Lesson 4 of 6
Where a channel check stops
A researcher may assemble public and immaterial pieces into a conclusion. What that does not cover is one material piece from someone who owed a duty.
By Javier Sanz
Founder of ValueScreener. Built Ninety Nine, a securities services provider, and sold it to a tier-one global bank; ran operations and internationalization at Alpaca, the brokerage API.
Who writes thisEditorial processLinkedIn (opens in a new tab)X (opens in a new tab)
In one minute
- Assembling many pieces, each public or individually immaterial, into a conclusion nobody else has reached is research.
- Receiving one material non-public piece from a person who breached a duty in telling you is a different thing entirely.
- The difference is not the value of the conclusion. It is where each input came from and what duty attached to it.
Scuttlebutt sends a researcher to customers, suppliers, competitors and former employees. That work is legitimate and it is also the point at which this academy's method touches a criminal statute, so the boundary deserves a lesson rather than a line.
The same conclusion, reached two ways
A mosaic
- Public filings, published prices, industry data anyone can buy.
- Observations a member of the public can make: a shop's queue, a job advert, a product change.
- Views from people speaking about their own industry in general terms, owing no duty in respect of what they say.
A tip
- One specific unreleased figure: this quarter's revenue, a contract about to be signed, an approval about to be granted.
- Passed by somebody who owed a duty to their employer or to the source, who got something out of passing it.
- Received by a person who knows, or should know, where it came from.
What separates them is the provenance of each input, not the quality of the conclusion.
Why the pieces matter more than the total
The tippee rule in Dirks is about each input, not about the conclusion. A tippee assumes a duty not to trade only when the insider breached a duty by disclosing and the tippee knows or should know of the breach. Nothing there depends on how valuable the resulting analysis turns out to be.
That is what makes a mosaic defensible in principle and a tip dangerous even when the trade loses money. The question is asked of every input separately.
The practical difficulty with expert networks
None of those four is a legal test, and this page cannot give one. They are the points at which a careful researcher stops and takes advice, which is the only honest instruction this lesson can give.
What the law's shape implies for research design
Two habits follow from the structure of the theories rather than from any rule of thumb.
Record where each input came from while you are gathering it. The question a regulator or a court would ask is about provenance, and provenance is impossible to reconstruct months later from memory.
Prefer inputs whose supply is unlimited. A published filing, a price series, a patent register or an industry statistic can be given to a thousand people without anybody breaching anything. An input that is valuable because only you have it is exactly the input whose origin needs explaining.
Check yourself
What makes a mosaic different from a tip?
- The mosaic is less valuable
- Where each input came from and what duty attached to it
- The mosaic uses only numbers
Show the answer
Answer: B.
The tippee rule asks about the provenance of the information and the duty breached in passing it, not about how good the resulting analysis is.Under Dirks, when does a recipient assume a duty not to trade?
- Whenever the information is useful
- When the insider breached a duty in disclosing it and the recipient knows or should know of the breach
- Only if the recipient paid for the information
Show the answer
Answer: B.
Liability is derivative from the insider's breach, and the recipient's knowledge of that breach is part of the test.Which research input is least likely to raise this problem?
- A specific unreleased revenue figure from an employee
- A published filing or an industry statistic anyone can obtain
- A decision described by a contractor before it is announced
Show the answer
Answer: B.
Inputs whose supply is unlimited can be shared with anybody without a duty being breached, which is why they carry no provenance problem.Sources
- Dirks v. SEC, 463 U.S. 646 (1983): tippee liability as derivative of the insider's breach, and the requirement that the tippee knows or should know of it.
- United States v. O'Hagan, 521 U.S. 642 (1997): the misappropriation theory, under which a duty to the source of the information is enough.
What this leaves out
- This is education about what published decisions say, not legal advice, and nothing here tells any reader whether a particular conversation is lawful.
- The four signals in the warning box are practical prompts to stop and take advice. They are not legal tests and no court uses them.
- These are United States authorities. Other jurisdictions treat the same conduct differently, and the law changes.
- Nothing here is advice about any security.