Evidence grade
A — broad and repeatedly examined
The idea has a clear definition, evidence across multiple markets or long samples, and enough public detail for independent reconstruction. An A does not imply a positive future return.
B — credible but narrower or fragile
The hypothesis has respected primary research, but results may depend more heavily on construction, sample, instruments or a particular historical environment.
C — exploratory
The idea may be interesting, but evidence is limited, difficult to replicate or dominated by assumptions that cannot be validated with public data.
Retail feasibility
We assess data availability, point-in-time requirements, shorting, leverage, derivatives, financing, rebalance frequency and cost sensitivity. A strong academic result can have poor feasibility.
Editorial workflow
- Start from a primary paper or an openly documented replication.
- Write the minimum rule without optimizing it.
- List information that must have been known at each historical timestamp.
- Identify costs, constraints and plausible failure regimes.
- Link the DOI or institutional source and record the review date.
What we do not do
We do not manufacture live performance, invent author experience, sell a signal, select only favorable studies or claim that an evidence grade predicts returns.