Standards
Research Standards
The standards below govern every FIBCOS research publication and every entry in the public Alpha record.
Applied to this website
These standards are why performance fields show an em dash rather than a number, why demonstration records carry a DEMO label and are excluded from every count, and why no client, credential or track-record claim appears anywhere on this site.
The standards
Select a standard to read it in full
No company enters a country Structural Alpha Index before the required research is complete and the investment thesis has been documented and approved. Inclusion follows research; it never precedes it.
Every structural driver must be supported by evidence that can be stated and checked. What cannot be evidenced is recorded as an assumption and treated as a risk.
Every thesis states in advance what evidence would prove it wrong. A thesis that cannot be wrong is not a thesis.
Theses are reviewed continuously and on defined triggers. Reviews produce published thesis updates, whether they strengthen or weaken the original conclusion.
Active, Under Review and Removed are public, dated states. Status changes are published when they are made, not when they are convenient.
Portfolio entries, exits and position changes are disclosed with the reasoning behind them.
Benchmarks are adopted before inception, derived from portfolio characteristics, and never selected or restated to improve a comparison.
Performance is reported on a stated basis, for identical periods, including periods of underperformance and positions that failed.
Key risks are published alongside structural drivers on every company record. A thesis presented without its risks is a marketing document.
FIBCOS does not backdate inclusions, does not quietly remove failed positions, and does not construct a record from the subset that worked.
FIBCOS does not publish invented returns, Alpha, CAGR, drawdown, AUM, clients, institutional relationships, licences, awards or credentials — including as placeholders.
The public record states the decision. The paid research states the work. Neither is presented as the other.