Our Methodology: How Every Number Is Calculated

Every number on this site is computed by code, from public data, the same way for all 8,000+ schemes. Nothing is hand-picked, nothing is sponsored, and no fund house pays to look better. This page explains exactly how each number is calculated, in plain English, so you can decide for yourself how much weight to give it.

If you only read one thing: our two headline ratings, RS Rating and King Score, are both percentile ranks within a fund's own category. They never compare a small-cap fund to a liquid fund. They answer one question: against its true peers, where does this fund stand?

The data everything starts from

All analysis uses Direct plan, Growth option schemes. Direct plans carry no distributor commission, so they show the fund's true cost and return. Data refreshes every morning; NAV dates advance on business days.

The one idea behind our ratings: percentile rank

Imagine a class of 100 students. If you scored better than 90 of them, you are in the 90th percentile. It does not matter whether the exam was easy or brutal - the rank tells you where you stand among people who sat the same exam.

We do exactly this with funds. Every rating below starts by ranking a fund only against its own sub-category (Small Cap vs Small Cap, Gilt vs Gilt). This matters because raw numbers lie across categories: a 12% return is ordinary for a small-cap fund and outstanding for a corporate bond fund.

RS Rating (1 to 99): recent momentum

RS stands for relative strength. It answers: how strong has this fund been recently, versus its peers?

The exact recipe:

1. Compute a momentum score for every fund: (2 x 3-month return) + (1 x 6-month return) + (1 x 1-year return). The last quarter is deliberately double-weighted, so the rating notices when a fund heats up or cools down early.
2. Rank all funds in the same sub-category by that score.
3. Convert the rank to a 1 to 99 scale. 99 means stronger than roughly 99% of peers right now. 50 is dead average. 5 means almost every peer has done better recently.

This is the same construction stock traders have used for decades (popularised by Investor's Business Daily for US stocks). To our knowledge, no other platform computes it for Indian mutual funds.

Two companions to the rating:

• Market RS uses the same momentum score but ranks the fund against the entire growth universe (equity, hybrid, gold, international, index), not just its category. A debt fund can be RS 99 in its category and still show a modest Market RS, because it is not in a race with small caps.
• The RS line on our charts divides the fund's NAV by its benchmark index and rebases to 100. Line rising: the fund is beating its index. Line falling: it is lagging. When the line reaches its highest level in 52 weeks, we flag a new RS high.

A high RS is a description of what has already happened, not a promise. Momentum persists sometimes and reverses sometimes; that is exactly why we also compute a slow, long-term score next.

King Score (0 to 100): long-term quality

King Score answers a different question: over the long run, has this fund delivered good returns without wrecking its investors on the way?

It blends four percentile ranks, each computed within the fund's sub-category:

King Score = 0.35 x return percentile + 0.25 x consistency percentile + 0.20 x drawdown percentile + 0.20 x Sharpe percentile.

Worked example: a fund at the 90th percentile on returns, 80th on consistency, 60th on drawdown and 70th on Sharpe scores 0.35(90) + 0.25(80) + 0.20(60) + 0.20(70) = 77.5, rounded to 78.

Why these weights: returns get the largest share because that is what you ultimately keep, but 65% of the score is about how those returns arrived - regularly (consistency), without devastating falls (drawdown), and without wild swings (Sharpe). A fund that shot up 80% in one year and then bled for two can have a great 3-year CAGR and still score mediocre on King, by design.

For the full explainer with reading bands and common mistakes, see What is King Score?

Returns: three different lenses

• Trailing returns (1M to 10Y): simple point-to-point growth. Over one year we show the plain return; over longer periods we show CAGR, the constant yearly rate that gets you from the start NAV to today's NAV. Trailing returns depend heavily on the end date, so treat them as a snapshot.
• Calendar-year returns: each year's return separately, next to the fund's benchmark index for the same year. This is where you see how bumpy the ride really was, and we count it up for you: "beat its benchmark in X of the last Y calendar years."
• Rolling returns: the fairest lens. Instead of one start date, we test every start date. For 3-year rolling returns we compute the return of every possible 3-year holding period in the fund's history and report the median, the worst window, the share of windows that were positive, and the share that beat the Nifty 500. One lucky or unlucky entry cannot distort this.

SIP returns are computed honestly: we simulate investing ₹10,000 on the first NAV date of every month, at that day's actual NAV, and report the XIRR - the annualised return that accounts for every instalment's timing. SIP XIRR and lumpsum CAGR legitimately differ; both are shown.

Risk: how we measure the downside

• Maximum drawdown: we walk the entire daily NAV history, track the running peak, and record the deepest percentage fall from any peak to the subsequent bottom. This is the single most honest risk number a fund has: it is what an investor at the worst moment actually lived through.
• Best and worst 1-year stretch: we compute the return of every rolling 365-day window in the fund's history and show the best and worst, with their start dates. It answers: how much did timing alone ever matter?
• Volatility: annualised standard deviation of returns over the last year. Bigger swings, bigger number.
• Sharpe and Sortino (3Y): return earned per unit of volatility. Sortino counts only downside swings, since upside "volatility" is not something investors complain about.
• Beta, alpha and R²(3Y): computed by regressing the fund's returns against its mapped Nifty benchmark. Beta is sensitivity (1.1 means it moves 10% more than the index), alpha is the extra yearly return unexplained by the index, and R² tells you how well the benchmark explains the fund at all.
• Up and down capture: in months the index rose, how much of the rise did the fund capture? In months it fell, how much of the fall? The dream combination is up-capture above 100 with down-capture below 100.

Bull and bear behaviour

We split history into regimes using a simple, transparent rule: the market is bullish when the Nifty is above its 200-day moving average and bearish when below. Each fund's annualised return is then computed separately inside each regime. This shows you, before the next fall happens, how the fund has historically behaved when markets crack.

Holdings and overlap

Holdings come from each fund's latest monthly disclosure: top positions, sector weights and concentration (share of assets in the top 10 holdings). Portfolio overlap between two funds is the sum, over every stock they both own, of the smaller of the two weights. If two "different" funds overlap 45%, buying both mostly buys the same portfolio twice.

Backtests: built to talk you out of bad ideas

The Strategy Lab backtester is deliberately strict:

• Point-in-time: every decision on a simulated date uses only data that existed on that date. No lookahead, no survivorship shortcuts.
• Costs modelled: exit loads are deducted when a simulated sale triggers them. Results are clearly labelled pre-tax.
• Multiple start dates: every backtest runs from several staggered anchors, because a strategy that only works from one lucky start date is a coincidence, not a strategy. The spread across anchors feeds a Robustness Score (0 to 100) that grades how fragile the result is.

What we deliberately do not do

• No advice. Every page describes; none recommends. Nothing here is a recommendation to buy, sell, hold or switch any fund. We are not a SEBI-registered investment adviser or research analyst.
• No sponsorship. No AMC pays us, no fund can improve its score, and rankings cannot be bought.
• No post-tax modelling. Your tax depends on your slab, holding period and asset class; pretending otherwise would be false precision. All figures are pre-tax.

Honest limitations

• Every number here is computed from past prices and disclosures. Past performance does not guarantee future returns, and no formula changes that.
• Percentile ranks need peers: in tiny categories (fewer than a handful of funds) ranks and averages are noisy, and we suppress category averages below three funds.
• Third-party fields (expense ratios, managers, holdings) depend on public disclosures, which can lag by a few weeks.
• Ratings move. A fund's RS can drop fast in a correction; that is the rating working, not breaking.

See it in action

Open any fund page, for example from the Equity Funds directory, and you will find every number described above, computed for that fund. Or open the terminal with a free account to screen, chart and backtest across the full universe. New to all of it? Start with the 5-minute guide.