QVM Quant

Methodology

A high-level overview of the models and calculations used across QVM Research.

Last updated: July 14, 2026

QVM Score

A relative quantitative score that ranks a security against its peer cohort using a blend of quality, value, and momentum inputs. Scores are normalized within cohorts so that a security is compared to similar securities rather than to the entire universe.

QVM Health

An absolute assessment focused on the security’s own risk-adjusted characteristics such as drawdown behavior, return consistency, and volatility profile. QVM Health is computed independently of QVM Score.

Performance calculations

  • Cumulative returns are used for horizons under one year.
  • Annualized returns (CAGR) are used for horizons of one year or more.
  • All returns use adjusted prices where available.

Risk analytics

  • Annualized Standard Deviation across 1M, 3M, 6M, 1Y, 3Y, 5Y and 10Y horizons.
  • Sharpe Ratio — excess return per unit of total risk.
  • Sortino Ratio — excess return per unit of downside risk.
  • Maximum Drawdown — largest peak-to-trough decline in the window.
  • Ulcer Index and Martin Ratio — depth and duration of drawdowns.
  • Calmar Ratio — CAGR relative to maximum drawdown.

Benchmark-relative analytics

Each security is compared to an assigned benchmark. Where a fund’s stated benchmark is not covered, a proxy benchmark is used and labeled as such. Beta, Alpha, R², Tracking Error, Information Ratio, and up/down capture are computed against the assigned benchmark.

Peer, category, sector and industry ranks

Ranks are computed within a cohort of comparable securities. Peer cohorts are defined by purity-focused rules (for example, sector-specific groupings for sector ETFs). Ranks depend on cohort composition and are recomputed as inputs change.

Style Box

Style classification uses holdings-level fundamentals aggregated up to the fund, applying cap-cohort ranks for size and net-margin/growth-relative ranks for style. Classifications are shown only when data confidence gates are met; otherwise the panel indicates a pending state.

Similar securities and replacement candidates

Similarity blends mandate, sector, risk profile, benchmark, and structural characteristics. Replacement candidates surface funds with comparable mandates. This is a comparison tool, not a suitability determination.

Data confidence and pending states

  • Values shown as “pending” are not yet published because inputs are missing or unverified.
  • Missing data is not converted to zero.
  • Proxy benchmarks are labeled.
  • Methodologies are versioned; results may change as models evolve.