Smart Insider← Purchase feed

For your own research

Research pool

Research list, not a buy signal. In our 2020–2024 backtest, insider buys at financially decent companies did not beat the S&P 500 once delisted companies were counted (2024: 18–30% of cases beat SPY). Use this to choose companies for your own research.

Loading cached research data…

Screen rules, timing and price limitations

F1: positive net income, operating cash flow and free cash flow; FCF yield ≥ 6%; ROIC ≥ 10%; net cash or net debt/EBITDA < 3. Financials and REITs excluded. Buyback: positive TTM repurchases and share count down at least 1%. Cluster: at least two other independent reporting parties bought in the prior 90 days, known by that purchase’s public time.

Fundamentals use only SEC facts filed before each purchase’s public date. Each company’s metrics describe its latest qualifying buy; all qualifying buys in the selected window are grouped. Market cap, yields and DCF use the purchase-price proxy when a pre-public close is unavailable, with a visible flag. Missing plan disclosure does not prove a spontaneous purchase. Current quotes are retained observations only; missing or stale quotes are shown explicitly. No investment recommendation.

Definitions and data quality

ROIC = 0.79 × EBIT / (debt + equity − cash), requiring positive invested capital. FCF = operating cash flow − capital expenditure. TTM uses annual + current YTD − prior YTD, with matched periods. Debt is the highest supported period-aligned total among Job 4B, carrying amount, notes/facilities without double counting, and current + noncurrent lines. More than 20% divergence is flagged. Ambiguous component overlap is withheld. Short-term debt assumed zero is an assumption, not confirmation of no short-term borrowing.

DCF uses four consecutive annual FCF endpoints, positive first and last endpoints, three-year CAGR clamped to 0–8%, five years of that growth, a linear fade to 2.5% at year ten, 10% discount rate, 2.5% terminal growth, less net debt. Share-count change uses Job 4B’s same-tag prior count filed 9–15 months earlier, with no known intervening split. Figures are screening estimates and may be incomplete.