Risk disclosure and disclaimer

Aperture publishes statistical research, model output and general market commentary about publicly traded securities. It is not a registered investment adviser, a broker-dealer, a commodity trading advisor or a bookmaker, and it is not registered with the SEC, FINRA, the CFTC or any state securities regulator.

Nothing published here is personalised investment advice. We do not know your financial position, your tax situation, your risk tolerance, your time horizon or your other holdings, and we do not take them into account. All content is impersonal and is directed at the general public rather than at you specifically.

Aperture is a measurement tool. It does not maintain a buy list, it does not tell you to open or close a position, and it has no view on your account. What it publishes is a probability, a two-sided move forecast and the evidence each was read from.

That is a calculation, not a recommendation to buy or sell anything. Nothing here is sized to your account, your capital or your tolerance for loss. Acting on any of it is your decision alone.

• A failed clinical readout routinely takes 60-90% off the equity of a single-asset company in one session, and the stock may be unhalted at a price no stop order could have protected you from.
• An approval that arrives on time can still be followed by a fall, if the label, the timing or the commercial opportunity disappoints what was already priced in.
• Dates move. An FDA action date can be extended, a readout can slip a quarter, and a trial can be stopped. A position held for an event can be destroyed by the event simply not happening when expected.
• These are thinly traded names. Order books in microcap biotech are shallow, gaps are routine, and your fills may be materially worse than the last printed price.

A calibrated 70% means roughly three in ten of those events resolve negative. That is the model working correctly, not failing. Any individual forecast on this site may be wrong, and some will be badly wrong. The published calibration curve is the honest summary of how often.

Past performance of a statistical model — ours or anyone's — does not indicate future results. Model performance decays as a field matures and as other participants price the same information.

Every input is a public filing, a registered trial protocol, a published regulatory notice, a market price, or our own calculation on top of licensed market data. We do not receive, seek or act on non-public information about any company, drug or regulatory decision.

Aperture and its personnel may hold positions in securities discussed on this site. Where a material position exists in a named security, it is disclosed on the relevant report. We do not accept payment from issuers for coverage, and no issuer has any influence over a score.

Content on this site is not directed at any person in a jurisdiction where its publication or availability would be contrary to local law. It is your responsibility to know whether you may lawfully access and act on it.