A 4-phase systematic framework covering the full active SEBI mutual fund universe. Phases 1 & 2: 3-layer scoring model (pass/fail gates, quantitative scoring, analyst overlay) applied to 273 funds across 15 equity and hybrid categories. Phase 3: structured comparison tool across 20 sector and thematic themes (169 funds, not scored). Phase 4: solution-oriented and FoF categories (64 funds) with lock-in, fee-layering, and structural heterogeneity analysis. Built from ACE MF data (sourced from AMFI and AMC disclosures). Data: March 2026.
India's mutual fund industry has ~₹67 lakh crore in equity AUM across hundreds of schemes. Most retail investors pick funds based on trailing 1-year returns or star ratings from aggregators. The problem: trailing returns are backward-looking, and star ratings collapse multi-dimensional fund quality into a single number with opaque methodology.
This model was built during my time at Omega Portfolio Advisors / MoneyWorks4Me to power fund recommendations for Third Rock Wealth's (TRW) client-facing dossiers. The goal was a repeatable, auditable framework that separates genuine alpha generators from funds riding beta in a rising market.
The core design choice: score funds within their own SEBI category using percentile ranks, not absolute values. A large-cap fund generating 1.5% alpha is doing something meaningfully different from a small-cap fund generating 1.5% alpha. The benchmark difficulty, volatility profile, and capacity constraints are entirely different. Category-relative scoring is the only honest comparison.
Every fund in the starting universe passes through three sequential layers. Each layer eliminates or scores, so that the final output is a ranked shortlist with both quantitative rigour and qualitative context.
Binary filters that eliminate funds with structural disqualifiers. Funds must pass all gates to enter the scoring engine. The gates remove funds that are too small, too young, or too expensive to be taken seriously for long-term allocation.
Each surviving fund is scored on 5 weighted factors, producing a composite score from 0 to 100. All sub-metrics are converted to within-category percentile ranks before weighting, so the scoring is relative to peers in the same SEBI category.
Hybrid categories use adapted factor sets. Benchmark-distorted metrics (capture ratios, alpha, info ratio) are excluded where stated benchmarks are structurally invalid, e.g. arbitrage funds scored on cost and yield consistency only.
Numbers catch pattern, but they miss context. The qualitative layer adds a /10 analyst score based on dimensions that purely quantitative screens cannot capture. This overlay can upgrade or downgrade a fund's final rating.
After running 273 funds through the full scoring engine across 15 SEBI categories, several patterns emerged that challenge conventional fund selection logic, in both equity and hybrid universes.
81% of the 177 scored equity funds generated positive 3-year alpha. But positive alpha alone is not the bar — only 37% (66 equity funds) earned Strong Buy or Buy. Generating modest alpha in a rising market is not difficult. Doing it consistently, with strong downside protection and low volatility, is rare.
Of 273 scored funds, only 63 earned Strong Buy (23%) and 37 earned Buy (14%). The remaining 173 funds — 63% of the universe — scored Hold or below. The model is deliberately demanding: it separates genuine quality from noise, and most funds are noise.
Across all 8 equity categories, Sortino ratio, return per unit of downside deviation, is one of the clearest separators between top and bottom-quartile funds. Unlike Sharpe, Sortino only penalises downside volatility, not upside. Funds that protect capital during drawdowns compound more powerfully over time than those generating equivalent returns with higher drawdown depth.
Small Cap funds show the widest dispersion (11.3 to 97.1), while Mid Cap and Flexi Cap fund scores cluster more tightly. Small-cap alpha is highly manager-dependent: stock selection matters far more in illiquid segments where passive replication is expensive. The gap between the best and worst small-cap manager is the largest in the equity universe.
ELSS funds carry a mandatory 3-year lock-in, which means investors cannot exit during a drawdown. The model raises the Consistency weight from 25% to 30% (and reduces Alpha to 20%) for ELSS. A fund that is highly volatile but occasionally generates strong alpha scores lower here than a fund that beats its benchmark steadily — because the investor has no exit if volatility turns negative.
The gap between Regular and Direct plan expense ratios (ER Spread) is a direct proxy for distributor commissions. Funds with unusually high ER Spreads are spending heavily on distribution. This raises a question that quantitative models can flag but not answer: is AUM growth organic or fee-driven?
Across the 7 hybrid categories, benchmark-relative metrics (alpha, capture ratios, information ratio) are unusable for cross-fund comparison in 4 of 7 categories. Equity Savings, Multi Asset, and Conservative Hybrid funds are commonly benchmarked against pure bond indices (CRISIL 10Y Gilt), making their "alpha" arithmetically meaningless. Multi Asset funds use 9 different benchmarks including Gold, Silver, and MSCI World. The model detects and excludes distorted metrics per category, using absolute risk metrics (Sharpe, Sortino, Std Dev) as the scoring backbone for hybrid funds.
The 16 scored arbitrage funds returned between 5.82% and 6.23% over the trailing year, a spread of just 41 basis points. The ER spread across the same funds is 33 basis points. In a product where the strategy itself is commoditised, up to 80% of observable performance difference is attributable to cost alone. The model scores arbitrage funds primarily on ER% (50% weight) and yield consistency (40%), omitting Sharpe and Sortino entirely, both are artifactually elevated by near-zero volatility and provide no useful ranking signal.
Composite scores are mapped to 5 rating bands. The thresholds are calibrated to produce a demanding distribution where most funds fall below "Buy." Distribution shown across all 273 scored funds (equity + hybrid).
Phases 1 & 2 cover 15 SEBI fund categories across equity and hybrid segments, each fund scored against category peers using percentile ranks. Hybrid categories use adapted factor sets where benchmark-relative metrics are structurally unreliable. Phase 3 adds sector & thematic coverage as a comparison tool across 20 additional themes. Phase 4 covers solution-oriented funds (retirement, children's) and fund-of-funds (domestic + overseas), presented as structured reference tables given the structural differences that prevent standard scoring.
Equity Categories — 8 categories, 177 scored funds
The Nifty 100 is highly efficient. Genuine outperformers stand out starkly against significant underperformers in this category.
Largest category by fund count. No market-cap mandate means manager flexibility is high — and style drift risk is highest here.
AUM capacity constraints become material in mid cap. Large funds face liquidity drag that compresses returns.
Widest composite score spread of any category. Highest beta. Downside protection and capture asymmetry are the critical differentiators.
Tax-saving funds with mandatory 3-year lock-in. Consistency weighted at 30% (vs 25% default) to reflect locked-in investor exposure. AUM gate lowered to ₹300 Cr.
SEBI mandates minimum 25% each in large, mid, and small cap. The mandatory small-cap exposure introduces higher volatility than flexi cap funds with similar names.
Mandated 35% minimum in both large and mid cap. Less flexible than flexi cap but broader than pure large cap. Good risk/return profile for moderate risk tolerance.
Pure equity funds with a maximum 30-stock mandate. High-conviction portfolios. Idiosyncratic risk is structurally higher — a wrong sector call is amplified by concentration.
Hybrid Categories — 7 categories scored, 96 funds · 1 display-only
65–80% equity. Information Ratio leads (25%) to reward genuine benchmark outperformance. Raw alpha replaced by IR to correct for benchmark-selection distortion common in this category.
BAF and DAA treated as a single 36-fund peer group. Alpha gate removed, negative alpha is structurally expected. Capture Asymmetry (Up Cap % − Down Cap %) directly tests whether the dynamic allocation model is adding or destroying value.
11/13 post-gate funds use CRISIL 10Y Gilt as stated benchmark. All benchmark-relative metrics excluded from scoring. Std Dev (3x spread: 0.71–2.26) is the strongest differentiator, funds that cannot control volatility fail their own mandate.
Extreme benchmark fragmentation: Gold, Silver, Nifty 50, BSE 500, MSCI World, bond indices, 9 different benchmarks across 30 funds. Down Capture range of −237% to +7,675% confirms that capture ratios are noise. Age gate (≥3Y) leaves only 10 funds, a new category with limited track records.
75–90% debt. Two funds with NaN alpha (Nifty 50 Hybrid Composite Debt 15:85 benchmark has no data in ACE MF export) are treated as gate-pass, data gap, not underperformance. On a ~8–9% gross-return product, cost efficiency has outsized impact on net investor returns.
Entirely different framework. Delta-neutral product: no directional equity risk. Sharpe, Sortino, and Std Dev excluded, all are artifactually elevated by near-zero volatility. Return spread of 32–41 bps across all horizons. ER range of 33 bps explains most of observable performance difference.
Only 2 AMCs run this category. Both funds are under 3 years old with no 3Y CAGR data. Percentile ranking requires a meaningful peer group, not yet available. Revisit when category reaches ≥5 funds with ≥3Y history.
Phase 3, Sector & Thematic Comparison Tool
Not a scoring model. Sector and thematic funds cannot be ranked with a single composite, a Pharma score and an Infrastructure score are not comparable. Phase 3 is a comparison reference: 20 themes across 169 funds in universe, 134 AUM-gated (≥₹500 Cr), sorted by Sharpe within each theme. Benchmark-relative metrics are shown for reference only, benchmarks are inconsistent within most categories (e.g. many funds use Nifty 50 regardless of actual sector focus).
Sector Funds — 7 themes · 52 gated funds
Mixed benchmarks (Nifty Fin Svcs / Nifty 50 / BSE BANKEX). Sharpe comparison valid within theme, risk-adjusted returns vary significantly despite similar mandates.
Strongest Sharpe ratio of all sector themes. BSE Health Care is the appropriate benchmark, avoid comparing alpha across funds using Nifty 50 as stated benchmark.
Low Sharpe across the theme reflects IT sector underperformance in the 2022–2024 period. Tata Digital India is the oldest fund with the longest track record (10Y).
Heterogeneous benchmarks (BSE Infra / Nifty Infra / Nifty 50). Several older funds (18–22Y) with full track records. High dispersion in risk-adjusted returns.
Largest sector theme by fund count. Mixed SEBI classification (some as Sector Funds, some as Thematic). Nifty India Consumption is the most appropriate benchmark, ~40% of funds use Nifty 50 instead.
Benchmark diversity extreme: Nifty Energy / MSCI World Energy / Nifty Commodities / Nifty 50. Alpha comparisons across funds are meaningless. Use Sharpe only.
Small peer group. Sundaram Services (7.5Y) vs SBI Automotive (1.8Y). Peer comparison limited, use as directional reference only.
Thematic Funds — 12 themes · 82 gated funds
New SEBI category (post-2022 launch for most funds). Majority benchmark vs Nifty India Manufacturing TRI, more consistent than most thematic themes. Sharpe data based on <3Y history for 9 of 11 funds.
All funds benchmarked to Nifty 500 or Nifty 50, broader index, not sector-specific. Effectively a flexible equity strategy with a business cycle narrative. Most launched 2021–2022.
Strong Sharpe driven by PSU sector re-rating (2021–2024). SBI PSU is the oldest (15.7Y) and largest. ABSL PSU has grown rapidly (₹6,086 Cr). Returns highly correlated with government capex cycle.
Heterogeneous group, mandates range from structured innovation plays (ICICI Pru Innovation, ₹7,487 Cr) to defence (HDFC Defence, ₹8,097 Cr) to recently-listed IPOs. Not a comparable peer group.
Mixed ESG approaches: integration strategy, exclusionary, Shariah-compliant, ethical screens. Benchmarks range from Nifty 100 ESG to Nifty 50. Low AUM concentration, weak investor adoption.
Most consistent benchmark (Nifty Transportation & Logistics TRI) of any thematic theme, meaningful alpha comparisons possible within this group. Category launched 2022–2024.
Small peer group. ICICI Pru (₹2,437 Cr) vs HDFC (₹1,272 Cr) vs Tata (₹502 Cr). Housing theme has benefited from real estate cycle. Benchmarks inconsistent (Nifty Housing vs Nifty 50).
Young theme, all 5 post-gate funds have <3Y history. Sharpe is based on <3Y data. Low Sharpe reflects the 2024 momentum reversal that hurt this strategy. Do not compare to older category funds.
Diverse strategies, quantitative screening, multi-factor, minimum variance, quality. Cannot meaningfully compare a quant fund vs a minimum variance product. Use within sub-strategy only.
ABSL MNC is the oldest fund (26Y) with the longest track record. HDFC MNC and Kotak MNC are relatively new (<3Y). Nifty MNC is the appropriate benchmark for this theme.
Minimal peer group. SBI Comma (20.7Y, ₹999 Cr) vs ICICI Pru Commodities (6.5Y, ₹3,677 Cr). Very limited comparison value. ICICI has 3x the AUM but lower Sharpe.
Miscellaneous group, Nippon India Taiwan Equity (overseas), HDFC Defence, ICICI Pru Rural Opportunities. No meaningful within-group comparison. Displayed for completeness.
Phase 4, Misc & Special Categories
Solution-oriented and FoF categories with structural differences that prevent standard scoring. No composite scores, peer groups are too small and mandates too heterogeneous for percentile ranking. Funds presented as comparison references within logical sub-groups.
Benchmarks are structurally heterogeneous across plan types (equity plan vs debt plan). Sub-grouped by equity allocation: Equity-Oriented (65%+), Balanced (30–65%), Conservative/Debt (<30%). Sorted by Sharpe within each sub-group. Lock-in amplifies ER impact dramatically over a 20–30Y horizon.
Small N, mixed equity/hybrid/debt mandates, no composite scoring. AUM is highly concentrated: SBI Children's Investment Plan (₹5,354 Cr) holds ~74% of category AUM. Tata and SBI Children's Savings Plan have 24–30Y track records. Displayed for completeness.
5 sub-strategies, each structurally distinct. BHARAT Bond: passive ETF wrapper (ER 0.08% = wrapper only; underlying ETF ~0.0005%); government bonds, target-maturity structure, largest domestic FoF category by AUM. Income Plus Arbitrage: dynamic 60% debt + 40% equity/arbitrage allocation; classified as debt for tax. Silver ETF FoF: SIP-accessible silver exposure, no demat required. Passive FoF: multi-index or index-of-index wrappers. Active Multi-Asset: underlying active funds, highest fee drag. ER shown is wrapper cost only, underlying fund TER is deducted from NAV, not displayed separately. Compare ER within sub-strategy only; BHARAT Bond 0.08% vs Active Multi-Asset 1.51% are not comparable products.
No new investments permitted. RBI's USD 7 billion industry-wide overseas mutual fund investment cap was breached in February 2022. All 13 overseas FoFs have been closed to new lump sums and new SIP registrations since then. Existing investors can continue existing SIPs and can redeem freely, but cannot add fresh capital. RBI has not raised the cap. Displayed here for completeness and for existing investors monitoring performance. Two sub-types: Equity/Specialty (US equity, REIT, clean energy, Asia Pacific) and Overseas Debt (US Treasury ETF wrappers). Active overseas FoFs carry double fee drag: FoF ER (1.4–2.4%) plus underlying overseas fund ER, compounding to 2.0–2.5% total annually.
The full methodology is now consolidated in a single combined document (v4.0) covering all four phases. It includes the complete Phase 1 & 2 scoring engine (equity + hybrid, 273 funds), the Phase 3 sector & thematic comparison framework, and the Phase 4 solution-oriented and FoF reference tables, with all corrections applied. The original phase-specific PDFs are retained below as archived references.
Archived, Phase-Specific Documents
The underlying Excel model is not published. It contains proprietary scoring logic built during my work at Omega Portfolio Advisors and is used internally for TRW client recommendations. This page and the PDF appendix document the methodology and findings for portfolio demonstration purposes.
All fund data is sourced from ACE Mutual Fund, which aggregates data from AMC filings with SEBI and AMFI. The dataset contains 47 variables per fund across returns, risk ratios, portfolio composition, rolling returns, and benchmark data. Data is as of the most recent available filing period.
Key limitations: The model uses point-in-time snapshots, not time-series panels. Rolling return data is pre-computed by ACE MF at the 1-month frequency (1Y, 3Y, 5Y, 10Y windows). Alpha is calculated as fund CAGR minus benchmark CAGR, which is an arithmetic approximation, not a regression-based Jensen's alpha. Survivorship bias is partially present since the ACE MF universe only includes currently active schemes.
The model is designed for relative ranking within categories, not absolute return prediction. A "Strong Buy" rating means a fund scores well relative to peers on the chosen factors. It is not investment advice.