The research behind the data
Every filing algoum reads is public on SEC EDGAR. So why should the information in it be worth anything? This page summarises what decades of peer-reviewed research in finance and accounting have found about insider transactions, material events and institutional holdings, and how each finding shaped the data we provide. Whether you read filings yourself or test a signal as an algorithmic trader, every claim names the study it rests on, so you can check it yourself.
Why public information can still matter
If prices instantly reflected everything that is known, nobody would have a reason to gather and analyse information, and then prices could not reflect it either. Economists therefore expect markets to be efficient only up to the cost of processing information: those who do that work are compensated for it (Grossman & Stiglitz, 1980). For SEC filings, that cost is concrete: thousands of documents a day, written for lawyers, in formats that change from form to form. The research below shows that what is in them does reach prices, but not instantly and not evenly.
Insider transactions
Officers, directors and large owners know their company better than anyone outside it. When they buy or sell its stock with their own money, does that tell the rest of the market something?
What the research finds
- Insider trades predict later stock returns. The question has been studied since the 1970s (Jaffe, 1974). Across all companies listed on the NYSE, AMEX and Nasdaq over twenty years, insider activity predicted how stocks went on to perform, most clearly in smaller companies. The trade and its report caused little price movement at the time; the information showed up later (Lakonishok & Lee, 2001).
- Purchases say more than sales. An insider sells for many reasons that have nothing to do with the company: diversification, taxes, a house. There is essentially one reason to buy. Accordingly, the predictive power comes from purchases, while sales show little or none (Lakonishok & Lee, 2001; Jeng, Metrick & Zeckhauser, 2003).
- Not every trade is equal. Insiders who trade at the same time every year, out of habit, carry little information. When an insider breaks such a pattern, the trade is informative (Cohen, Malloy & Pomorski, 2012).
- Speed matters. Since 2002, Form 4 has to be filed within two business days instead of weeks. After the change, prices and trading volume reacted significantly more strongly to filings of insider purchases: more of the information now reaches the market through the report itself (Brochet, 2010).
What this means for the data
The research only works once purchases are separated from sales, planned trades from discretionary ones and a chief executive from a passive shareholder. That is why every transaction carries side, the raw SEC code, the insider’s role and, where the filing allows it, is_10b5_1 for trades under a pre-arranged plan. This way the distinctions above are part of the data, not something you have to guess.
Material events
A company has to report a merger, a change of leadership, an auditor resignation or a bankruptcy within four business days on Form 8-K. Are these reports news, or do they only confirm what is already known?
What the research finds
- They are news. In 2004 the SEC widened the list of reportable events and shortened the deadline to four business days. Under these rules, every kind of item, old and new, comes with unusual trading volume and price swings around the filing, a sign that investors revise their view when the report arrives (Lerman & Livnat, 2010).
- The kind of event matters. For some items, prices keep drifting in the days after the filing, so the market has not finished processing the news when the report appears (Lerman & Livnat, 2010).
- A delay is information too. When a company announces that its quarterly or annual report will be late, that signals continued weak performance that prices do not fully reflect at the time. Accounting problems as the stated reason mark a particularly serious delay, and investors do not take management’s explanations at face value (Bartov & Konchitchki, 2017).
- Filed is not the same as noticed. Investors have limited attention, and theory predicts that information which is public but hard to process reaches prices more slowly (Hirshleifer & Teoh, 2003). The data bear this out: earnings announced on a Friday, when attention is lower, get a weaker immediate reaction and a stronger delayed one (DellaVigna & Pollet, 2009). A dense legal filing is hard to process by nature.
What this means for the data
Because reactions depend on the kind of event, each filing is normalised from its SEC item number into an event_type, with the original item_code kept alongside. Notifications of late filing are their own type, late_filing. Each event carries the time of its filing in filed_at, so you can tell what was public at any given moment.
Institutional holdings
Large investment managers have to disclose their US stock holdings every quarter on Form 13F, up to 45 days after the quarter ends. Can a report that late still be useful?
What the research finds
- Institutional demand moves prices. Large institutions nearly doubled their share of the US stock market between 1980 and 1996. Their preference for large companies alone can explain much of why those stocks rose relative to small ones over that period (Gompers & Metrick, 2001).
- Direction and breadth both matter. Stocks that many mutual funds buy at the same time have done better over the following six months than stocks that many funds sell (Wermers, 1999). When fewer funds hold a stock than in the quarter before, that has predicted weaker returns, and more funds stronger ones (Chen, Hong & Stein, 2002).
- Part of the effect is already priced in by the time the report appears. Changes in institutional ownership go hand in hand with returns in the same quarter. Part of this price effect is temporary, part of it lasts, and the lasting part is tied to information (Sias, Starks & Titman, 2006). A 13F shows the state of a portfolio and how it changed, not the trades as they happen.
- Amendments are not noise. Managers may ask to keep positions confidential and disclose them later in an amendment. These hidden positions are disproportionately tied to information-sensitive situations and have shown superior performance for up to twelve months (Agarwal, Jiang, Tang & Yang, 2013).
What this means for the data
The value is in counting, not in a single fund: per stock and quarter the API reports how many funds hold it, how many increased, reduced, opened or closed a position. Amendments are applied to the quarter they belong to, so you see the corrected picture, including holdings disclosed late. quarter_complete tells you whether the filing deadline has passed or whether funds are still missing.
What the research does not say
- These are averages over many thousands of filings and many years. They describe a tendency, not what will happen to a particular stock.
- Effects that have been published tend to weaken afterwards, partly because more people act on them (McLean & Pontiff, 2016). Past patterns are no promise.
- Using information costs something: time, trading costs, and the risk of being wrong. The studies above measure information content, not what an individual investor can earn after costs.
algoum provides data, not recommendations. See the disclaimer for what that means, and data sources for which forms each area is built from.
References
- Agarwal, V., Jiang, W., Tang, Y. & Yang, B. (2013). Uncovering hedge fund skill from the portfolio holdings they hide. Journal of Finance, 68(2), 739–783.
- Bartov, E. & Konchitchki, Y. (2017). SEC filings, regulatory deadlines, and capital market consequences. Accounting Horizons, 31(4), 109–131.
- Brochet, F. (2010). Information content of insider trades before and after the Sarbanes-Oxley Act. The Accounting Review, 85(2), 419–446.
- Chen, J., Hong, H. & Stein, J. C. (2002). Breadth of ownership and stock returns. Journal of Financial Economics, 66(2–3), 171–205.
- Cohen, L., Malloy, C. & Pomorski, L. (2012). Decoding inside information. Journal of Finance, 67(3), 1009–1043.
- DellaVigna, S. & Pollet, J. M. (2009). Investor inattention and Friday earnings announcements. Journal of Finance, 64(2), 709–749.
- Gompers, P. A. & Metrick, A. (2001). Institutional investors and equity prices. Quarterly Journal of Economics, 116(1), 229–259.
- Grossman, S. J. & Stiglitz, J. E. (1980). On the impossibility of informationally efficient markets. American Economic Review, 70(3), 393–408.
- Hirshleifer, D. & Teoh, S. H. (2003). Limited attention, information disclosure, and financial reporting. Journal of Accounting and Economics, 36(1–3), 337–386.
- Jaffe, J. F. (1974). Special information and insider trading. Journal of Business, 47(3), 410–428.
- Jeng, L. A., Metrick, A. & Zeckhauser, R. (2003). Estimating the returns to insider trading: a performance-evaluation perspective. Review of Economics and Statistics, 85(2), 453–471.
- Lakonishok, J. & Lee, I. (2001). Are insider trades informative? Review of Financial Studies, 14(1), 79–111.
- Lerman, A. & Livnat, J. (2010). The new Form 8-K disclosures. Review of Accounting Studies, 15(4), 752–778.
- McLean, R. D. & Pontiff, J. (2016). Does academic research destroy stock return predictability? Journal of Finance, 71(1), 5–32.
- Sias, R. W., Starks, L. T. & Titman, S. (2006). Changes in institutional ownership and stock returns: assessment and methodology. Journal of Business, 79(6), 2869–2910.
- Wermers, R. (1999). Mutual fund herding and the impact on stock prices. Journal of Finance, 54(2), 581–622.