Method · 2026-09-08Why a backtest on today's index is not evidence
Most quantitative work below the institutional tier is tested against data that did not exist on the decision date. Two errors do most of the damage. The first is survivorship: a universe built from today's index members excludes every company that was delisted, acquired, or demoted along the way, and those are disproportionately the ones that performed badly. The second is restatement: fundamental data that has been revised after the fact is used as if the original figure had never been published.
Both errors flatter the result in the same direction, and neither is visible in the output. A clean-looking backtest with a high Sharpe ratio is exactly what these errors produce. An examiner does not need to understand factor models to ask the question that exposes them: what data was available on the date this decision was made, and can you show it?
The fix is not a better model. It is a data layer that preserves as-reported values with their original publication timestamps and keeps monthly snapshots of index membership including the names that later disappeared. That layer is expensive to build and boring to maintain, which is why it is rarely maintained below the enterprise price tier and why Marnello describes it before describing any factor. Our demo builds, generated on free data, disclose the survivorship problem on every page rather than hide it.
Method · 2026-09-08What has to be true before we publish a factor
A factor is a rule for ranking securities. Momentum ranks by trailing return; low volatility ranks by trailing realized volatility. The rule itself is public knowledge and nothing to be proud of. What matters is whether the ranking is built correctly and whether it has done, out of sample and after costs, what the literature says it should.
Marnello applies four tests before a factor is presented as a source of return. A Fama-MacBeth cross-sectional regression has to show a risk premium with a t-statistic above 2. Decile portfolios have to show a positive top-minus-bottom spread in a holdout period that was not used to build the factor. The spread has to survive 20 basis points per side of transaction cost. And the in-house series has to correlate with the corresponding public benchmark from the Kenneth R. French Data Library, so that anyone can check the construction independently.
A factor that fails is still computed and reported, because a client's exposure to it is a fact about their portfolio whether or not the factor earns a premium. It is simply labeled as a risk descriptor rather than a return source. In our September 2026 demo build, only one of four factors passed on the free-data universe. That is written in the validation record, not smoothed over.
Practice · 2026-09-08What "document your process" means in an examination
The request usually arrives as a list: for a sample of accounts, show the basis for the allocation, when it was decided, what information was considered, and who reviewed it. Firms with a policy manual and a good story tend to be surprised by how little of that counts. Policies describe what should happen; the examiner wants evidence of what did.
Quantitative evidence answers a narrow part of that request, but it is the part small firms most often cannot produce internally: a dated measurement of what the portfolio's characteristics actually were, with the method written down. A factor exposure report from the quarter in question, with its methodology appendix and a reviewer's sign-off, is a contemporaneous record. A spreadsheet built the week before the examination is not.
The same logic applies to marketing. A claim that the firm "applies a disciplined factor framework" is a capability claim, and the Marketing Rule requires that it be substantiated. The substantiation is the measurement, dated, with the method attached. This is the reason Marnello's products are built around documentation first and analysis second, and why every deliverable carries a governance record on its last page.
Insights are general research and commentary. They describe method and practice; they are not legal or compliance advice and not tailored to any firm. Standard disclosure applies.