0–100, mostly how each is changing against the company’s own history, not its level
P/E (TTM) and P/B against the company's own history
The two disagree: P/B is 4.7σ further from its own 5Y norm than P/E.
Price against trailing twelve-month earnings — bands are the 5Y robust median ± σ
Price-to-Earnings ratio = stock price ÷ earnings per share. The number of years of current earnings the market is paying for one share. A higher P/E means investors expect more growth — or are paying more for the same earnings, depending on your perspective.
TTM EPS = trailing twelve months EPS, summed across the last four reported quarters. Updates only when a new quarterly filing lands, so the band edges step up or down each quarter.
The shaded band is the ±1σ̃ envelope around the fair-value line (price = TTM EPS × the company's 5Y median P/E). Dashed reference lines at ±2σ̃ mark the wider "typical range". Median + robust σ̃ (scaled median absolute deviation) is used instead of plain mean + σ because a single quarter of near-zero EPS can produce extreme P/E outliers (price ÷ 0.01 = 6000x) that wreck the stats. When today's P/E falls outside the ±2σ̃ envelope the chart's y-range extends so the price line stays on-chart.
+2σ̃ would be about two above it.The bands' height on any given day = multiple × that day's TTM EPS. When TTM EPS collapses (e.g. a cyclical trough), every band shrinks proportionally — but the price doesn't follow EPS one-for-one because the market often "looks through" the trough to expected recovery. Result: the price line floats above the entire band stack until EPS recovers. That's not a charting bug — it's the chart correctly showing that the company traded at multiples outside its own historical range during the trough, which is itself a useful signal (the market was willing to pay an unusually high P/E in anticipation of recovery).
Price-to-Book = stock price ÷ book value per share (shareholder equity ÷ shares outstanding — the accounting "net worth" behind each share). P/E reflects current earnings power; P/B reflects the balance sheet, and the two routinely disagree — a company can look expensive on P/E because earnings are temporarily depressed while still trading near book value. P/B is measured against its own history exactly like P/E (robust median ± σ̃ over the same window), so the two σ̃ numbers in the header are directly comparable; the sentence under them says how far apart they sit. It has no chart of its own because in price terms the P/B band is a near-twin of the P/E one.