Skip to main content

Methodology

Last updated: November 18, 2026

Every number on Redstone comes from the same closing prices and the same formulas, whether it shows on a ticker page, a comparison, a portfolio, a screener, a ranking or in an AI assistant. This page gives each formula as it is computed, and the reasons a figure here can differ from the same metric on another site.

Data

  • Prices, dividends and splits come from Massive (formerly Polygon.io). They are end-of-day figures, updated every trading day after the US market closes. Nothing is real-time.
  • Figures over a year or less are built on daily closes. Five-year figures are built on weekly closes, the last close of each week.
  • History goes back up to five years.
  • An ETF's expense ratio, objective and strategy come from its prospectus filed with the SEC.
  • A closed-end fund's objective and expense ratio come from its latest annual report filed with the SEC. The expense ratio is all it spent over its last fiscal year, borrowing costs included, as a share of its net assets, before any income taxes. Until a report gives that figure for a full year, as for a fund in its first months, we show the estimate in the fund's prospectus. Its market cap is its latest close times the shares outstanding in its newest SEC filing that gives them.
  • Where the feed lacks a dividend, lists one twice or misses a spin-off, we correct the record and check each correction against a second source.

Returns

Every return is a total return: the price change plus dividends, with each dividend reinvested on its ex-dividend date. Prices are adjusted for splits and spin-offs. In the formulas below, a security's value is its close on that basis.

Return = (ending value ÷ starting value − 1) × 100

  • 1D starts from the previous close.
  • YTD starts from the last close of the previous year.
  • 6M and 1Y start from the last close on or before the same date six months or a year earlier.
  • 5Y starts from the last weekly close on or before the same date five years earlier, and is shown as a yearly rate.

5Y return a year = ((ending value ÷ starting value) ^ (1 ÷ years) − 1) × 100

years = days between the two closes ÷ 365

Each window ends at the security's own last close, not at today's date.

Volatility

Daily return = (value ÷ previous day's value − 1) × 100

Volatility = standard deviation of the daily returns × √252

  • The daily returns are those of the past year.
  • The standard deviation divides by the number of returns, not by one less.
  • 252 is the number of trading days in a year, which turns a daily figure into a yearly one.
  • Five-year volatility uses weekly returns and √52.
  • The volatility chart shows the same figure over a moving window of about a month: 21 trading days, or 4 weeks on the five-year chart.

Sharpe ratio

Sharpe ratio = (average daily return × 252 − 3.7%) ÷ volatility

  • It uses the past year of daily returns.
  • 3.7% is the risk-free rate, fixed at about what Treasury bills yield.
  • The return in the formula is the average daily return times 252, not the 1Y return shown beside it, so working the ratio out from the 1Y return gives a different number.

The Sharpe ratio guide explains how to read it.

Sortino ratio

Sortino ratio = (average daily return × 252 − 3.7%) ÷ downside deviation

Downside deviation = √(average squared shortfall) × √252

A day's shortfall is how far its return fell below the daily risk-free rate (3.7% ÷ 252), and zero on a day it didn't. The average is taken over every day of the past year, not only the days with a shortfall. Sharpe vs. Sortino covers when to use each.

Max drawdown

Drawdown on a day = (value − highest value so far) ÷ highest value so far × 100

Max drawdown = the lowest drawdown of the past year

It is the largest fall from a peak to a later low within the past year, on daily closes with dividends reinvested. The drawdown chart shows the same measure over the period you pick.

Calmar, Omega and Martin ratios

Calmar ratio = 1Y return ÷ size of the max drawdown

Omega ratio = sum of the days' gains over the daily risk-free rate ÷ sum of the days' shortfalls under it

Martin ratio = (average daily return × 252 − 3.7%) ÷ Ulcer index

Ulcer index = √(average squared daily drawdown)

All three cover the past year. The Ulcer index grows with both the depth of the drawdowns and the time spent in them.

Beta and alpha

Both are measured against SPY, the S&P 500 ETF, over the past year of daily returns. An excess return is a daily return less the daily risk-free rate.

Beta = covariance of the excess returns with SPY's ÷ variance of SPY's excess returns

Alpha = (average excess return − beta × SPY's average excess return) × 252

Only the days both have a close are counted. Alpha is in percent a year.

Dividend yield

Dividend yield = regular dividends with an ex-dividend date in the past 12 months ÷ latest close × 100

  • Special and supplemental payouts are left out of the yield, because they would overstate what a security usually pays. They still count in its total return.
  • A portfolio's yield is its holdings' yields, weighted.
  • A security that pays dividends in another currency shows no yield. Its dividends still count in its total return.

Correlation

Correlation is the Pearson correlation of two return series, over the days both have a close. It runs from −1 (they move opposite) to 1 (they move together).

  • The one-year figure uses daily returns.
  • The five-year figure uses weekly returns, and shows only when both have five years of history.
  • Comparisons show both, because a single year can sit far from the usual relationship.
  • A portfolio's matrix of its holdings follows the window its chart shows.
  • Correlation to the S&P 500, as on the low correlation rankings, is the one-year figure against SPY.

Portfolios

  • A portfolio is bought at its weights at the start of each window (6M, YTD, 1Y, 5Y), then rebalanced back to those weights at the first close of each calendar quarter.
  • No trading costs or taxes are taken out.
  • Its volatility, Sharpe ratio, drawdown and the other risk figures come from the portfolio's own daily value, by the formulas above.
  • A figure shows only when every holding has closes for the whole window.
  • The optimizer looks for the weights, at least 1% in each holding, that would have given the highest Sharpe ratio, the highest return or the lowest volatility over the past year, with the same quarterly rebalancing. It describes the past year, not the next one.

New listings

A security listed under a year ago gets its year figures from its first close, once it has about three months of closes. Those figures carry an asterisk and the date they start from. Where securities are set side by side over a full year, as in the rankings, a new listing is left out.

Why a number can differ

The same metric often reads differently on two sites, and both can be right. These are the usual reasons.

  • Dividends. Most price charts, Yahoo Finance's among them, show the price change alone. Returns here add reinvested dividends, so a dividend payer's return is higher than its price change.
  • Period and sampling. The Sharpe ratio, volatility, beta and alpha here use daily returns over the past year. Morningstar calculates its Sharpe ratio from monthly returns, usually over three years or longer, and Yahoo Finance's beta for a stock uses five years of monthly returns. A fund can look very different over a different stretch.
  • Risk-free rate. We subtract a fixed 3.7% a year. A site that uses another rate, or none, gets a different Sharpe ratio from the same returns.
  • Dividend yield. Ours looks back 12 months and leaves out special payouts. Other sites may show a forward yield (the latest payout multiplied out to a year) or a fund's 30-day SEC yield.
  • Rebalancing. Portfolios here are rebalanced every quarter. A backtest that rebalances monthly, yearly or never ends with a different return from the same holdings.
  • Max drawdown. Ours covers the past year. Many sites report the worst fall since a fund launched.
  • Timing. Figures here are as of the last close. A site with live prices will differ during the trading day.

Corrections

If a number here looks wrong, tell us which one and where you saw it. Write to support@redstonemarkets.com.

Not investment advice. Redstone is for informational and educational purposes only. Past performance doesn't predict future results.