Price range. For each name we pull ~14 months of daily closes and compute the standard deviation of the last 20 daily log returns — call it σ (a 1-day realized vol). The headline band is close × e^(±1σ). That is the classic options expected move: absent news, price holds inside it about 68% of sessions. The wider bar spans ±2σ using Yang-Zhang volatility, which reads the whole OHLC bar (overnight gap + intraday range) rather than closes alone and so tracks the tails better: measured over 14.7k ticker-days it covers 94.5% of next sessions against a 95.4% target, versus 93.2% for close-to-close. At 1σ close-to-close is the better fit, so the headline band keeps it.

Vol source (RV / IV). The Vol toggle rebuilds the band from either realized vol (what the stock has actually done — the default) or implied vol, using IV₃₀ ÷ √252 as the 1-session sigma. IV mode is the market's expected move — it's what options are pricing, and it widens ahead of earnings when realized vol can't see the event yet. Names without an IV quote fall back to their realized-vol band (noted under the bar).

Trend tier comes from the 15 / 50 / 200-day SMA stack — STRONG BULLISH when price > 15 > 50 > 200, down to BEARISH when price is below the 50 and 200. Grey SMA ticks on the bar show where those averages sit relative to today's range.

Reading it for options. The band sizes the move; it does not call the direction. A wider band = richer realized vol = pricier options / wider break-evens, so its job is strike selection and break-even math — where to set a spread, whether a premium is rich or cheap against what the stock actually realizes. Turn on ⚡ Intraday to drop a live-price marker (amber ▲) onto each band and see how far today's tape is stretched versus the close-anchored range.

What the edges are not. The band is centred on the last close, which is the random-walk assumption: the best estimate of tomorrow's price is today's. So the edges are not mean-reversion levels — there is no mean here to revert to, and the whole band re-centres on price every session. We tested the reversion claim on 13,112 ticker-days (12 names, 5 years) and it does not hold: forward 5-day returns after a −1σ touch are 0.54% against a 0.48% all-days baseline, with a 54.8% win rate against 54.8% baseline. That is zero edge, not a small one. Treat a tag of either edge as "the move was a normal-sized move," nothing more.

Forecast vol (HAR). Realized volatility is one of the few genuinely predictable things in a price series, so the RVol column and the IV/RV ratio use a forecast rather than a backward window. The model is Corsi's HAR: next-month volatility regressed on realized vol at three horizons — yesterday, the last week, and the last month — using Garman-Klass daily variance off the OHLC bar. Fitted once on a 63-name, 5-year panel (74,921 observations) and frozen. Walk-forward out-of-sample it predicts the next 21 sessions of realized vol at R² 0.54 against 0.29 for the trailing-20-day σ, beat the trailing window on 60 of 61 individual tickers, and coefficients fitted on 50 names outside this watchlist transferred to the watchlist twelve essentially unchanged — so it is not tuned to these names.

Why this matters for IV/RV. The ratio used to divide a forward-looking 30-day implied vol by a backward-looking 20-day realized vol. That mismatch biases it in a predictable way: a name reads "rich" simply because the last month was quiet, and "cheap" right after a spike the trailing window is still carrying. Both sides now cover the same forward month. Note the honest limit — we validated that the denominator forecasts realized vol better, not that the resulting ratio is profitable to trade; that would need option-price history we do not store.

Caveats. This uses realized vol, not implied — around earnings, IV (and the real expected move) will be wider than this shows. Realized vol also trails a regime change: the band is too narrow just after volatility jumps and too wide as it decays. Ranges rebuild on every new close. Educational tool, not a signal service.

Ticker

Symbol and company name.

Last / Δ

Most recent daily close, and its % change vs the prior close.

Trend

Tier from the 15 / 50 / 200-day SMA stack: STRONG BULLISH (price > 15 > 50 > 200) → BULLISH (above 50 & 200) → PULLBACK (above 200 but below the 15-day) → NEUTRAL → BEARISH (below 50 & 200).

Low −kσ

Lower edge of the range = close × e^(−kσ) — the down-side of the expected move, i.e. a decline of this size would be an ordinary session. Not a buy signal: reaching it carries no measurable edge (see What the edges are not). k follows the Band toggle (1σ / 1.5σ / 2σ).

Daily Price Range

The visual band. Seafoam→amber shading spans −kσ … +kσ; the white line is the anchoring close; grey ticks mark the 15 / 50 / 200-day SMAs; the scale ends are ±2σ on Yang-Zhang volatility. With ⚡ Intraday on, an amber ▲ shows the live price plus its position in the band.

High +kσ

Upper edge = close × e^(+kσ) — the up-side of the expected move. Symmetric with the lower edge by construction, and equally not a signal to fade.

Width / RVol

Width = the band as a % of price (round-trip size of the expected move). RVol shows two numbers: trailing 20-day realized volatility → the HAR forecast for the coming month (annualized). Wider = pricier options / wider break-evens. When the forecast sits well below the trailing figure, vol has spiked recently and is expected to decay; well above, it is expected to expand.

IV₃₀ / RV

IV₃₀ = 30-day at-the-money implied volatility (annualized), interpolated from the option chain — what the market is pricing. The ratio = IV₃₀ ÷ the forecast vol for that same forward month (see Forecast vol below), so both sides of the ratio look forward over the same window: ≥1.1 rich (options pricing more move than the stock has realized → selling premium relatively favored), ≤0.9 cheap (buying relatively favored), ~1 fair. Requires a Massive options key; shows "—" without one.