Methodology

Everything shown on FindMF is computed from publicly available NAV data and regulatory disclosures using the formulas on this page. We want our math to be reproducible. If a metric looks wrong, this document tells you exactly how it was produced.

Canonical scheme identity

Every mutual fund is issued with one scheme code per plan/option combination (Direct/Regular x Growth/IDCW/Reinvestment). We collapse these to a canonical layer so “HDFC Flexi Cap Fund” appears once, not four times. Each canonical scheme has:

• Primary variant - the display/identity variant (Direct-Growth preferred, with fallback).
• Performance variant - the Growth variant used for all return and risk math. If a scheme has no Growth variant (rare), its performance metrics are suppressed.
• All variants - shown on the detail page for reference but never mixed into performance computation.

Return formulas

Trailing returns are anchored to exact calendar dates, not month indices:

• 1M / 3M / 6M / 1Y = (NAV_latest / NAV_lookback) - 1, where NAV_lookback is the last available NAV on or before (latest_date - N days).
• 3Y / 5Y CAGR = (NAV_latest / NAV_Nyrs_ago)^(1/N) - 1.
• Since-inception CAGR uses the variant's actual earliest NAV date, not the scheme's overall inception (Direct plans only launched in Jan 2013, so a 2008-vintage scheme has a shorter Direct history).

IDCW and IDCW-Reinvestment variants are excluded from all performance math because their raw NAV drops on payout dates, which corrupts return calculations without distribution-event data.

Risk metrics

Risk metrics are computed from completed-month returns (the current calendar month is excluded because it is partial):

• Volatility = annualized standard deviation of monthly returns = std(r_m) * sqrt(12).
• Sharpe = (mean(r_m - rf_m) / std(r_m)) * sqrt(12), with rf_m = 0.07/12 (monthly equivalent of 7% annual India G-Sec).
• Sortino = (mean(r_m - rf_m) / downside_dev) * sqrt(12), where downside_dev = sqrt(mean(max(rf_m - r_m, 0)^2)).
• Max drawdown = largest peak-to-trough decline of the daily NAV wealth index. Always shown as a negative percentage.
• Calmar = annualized_return / |max_drawdown|.
• Win rate = % of completed months with positive return.

Minimum observation thresholds

Metrics with too few data points are suppressed (shown as -) rather than computed on thin history:

• Volatility, Sharpe, Sortino, Max Drawdown, Win Rate - at least 12 completed months.
• Calmar, Beta, Alpha, Info Ratio, Up/Down Capture - at least 24 completed months.
• 3Y CAGR - at least 36 completed months. 5Y CAGR - at least 60.

Additionally, Sharpe and Sortino are suppressed when annualized volatility is below 1%. Liquid, overnight, arbitrage and fixed-maturity schemes sit near the risk-free rate with near-zero dispersion, which makes these ratios mathematically unstable (tiny excess return divided by tiny denominator) and semantically meaningless (they are cash equivalents, not risk-adjusted bets).

Benchmarks

Each equity SEBI sub-category has a default benchmark (e.g., Large Cap -> NIFTY 100 TRI, Flexi Cap -> NIFTY 500 TRI). Beta, Jensen's Alpha (annualized, arithmetic), Information Ratio, Tracking Error, and Up/Down Capture are computed against the scheme's default benchmark when three conditions are met: (a) the benchmark series is available, (b) the scheme's monthly history overlaps with the benchmark's monthly history for at least 18 months, and (c) the scheme has a Growth variant.

Benchmark metrics use a trailing window of 36 aligned months when available, falling back to the latest contiguous window. We treat 24 aligned months as the confident threshold; in the 18-23 month band the metrics are still shown but marked provisional (a short-history flag and a low-confidence note on the figures), because a regression slope on fewer than two years of monthly data has a wide standard error. Below 18 aligned months we null all benchmark-relative metrics. Some funds carry a broad-market proxy benchmark (e.g. solution-oriented plans mapped to NIFTY 500 TRI) rather than their exact SEBI benchmark; those are flagged so their alpha/beta read as approximate.

Benchmark coverage: we currently cover TRI (Total Return) variants of ~16 NSE equity indices. Debt and hybrid schemes have default benchmarks that are not yet in our feed; those schemes show no alpha/beta. Passive products (ETFs, index funds) use their generic category benchmark as a crude proxy rather than the fund's actual declared benchmark.

Data quality flags

Each scheme can carry one or more data-quality flags. Flagged schemes remain visible but are marked on the detail page so you can judge whether the metrics are trustworthy. Reasons include:

• Scheme has no Growth plan; all performance metrics suppressed.
• No NAV rows could be parsed for the performance variant.
• Fewer than 12 completed months of NAV history.
• At least one calendar-month gap in the NAV series.
• Latest NAV is older than 30 days from the as-of date.
• At least one monthly return above 100%.
• A NAV row at or below zero (dropped from the series).
• Possible corporate action detected: a single-day NAV step-change close to a common split factor (0.1, 0.2, 0.25, 0.5, 2.0, 4.0, 5.0, 10.0) AND the new level persisted the next trading day. Only flagged on ETF performance variants. Some metrics (since-inception, max drawdown, long-horizon returns crossing the event) may be distorted since our NAV is not corporate-action-adjusted.
• Scheme has a default benchmark but fewer than 24 aligned monthly observations with it.
• Scheme's default benchmark is not in our ingested feed.

Update cadence

• Daily: latest NAVs are appended.
• Weekly: scheme metadata (expense ratio, fund manager, AUM) refreshed where available.
• Monthly: category-level industry AUM refreshed from public regulatory disclosures.
• Per build: all metrics are recomputed from scratch using the latest data.

What this is not

FindMF does not sell mutual funds. We are not SEBI-registered investment advisors. Nothing on this site is investment advice. See the full disclaimer.