The view: down first to the floor of the measured zone, then a stepped recovery back to its ceiling. The grey path drops to about 0.06967 - the bottom of the 50% core - and then climbs in stages, with pullbacks along the way, to 0.07554, which sits just under the core top at 0.07567.
Say plainly what that shape is: a full round trip across the core zone inside one horizon. Both edges get visited. That is a much bigger claim about the route than an endpoint sitting inside the band suggests, and it should be judged as such. The endpoint is unremarkable - it is where the band already was. The journey is the actual assertion, and it is the part most likely to be wrong.
MEASURED, 24-bar horizon:
Median 0.07276
Core 50% 0.06967 - 0.07567
Wide 80% 0.06657 - 0.08006
125 independent windows, not the 3,001 overlapping ones
WHAT SUPPORTS A RANGE READ: skew is +0.44 here, a RIGHT tail, which is unusual - most of what I measure leans left. And the band widens more slowly than diffusion predicts: 8.3% at 24 bars, 10.4% at 48 (1.25x where root-t expects 1.41x), 13.5% at 72 (1.63x against 1.73x). Slower-than-root-t scaling is what rangebound behaviour looks like in the numbers, and it argues for travel back toward the middle rather than a sustained break in either direction.
A CAVEAT I WILL NOT HIDE: the regime line reads range with no trend, while ADX prints 28.9, which is nominally strong. Those two disagree. I am siding with the range read because the width scaling supports it, but a strong ADX in a market I am calling rangebound is exactly the sort of tension that resolves against whoever ignored it.
WHERE THIS BAND IS WEAK - AND IT IS THE OPPOSITE PROBLEM FROM MOST: out-of-sample coverage runs 67.3% in the core against a 50% target, and 94.3% wide against 80%. This band is too WIDE, not too narrow. Over-coverage is a failure in the same way under-coverage is: a range that swallows almost everything can never be caught being wrong, and being uncatchable is not the same as being right. So price finishing inside these zones should impress nobody, including me.
Also worth stating: sigma times root-t on this instrument comes out at 0.75x the empirical width. The textbook band here is materially too narrow - the reverse of what it does on gold, where the same formula runs too wide.
Invalidation: acceptance below 0.06657, the 80% floor. That is where the range read stops being a read.
The forecast is the band. The drawn path is one illustration of a route through it, and a band is scored on coverage across many calls, not on one arrow.
Educational content - not financial advice, and not a betting tip.
Say plainly what that shape is: a full round trip across the core zone inside one horizon. Both edges get visited. That is a much bigger claim about the route than an endpoint sitting inside the band suggests, and it should be judged as such. The endpoint is unremarkable - it is where the band already was. The journey is the actual assertion, and it is the part most likely to be wrong.
MEASURED, 24-bar horizon:
Median 0.07276
Core 50% 0.06967 - 0.07567
Wide 80% 0.06657 - 0.08006
125 independent windows, not the 3,001 overlapping ones
WHAT SUPPORTS A RANGE READ: skew is +0.44 here, a RIGHT tail, which is unusual - most of what I measure leans left. And the band widens more slowly than diffusion predicts: 8.3% at 24 bars, 10.4% at 48 (1.25x where root-t expects 1.41x), 13.5% at 72 (1.63x against 1.73x). Slower-than-root-t scaling is what rangebound behaviour looks like in the numbers, and it argues for travel back toward the middle rather than a sustained break in either direction.
A CAVEAT I WILL NOT HIDE: the regime line reads range with no trend, while ADX prints 28.9, which is nominally strong. Those two disagree. I am siding with the range read because the width scaling supports it, but a strong ADX in a market I am calling rangebound is exactly the sort of tension that resolves against whoever ignored it.
WHERE THIS BAND IS WEAK - AND IT IS THE OPPOSITE PROBLEM FROM MOST: out-of-sample coverage runs 67.3% in the core against a 50% target, and 94.3% wide against 80%. This band is too WIDE, not too narrow. Over-coverage is a failure in the same way under-coverage is: a range that swallows almost everything can never be caught being wrong, and being uncatchable is not the same as being right. So price finishing inside these zones should impress nobody, including me.
Also worth stating: sigma times root-t on this instrument comes out at 0.75x the empirical width. The textbook band here is materially too narrow - the reverse of what it does on gold, where the same formula runs too wide.
Invalidation: acceptance below 0.06657, the 80% floor. That is where the range read stops being a read.
The forecast is the band. The drawn path is one illustration of a route through it, and a band is scored on coverage across many calls, not on one arrow.
Educational content - not financial advice, and not a betting tip.
註釋
NOTE - the first leg landed, almost on the level. The second leg is the whole claim and it has notstarted.
Price 0.07012, from 0.07323 at publication. Down 4.25%. The drawn path called for a drop to the
50% core floor at 0.06967 before turning; the low printed 0.06932 - through the level by half a
percent - and price has since ticked back to 0.07012. As a called turning point that is close.
WHAT REMAINS IS THE BIGGER HALF. The path then climbs in stages to 0.07554, which from here needs
+7.7%. Nothing about the completed leg makes that more likely; the round trip across the zone was
always the ambitious part of this idea, and it is entirely unproven. I said at publication that the
endpoint was unremarkable and the journey was the real assertion. Half a journey is not a result.
THE CAVEAT I PUBLISHED IS RESOLVING AGAINST ME, PARTLY. At seal the panel read range with no trend
while ADX printed 28.9 - nominally strong - and I sided with the range read, writing that a strong
ADX in a market I am calling rangebound "usually resolves against the person ignoring it." The move
down was decisive rather than rangebound. So the trend side of that disagreement has been the more
useful one so far, and the range thesis now rests entirely on the recovery leg arriving.
THE BAND, AND WHY ITS BEHAVIOUR HERE PROVES LITTLE: price briefly left the core and came back. On a
calibrated band that is ordinary. This band is not calibrated - out-of-sample coverage reads 67.3%
against a 50% target and 94.3% against 80%, which is padded, and it is the same defect we found
across our whole live board this week: 31 of 36 resolved forecasts inside stated 50% zones, 86%,
traced by walk-forward test to missing regime conditioning and now fixed in the forecast engine.
The script drawing these zones has not been updated yet.
Invalidation unchanged: acceptance below 0.06665, the 80% floor, which is intact.
Educational content - not financial advice, and not a betting tip.
註釋
Note, 29 July: the first leg has played out as drawn. The second one, which is the actual claim, has not started.The idea described a round trip across the 50% zone: down through it first, then back up toward 0.0755. DOGE has done the first half. It worked down through the zone and is now 0.06933, having traded as low as 0.06882.
Against the published levels:
- wide 80% 0.06665-0.08016: holding, 3.9% above the floor.
- core 50% 0.06967-0.07567: price is 0.49% BELOW the lower edge. The descent slightly overshot the zone, so the core is currently out.
- median 0.07276: price is 4.7% below.
- stated target 0.0755: requires +8.9% from here. Not started.
So the honest read is half a structure, not a completed one. The shape of the decline matched what was drawn, and the outer band absorbed it, but the position is long and it is currently underwater against its own core zone. That is worth stating plainly now rather than after the fact.
What would make this idea right is the return leg, and nothing about the first half guarantees it. A round trip that only completes the outbound journey is a drawdown with a story attached.
Levels were fixed and hashed three days ago, before any of this.
Educational content - not financial advice.
手動結束交易
Closing this manually, and scoring it a miss on direction.The call leaned long for a round trip back up through the 50% zone. Nineteen days on, only the outbound leg happened: price fell and never made the return. DOGE is near 0.0700 now.
Measured against the levels published with the idea:
stated target 0.07554: not reached, roughly 8% above here, never started.
median 0.07276: price sits about 3.8% below it.
core 50% (0.0707 - 0.0751): price is just under the lower edge, so the core is currently out.
wide 80% (0.0685 - 0.0773): price is back inside it, low in the range.
So the honest read is half a structure, not a completed one. The shape of the decline was in the family of what the band allows, but the position is long and underwater against its own core zone, and that is worth stating plainly now rather than after the fact. A round trip that only makes the outbound leg is a drawdown with a story attached.
What would have made this idea right is the return leg, and nothing about the first half guarantees it. Levels were fixed and hashed at publication, before any of this.
Educational content - not financial advice.
免責聲明
這些資訊和出版物並非旨在提供，也不構成TradingView提供或認可的任何形式的財務、投資、交易或其他類型的建議或推薦。請閱讀使用條款以了解更多資訊。