বিষয়বস্তু · 35 / 35
Read Sector and Size Neutral Factors
হালনাগাদ 22 আগস্ট 2026 · 3 মিনিটের পাঠ · 14 বিষয়বস্তু
সহজ ভাষায় অর্থ
Factor Neutral starts with Factor Lab's raw composite score and removes the part jointly associated with sector membership and log market capitalisation. The residual neutral score is what those controls did not explain; it is not a purer truth, causal alpha or recommendation.
Why it is useful #
It helps reveal whether a raw factor ranking mainly reflects sector or company-size exposure. The response publishes selected factors, universe and scored counts, sector and size adjustments, cohort coverage and whether size control could be fitted.
Where it appears and the workflow #
Open Sector and Size Neutral Factors, choose one equity exchange and supported Factor Lab inputs, then set a result limit. Compare raw and neutral scores, inspect sector/size adjustments and coverage, and open a listing's Factor Lab evidence before interpreting a rank. Treat rows without a neutral score as unavailable.
Calculation or source #
The service reuses Factor Lab's persisted filing and completed-price inputs, then fits sector indicators and demeaned log market capitalisation jointly. Neutral score is the regression residual. If size cannot be controlled, the response says so rather than implying a full neutralization.
Prerequisites, inputs, period and unit #
Prerequisites are a supported non-crypto equity exchange, enough scored listings, sector identity and valid market capitalisation for the requested controls. Raw and neutral scores are dimensionless model outputs; market cap retains the exchange's native basis. EvidenceAsOf and method version define the sample.
Worked example #
Illustrative only: a raw score of 78 can become a neutral score of 12 after a +40 sector contribution and +26 size contribution are removed. The listing did not lose market value; the second number answers a different relative-ranking question within that fitted cross-section.
What high and low mean #
Higher neutral score means a larger positive residual under the fitted controls, not higher expected return. A large size coefficient means the raw cross-section carried material size exposure. Sector adjustment direction describes the model decomposition and is not a sector forecast.
Positive, negative and zero #
Raw, neutral and adjustment values can be positive, negative or genuinely zero. A null neutral score means the row could not be neutralized, never that it sits at the cross-sectional centre. Missing sector or market-cap evidence remains visible through coverage and unavailable reasons.
Limitations and common mistakes #
Residualization removes only the specified linear sector and log-size exposures in one sample. It can amplify noise, depends on classification quality and does not remove industry, country, liquidity or other exposures. Common mistakes are comparing residual and raw scores as the same scale or calling neutral a risk-free factor.
Market-specific differences #
Every run is confined to one exact equity exchange. Sector vocabularies and filing coverage differ by venue, and unknown sectors remain unclassified. Scores from different exchanges or evidence dates are separate cross-sections and should not be combined into one ranking.
Suggested next steps #
Choose one factor set, record universe/scored/neutralized counts and compare five rows before and after control. Identify which changes come from sector and which from size, then write one limitation the model does not remove.
Educational use only #
This guide explains descriptive research evidence. It is not investment advice, a price prediction, a recommendation, a claim of predictive accuracy, or an instruction to buy, sell, rebalance or place an order.