Find funds, download NAV history and build your research library.
Choose the exact scheme and plan before downloading. Search includes ETFs and mutual funds.
USD listings only. Full available adjusted-price history × historical INR per USD. This is market-price data, not NAV. Uses prior-date FX (maximum 7-day age); unavailable FX dates are omitted. Source access may be rate-limited.
Downloaded instruments appear below and in Compare Funds / Portfolio Backtest. FX audit CSV includes the USD price, FX rate, FX date and converted INR value.
| Instrument | Type / source | History starts | Latest date | Observations | Actions |
|---|
Compare buy-and-hold returns over a shared period.
Rotate into the strongest instrument and inspect every signal and trade.
Hold the top-ranked eligible assets in equal weights. A holding exits at or below the exit rank; retained holdings can remain outside the entry top N. Rebalance-date choices apply to monthly/quarterly reviews; weekly schedules use first/last common trading day. Signals execute at the next common close.
Buy-and-hold comparison only; does not change the rotation universe. Uses the same starting capital and full test period. If coverage is incomplete, the benchmark is marked unavailable.
At each rebalance, hold eligible assets only when their daily close is above their SMA; otherwise the unfilled allocation moves to cash. Cash return: 0%.
0 = off. A replacement needs to beat the current holding by this score gap unless that holding fails the exit-rank or SMA rule.
Periods use common trading observations. Positive weights are normalized to 100%.
Understand the histories, assumptions and calculations behind your results.
Full downloaded histories are stored in the shared research library. Updates persist across visits. Use Export CSV for a spreadsheet copy.
Update history downloads the full source history again and replaces the saved version only after a complete, validated response. Existing source histories remain available as separate instruments.
mfnav.in API documentation ↗Global Yahoo ETFs use USD adjusted closes multiplied by historical USD-INR (INR=X, INR per USD). Only USD-denominated ETF listings are accepted. The latest FX observation strictly before each ETF price date is used, at most seven calendar days old; no future FX or current-day unfinished bars are used. Available history is limited by the ETF/FX overlap. Export FX audit CSV to inspect every conversion. Raw closes and adjusted closes are saved alongside conversion data.
Yahoo adjusted closes reflect splits and distributions; no extra dividend cash is credited. These are return-proxy prices, not NAVs or executable quotes. US and Indian closes are asynchronous: mixed-market backtests remain daily-close research approximations and do not model intraday execution, settlement, FX conversion costs, foreign withholding tax or remittance restrictions. Source histories may be revised. The public Yahoo chart endpoint is not a guaranteed API service.
Yahoo adjusted-close definition ↗mfnav.in supplies fund NAVs. ETFs can trade at a premium or discount to NAV. Original GoldBeES and LIC data use exchange closes; Nifty Alpha 50 and NYFANG are index proxies. NYFANG data is already in INR.
NAVs include fund expenses. This simulation applies no additional fees, tax, exit load or slippage. Apart from Yahoo adjusted-close histories, it applies no dividend reinvestment adjustment. Distribution-paying plans are therefore not total-return comparisons.
The switching buffer is an absolute percentage-point difference between the leader’s score and the current holding’s score. Switch when the gap is at least the buffer; exact ties retain the holding. It applies at scheduled reviews, with next-common-close execution. Cash entries have no buffer. The winner failing its SMA still triggers cash; a held asset failing its SMA cannot be retained by the buffer. For ratio models, use the displayed model score versus the zero reference, not separate asset ROCs.
Optional downtrend protection: Off, 100-day or 200-day SMA. At each scheduled review, the highest momentum candidate must close strictly above its own SMA to be held; otherwise the entire portfolio holds cash at 0% return. The SMA averages the instrument’s most recent actual daily observations, including the signal date. Insufficient SMA history means cash. Re-entry follows the same rule at a later review. Exits execute at the next common close, not immediately on an intra-period breach. Lower-ranked funds are not substituted when the winner fails. This applies to GoldBeES too.
Allocation is long-only, 100% in one selected asset. ROC = current price ÷ lookback price − 1. Blended scores are weighted averages of those ROCs.
The initial signal uses the first eligible date in your range, with earlier history used for warm-up. Subsequent reviews occur at the last common observation in each calendar period. Weeks end Sunday; two-week groups are anchored to Monday 5 January 1970.
Orders execute at the next common available close. Cash earns zero before the first execution. Exact score ties retain the current holding. There is no HDFC-to-Alpha historical splice in this universe model.
Ratio strategies require two assets in displayed library order: the first is held when its ratio change, or ratio/EMA minus one, is positive; otherwise the second is held.
Tests use the overlapping history of all selected instruments. Signals require actual observations for every selected instrument. Daily valuation uses the union of dates and carries the last known price across holidays; no future values are used.
CAGR = (ending value / starting capital)^(365.25 / elapsed days) − 1. Drawdown = daily value / highest value to date − 1. Calendar-period returns are not called CAGR.
Trade ROI is exit price / entry price − 1. Win rate uses closed trades only. The final position is marked to market and shown as open. The selected benchmark is bought and held from the same test start with identical starting capital. It does not enter the rotation universe unless separately selected there. Full-period history is required; otherwise the benchmark is marked unavailable without shortening the strategy test. None disables benchmarking.