Our
process.

I. Sourcing

How we find opportunities.

A public company’s disclosure record is the most widely distributed and the least carefully read body of material in the capital markets. Our edge derives from that asymmetry—not from superior information, faster execution, or a proprietary data source. We read the whole record: annual, quarterly, and current reports, insider and five percent ownership filings, proxy statements, and registration statements. Candidates reach us through defined channels and advance through a defined sequence, so that our process can be assessed as a process rather than as a series of individual judgments.

1. 01

   Why the Record Goes Unread

   Sell-side models are built from company-guided figures. Quantitative screens ingest reported line items and cannot detect that a line item is itself misstated, because the misstatement is inside the input. Auditors test the reasonableness of management’s estimates; they do not value assets independently of them. The disclosure that reveals a problem is usually present—in a footnote, a supplementary schedule, a segment reconciliation—but finding it means reading an issuer’s full filing history in sequence, against its peers. That work is slow, does not scale through automation where it matters, and is not rewarded on a quarterly cycle.

2. 02

   Tracking of Known Bad Actors

   Reporting failures are not distributed randomly across people. Auditors, directors, and officers associated with one failure reappear at other issuers. We maintain a record of those individuals and follow them across the public company universe—not as a presumption of wrongdoing, but as a starting point for examination. The signal is in the personnel rather than the financial statements, so it is available before anything appears in the numbers. It is a record that takes years to build and cannot be purchased.

3. 03

   Two Screens, Neither Valuation-Based

   Proprietary systems, built and owned by the firm, compare filings across issuers, peers, and time. The first screen searches the numbers for patterns that historically precede a restatement, an impairment, or an enforcement action: a change to a materially smaller auditor, a disclosed material weakness, a late filing, earnings that persistently outrun cash, impairment assumptions left unrevised as a business deteriorates. The second searches the reporting record itself—Section 16 reports filed late or not at all, events with no corresponding Form 8-K, related-person transactions absent from Item 404. A required filing either exists or it does not.

4. 04

   Inbound, With Discipline

   Whistleblowers and market participants bring situations to us because our public record shows that we will act on them. Everything that arrives is held to one principle: a submission from outside the firm is a direction to look, and never a basis on which to act. Every conclusion must be independently reconstructible from the public record, and where it cannot be, we do not reach it. We do not solicit information a source is obliged not to disclose, and where a submission may carry a restriction, we consult counsel and restrict dealing in the issuer’s securities until the question is resolved.

III. Portfolio

Portfolio construction and risk management.

We do not define risk as concentration, and we do not define it as volatility. We define it as what an investor does not know about what it owns. A portfolio of two hundred securities about which little is known is not safer than a dozen about which a great deal is known—the ignorance has been distributed rather than reduced. We hold a small number of issuers because the number we can know to our standard is small. The portfolio is long-oriented because, on the long side, we can be the agent of a correction rather than a spectator waiting for one; we hold short positions selectively, where a defect conceals a business worth materially less than reported.

1. 01

   Capital Follows Conviction

   We do not establish a full position and then investigate. We establish an initial position, put our questions to management, and add only as the answers—or a refusal to answer—confirm the analysis. Where the answers dissolve the concern, the position stays small and we move on, having spent research time rather than capital. The result is that our largest exposures are our most tested ones, and a thesis that fails does so while the position is still small.

2. 02

   Liquidity and Disclosure

   We invest across the capitalization spectrum, weighted toward smaller companies, where coverage is thinnest and the record is read least. Thin trading lengthens both accumulation and exit, so we size positions against the possibility of a forced exit rather than an orderly one. Ownership above five percent requires a public Schedule 13D filing, after which any further accumulation takes place in view of the market.

3. 03

   Passivity Is a Risk

   A holder who does not ask questions is not avoiding exposure. The exposure is unchanged, and the holder has declined the only means of learning what it is in time to act. We size every position against the possibility that no catalyst materializes, keep the capacity to exit where an analysis is invalidated rather than merely unrecognized, and decline to escalate where escalation would commit capital to a timeline it cannot support. Where we are wrong, we intend to be wrong about something we examined—not something we never looked at.