📊 Closing Bell: $637.32, +$8.72, +1.39%
Post-close follow-up to the 2:50 PM piece, and the tape decided to settle the argument in the last seventy minutes of the session. When I filed that update, QQQ was sitting at $634.28, the 60-minute noon fade had just been fully absorbed, and the question I left on the table was whether the bid underneath $632 was patient enough to drag price to a new session high into the bell. Answer: yes, and then some. QQQ printed $637.14 at the 3:30 PM bar, marked $637.32 on the close, and tacked on another +$3.04 from where the 2:50 PM post caught it. Day eleven of the EMA override finishes green by +$8.72 on the session (+1.39%), and the daily bar closes at the high of the day.
The score did something I actually wanted to see: it moved. After sitting on 4.94 all morning and grinding to 4.93 at 1:21 PM, the engine lifted one tick to 4.95 at 2:54 PM ET with QQQ at $636.01, then cracked lower to a new cycle low of 4.91 at 3:35 PM ET with QQQ at $637.14. That's not noise. That's the price-level math doing exactly what it's designed to do — as QQQ punched through each successive overbought threshold into the close, the engine subtracted points. The 4.91 print is the lowest reading of the entire override window that began April 4. Deeper Extreme Risk-Off, at the exact moment QQQ is printing a new closing high. The pure signal and the trend filter are at peak disagreement, and the trend filter just booked its eleventh win in a row.
Final Recommendation: 100% QQQ (EMA override holds). Today's EMA 70 closes at $604.54. Cushion to the trend filter: $637.32 − $604.54 = $32.78. That's the widest cushion of the entire cycle — about $3 bigger than where yesterday's close left it, and $5 wider than the lunch-fade low. Ref remains 4.94 from April 11, the 0.07 rule still targets 5.01 on the upside exit, and no gateway fired — the engine was already maxed in the 100% SQQQ zone, so the 4.95 → 4.91 dip doesn't add anything it doesn't already have.
⏱️ Day Eleven in Four Trades — And What the 2:54 PM Tick Actually Was
The afternoon's score path is worth walking through, because it captures the full 2026 override story in a ninety-minute window. At 1:21 PM the score ticked to 4.93 with QQQ at $633.34 — that was the post-fade read from the 2:50 PM piece. Then at 2:54 PM the engine lifted back up to 4.95 with QQQ at $636.01. On paper that's the exact kind of fractional lift I was watching for in the lunch post — a tiny nudge toward the upside exit at 5.01. But it lasted forty-one minutes. By 3:35 PM the score had rolled straight back down to 4.91 with QQQ at $637.14, and that's where it closed the day's book.
The 4.95 tick was not a conviction change. It was a price-level artifact — a single input in the engine probably rolled over intraday (a volatility read, a short-end rate print, something with a noon-hour refresh) and briefly offset one of the overbought subtractions the tape was piling on. When QQQ tacked on another dollar and a half into the 3:30 bar, the overbought math reasserted itself and then some. A -0.04 drop from 4.95 to 4.91 in forty-one minutes during a +$1.13 push in QQQ is the clearest signature you'll get that this is a price move, not a macro move. The data did not get better. The price got more expensive relative to what the data justifies. The model is not confused — it's cataloguing, in real time, exactly how far the tape has outrun the fundamentals.
Treat the afternoon as one trade, because that's what it was: the override entered the session long at yesterday's $628.60 close, held through the noon fade to $632.37, held through the re-tag at $634.28, and is now marked out at $637.32. That's a +$8.72 / +1.39% day for the override position against a pure signal that spent the entire session insisting the right trade was 100% SQQQ. Eleven sessions in a row where that trade would have bled. The math isn't close anymore.
🎯 My Take: The Widest Divergence of the Cycle, and Why I'm Not Celebrating
Let me be honest about what today actually was. At 2:50 PM I was writing about a noon fade that got absorbed and a cushion rebuilt to $29.82, and framing it as the override doing a clean job. By 4 PM the cushion is $32.78, the close is at the high of the day, and the score is cracking to a new low. If you're scoring this strictly on P&L, the trend filter just delivered another textbook win on a position the pure signal said not to take. And yet — I want to flag something, because it is my job to flag it.
The thing that makes me pause is not the direction. It's the shape of the disagreement. When QQQ closed at $617.39 on Monday with the score at 4.94, the override worked. When it closed at $628.60 yesterday with the score at 4.93, the override worked. Now it's $637.32 with the score at 4.91, and the override has worked for eleven consecutive sessions while the pure signal has been getting louder, not quieter, in its objection. The engine is telling me — with more conviction every day — that the macro substructure under this rally has deteriorated, not improved. The tape has booked $20 of upside against a model that is simultaneously marking everything lower. That's a coiled spring, not a trend. And I've seen coiled springs resolve both ways: either the data finally catches up and blesses the price (in which case the score lifts through 5.01 and nobody talks about the divergence again), or the spring unwinds in one session and the override pays back three weeks of cushion in three hours.
Which one is it going to be? I genuinely don't know. What I do know is that the April 21 ceasefire clock is six trading sessions away, that the Pakistan-talks rumor the mid-day post covered is still a rumor and not a signed framework, and that every dollar QQQ prints above $637 from here is a dollar the model likes less, not more. If tomorrow is another melt-up and the score drops below 4.90 into the 4.85-4.87 zone, the divergence goes from historically wide to genuinely unprecedented for this cycle. That is not a prediction. That is a scenario I'm writing down tonight so that if it happens tomorrow, I can come back to this paragraph and say I saw it coming. The trend filter is still the trade. But the size of the disagreement is now the story, and the story has an expiration date.
⚠️ Bottom Line
Day eleven closes at $637.32, +1.39% on the session, at the high of the day. The score printed a new cycle low of 4.91 at 3:35 PM with QQQ at $637.14 — its loudest objection yet, registered at the exact moment the override booked its biggest win. Cushion to EMA 70 ends the day at $32.78, the widest of the cycle. Ref unchanged at 4.94, exit still at 5.01, and nothing about the final recommendation moves: 100% QQQ. Six sessions to the ceasefire date. The trend filter is still right, and the pure signal has never been louder about why it thinks it won't be for much longer.