Readout Subscribe
Disclaimer. This site reads publicly available research papers and corporate filings and explains what they say. It is not investment advice and does not recommend buying or selling any security. Historical cases shown here are computed from public data; the sample sizes are small and are not a basis for prediction. We do not answer questions about individual stocks.

This page was machine-translated from the Korean original and has not been reviewed by a person. Quoted source material is left in its original language and is not translated. Korean original

Bio

Readout case C014 · retroactive reading · abstract basis

Pharmaceutical filing citing a Journal of the American College of Cardiology paper — Amarin \UK, 2019-05-01

Amarin UK cited a paper published in the Journal of the American College of Cardiology as the basis for its claims in a regulatory filing submitted to the U.S. Securities and Exchange CommissionU.S. Securities and Exchange Commission
Regulator · Website
. The company read the manner in which it wrote the paper and the scope it actually demonstrated in sequence. It described where the stock price movement over the following month compared to past cases of the same type. This is a retrospective judgment written after knowing the results. The paper was read only up to the abstract.

Summary of facts. Amarin UK cited Journal of the American College of Cardiology paper in the 8-K (Item 2.02 Earnings Release · 9.01 Financial Statements·Attached Documents) filed on 2019-05-01. The paper was published on 2019-03-18. The abnormal return on the day of the filing was −5.6%, and the cumulative return over the following 20 trading days was +10.5%. When arranging 142 cases in order, it ranks in the 83% percentile from the bottom. This number is a past fact, not a rule to apply to future cases.
−5.6%
Excess return on filing day
+10.5%
20-trading-day cumulative excess return
83%
Percentile within 142 samples (lower means faster)
−7.4%
Median of samples (20-trading-day)

The disclosure is written in the press release as follows:

When a U.S. listed company has something to inform its shareholders, the form it files with the Securities and Exchange Commission is called 8-K. Amarin UK's item number in the document filed on 2019-05-01 was 2.02 Earnings Release · 9.01 Financial Statements·Attached Documents. The company cited a paper published in Journal of the American College of Cardiology in the press release attached to it. If the sentence citing the paper is translated verbatim, it reads as follows.

Recurrent event analyses for the total primary endpoint events and for the total key secondary endpoint in REDUCE-IT as published in the Journal of the American College of Cardiology9 were conducted using a series of statistical models.

The reader should first check where the company placed the paper in the press release. Next, they should determine whether the paper supports the company's entire claim or only part of it. To answer the latter question, one must read the full paper. This article, which only read the abstract, cannot provide an answer.

The scope of the paper discussed in the abstract is limited to this.

The paper's title is Effects of Icosapent Ethyl on Total Ischemic Events, and it was published in Journal of the American College of Cardiology on 2019-03-18. The authors are affiliated with Brigham and Women's Hospital · Harvard University · Inserm · Délégation Paris 7. The summary at the beginning of the paper is called abstract. This article only read up to that point. The paragraph was written by a person after filtering a machine-generated draft by numerical comparison, and thus differs in quality from a full reading of the paper.

This is an additional analysis of a large trial involving 8,179 high-risk cardiovascular patients with high triglycerides who were taking statins. While the previous analysis only looked at the time to the first event, this one counted all events that recurred in the same patients. The group taking four grams daily of prescription omega-3 formulation had fewer overall ischemic events than the placebo group. Although this is a pre-specified additional analysis, it redefines the primary endpoint, so caution is needed in interpretation.

What the abstract does not answer remains in the full text and appendices. This article, which did not go beyond the abstract, leaves gaps as they are. Readers should compare the two citations above to determine which sentence in the abstract corresponds to the excerpt cited by the company.

One month later, the stock price was within the top quartile of the sample.

The remaining return, after subtracting the portion that moves with the overall market, is called excess return. The date is written as a trading day, counting only the days the stock market is open. In calendar terms, 20 trading days is about a month. The excess return for the day of the regulatory filing and the following month are listed in the table below.

PeriodExcess return relative to the marketMeaning
Five trading days before the filing+1.7%The extent of movement before the filing
On the day of the filing−5.6%On the filing day or the first trading day thereafter
Accumulated over +1 to +5 trading days−6.0%One week
Cumulative +1~+20 trading days+10.5%One month

A bundle of past cases of the same type is called a sample. The position counted from the bottom is expressed as a percentile. The one-month excess return of +10.5% in this case is in the top quarter, corresponding to the 83% position from the bottom among 142 samples.

Sample scale (20 trading days)Tenth from the bottomFourth from the bottommediantop 25%top 10%This case
Excess return−31.9%−20.1%−7.4%+5.5%+19.5%+10.5%

Half of the sample falls between the two central marks on the chart. One checks from the chart whether this case lies within or outside that range. The sample collects cases where U.S.-listed companies cited papers in 8-K filings over the past ten years. The company itself is excluded from the sample. The numbers listed here are past facts, not rules to apply to future cases.

After the disclosure, the company issued such documents

After the disclosure date, the company issued three 8-K filings, listed here only by item number. In 8-K filings without titles, the item number indicates the type of content. One can determine whether it was an earnings announcement or a contract based on that number. The nature of the event is read in the original text of each link.

What this article does not answer

This article, based only on the abstract, does not answer what the paper reviewed, by what method, and to what extent. How often the paper cited by the company was cited afterward can be verified again through the OpenAlexOpenAlex
Data source · Website
citation count. What will be seen next is whether the company's next 8-K again cites the same paper.

How was it measured? The sample consists of 144 cases selected from 1,561 cases of U.S. listed companies' 8-K filings between 2015-01-01 and 2026-09-03 that included a paper DOI, where a paper was published between -3 and +60 days from the filing date and stock prices could be attached. Excess return is the cumulative daily return minus the market return, and the window is +1 to +20 trading days with the filing date set to 0. Stock prices are Sharadar adjusted close prices, paper metadata and abstracts are from OpenAlex, and filing searches are from SEC EDGAR professional search. Abstract summaries are generated by a machine, and the numbers in the summary are compared with the abstract by a program and then reviewed by a person. Review date 2026-10-07.

Glossary

8-K
A voluntary report submitted to the Securities and Exchange Commission when a U.S. listed company has news to inform shareholders. Types are indicated by item numbers such as earnings, contracts, and executive changes. All events in this article come from this document.
Excess return
The remainder after subtracting the overall market return from the stock return. It is used to observe the movement of the company alone by removing the portion where the market as a whole rises or falls.
Trading day
A date counting only the days when the stock market is open. Since weekends and holidays are excluded, 20 trading days correspond to roughly one calendar month.
Sample and percentile
A bundle of past cases of the same type is a sample, and the position from the bottom is expressed as a percentile. A percentile of 10 means that nine out of ten cases performed better than this case.
Abstract
An abstract is placed at the beginning of a paper. The author writes in a few sentences what was done, how it was done, and what results were obtained. Without reading the full text, one cannot know the details or limitations of the methods.