90 weeks: -37.02% from the first open to the last close, an average -0.39% a week

41 weeks closed above the week before, 45 below. The median week moved -0.17%, and a typical week's move either side of the average was 3.87%. The best week was +5.00% (week of 2025-10-13), the worst -5.26% (week of 2025-03-17). The second half of the window ran -0.27% a week against -0.51% in the first: about the same pace.

The slope: closes falling 0.18% a week, angle -10°, R² 0.14

A straight line through the weekly closes (on a log scale) falls 0.18% a week. On the page's fixed scale, where 1% a week is 45°, that is -10°. R² 0.14 says how closely the weeks hug the line: well below 1, so the weeks wander a long way from it and the slope says less. The line through the weekly highs runs -0.18% a week (-10°), the one through the lows -0.19% (-11°).

The shape: falling channel

Highs and lows both sliding at about the same pace: a steady downtrend. Week on week, 39 of 89 weeks made a higher high and 39 of 89 a higher low. The gap between the two lines went from 6.9% to 8.1%.

The patterns on the weekly candles: 8 in play, 8 with a record worth counting

Bollinger squeeze, then a break (just completed, up): four weeks later it went its way 348 of 822 times across 641 names; Gap at the open (just completed, up): four weeks later it went its way 98 of 169 times on this name; Move of X% within M candles (just completed, up): four weeks later it went its way 28 of 51 times on this name; Marubozu (completed recently, down): four weeks later it went its way 34 of 48 times on this name; Lower highs and lower lows (completed recently, down): four weeks later it went its way 15 of 27 times on this name; Moving-average cross (completed recently, up): four weeks later it went its way 8 of 13 times on this name.

Four weeks on, by the arithmetic: middle of the readings 3.38 (-1.23%), 48% chance the close is higher

From the last close of 3.42: if the window's average week carried on it would print 3.36 (-1.83%); if the close line's slope carried on, 3.40 (-0.73%). Four in five of this name's earlier four-week windows ended between 3.19 and 3.75 (-6.79% to +9.60%). The middle figure is the plain average of those three centres. None of these is a level the price is going to; they are what the record prints.

Our arithmetic from stored candles: no recommendation, no forecast of any price. The paths below are what the numbers would print if the past simply carried on, and the range is where this name's own earlier moves ended. The decision, and the risk, are yours.