The research behind the clubs

Each of our two clubs follows one idea from published academic research. This page summarises both ideas in our own words, explains what our alerts take from them and where we deliberately differ. It is a summary, not a substitute for the original studies – please read those too.

Alpha Insider Club – "Decoding Inside Information"

Study: Lauren Cohen, Christopher Malloy and Lukasz Pomorski, NBER Working Paper 16454, 2010 (US insider trades). Read the paper

What the authors show

Insiders trade for many reasons – tax planning, diversification, habit. The authors identify routine trades and set them aside. According to their abstract, the remaining "opportunistic" trades hold the predictive power in the insider-trading data and predict future news at the company level.

Why this could work

People who run a company know more about its prospects than the market does. When they buy shares with their own money outside any routine, that choice may reflect what they know.

What our alerts take from it

• We use only notifications that insiders are legally required to publish (MAR Art. 19).
• We alert when two or more different insiders of the same company buy within 14 days – a stronger signal than a single purchase.
• The score weighs the size of the purchase relative to the company, CEO/CFO involvement and parallel share buybacks. Option exercises and duplicates are removed.
• We alert only when the share price is less than 20 % below its 52-week high.
• Every alert shows daily turnover and warns when the stock is hard to trade.

Where we differ

• The study covers US data. We screen small and mid caps in seven European markets.
• We do not currently classify trades as routine or opportunistic. The cluster of several buyers is our practical stand-in for an information-rich purchase.
• The study measures portfolio returns. An alert is a starting point for your own check.

Redline Alpha Club – "Lazy Prices"

Study: Lauren Cohen, Christopher Malloy and Quoc Nguyen, Journal of Finance 75(3), 2020 (US annual and quarterly reports, 1995–2014). Read the paper

What the authors show

When companies actively change how their annual and quarterly reports are written, this carries information about future operations and returns. Changes in language about the executive team, about litigation and in the risk factor section are especially informative. The authors find no immediate price reaction; the effect appears only later, when the information becomes visible through news, events or earnings. They conclude that investors pay little attention to such changes.

Why this could work

Few people compare the fine print of a report with the previous one. A new risk, a softened warning or a new legal matter can slip through unnoticed.

What our alerts take from it

• Each new 10-K or 10-Q is compared with the company's previous report of the same type.
• An AI rates risk factors, legal proceedings and management's discussion as better, worse or unchanged. A human reviews every alert.
• Revenue and EPS direction from official XBRL data are checked against the wording; we flag when numbers and tone disagree.
• Market capitalisation and daily turnover appear in every alert.

Where we differ

• The study built a portfolio across a broad set of US companies, including short positions. We alert on US micro caps only and do not recommend shorting.
• The study measured changes in the text itself. We use an AI to judge the direction of the change, and an AI can misread a text.
• The club is in a pilot phase. We have no track record yet.

What this page does not say

The studies describe past markets, mainly in the US. Their results are no promise for our clubs, and a study result is not a recommendation to buy or sell any security. We track every alert against a market index and will publish the results once there are enough alerts to be meaningful. SmallCapAlpha is not affiliated with the authors or publishers of the studies.