcher database has all insider trades back to 2009. Before any analysis (unless otherwise noted), we remove all scheduled trades (Employee Stock Purchase Plans, Tax Sales, 10b5-1 Plans, Dividend Reinvestments, etc) and all Public Offerings, Private Placements, and Purchase Agreements. We also remove any erroneous trades (mismatching transaction codes, when the filing date is more than one month after the transaction date, etc) and any super small trades (<$5,000).
Looking at insider purchases after one-month stock price dips, you’ll see that dip buys outperform the S&P across all timeframes (next 1M/3M/6M/1Y) with improving performance as the dip size increases. (image 1)
These returns are more volatile than the S&P, which we can see by looking at the median returns. The S&P mean and median returns are very similar (as expected), while the median return for insider dip buys falls a bit due to some insider dip buys that massively outperform, pulling the mean up.
However, for dips greater than 20%, the median dip buy still outperforms the S&P. (image 2)
Unlike much of the academic research, which finds that insider purchases only outperform for small caps (generally speaking, not necessarily for dip buys), we see significant outperformance for both small and large caps after dip buys. (<$2B mkt cap is small cap, >=$2B is large cap). (images 3+4)
The data also has fairly large sample sizes, with the bottom end ranging from 868 purchases for the Large Cap 1m_ago < -30% group to 7126 for the All Cap 1m_ago < -30% group (the most restrictive 1m cutoff).
[appendix] I'm going to respond to this post with different cuts of the data. Feel free to ask for any specific cut you would like to see
English
