📊 The 4:19 PM Tick: First Up-Move Off the 4.91 Cycle Low
Following up the 4 PM close piece because the engine wasn't quite done for the day. Forty-four minutes after the bell, at 4:19 PM ET, the score printed 4.94 — the first up-tick since the cycle low of 4.91 that hit at 3:35 PM with QQQ at $637.14. That's a +0.03 lift in under an hour, which in the context of an eleven-session override window that has mostly walked downhill is the first real sign of life on the upside from the pure signal.
Here's why it matters without overselling it: the score is now sitting back exactly at the ref. Ref is 4.94 from April 11, and the latest print is 4.94. The entire afternoon — 4.93 at 1:21 PM, the 4.95 blip at 2:54 PM, the 4.91 crack at 3:35 PM — collapses into a round trip to exactly where this position was entered four trading days ago. That doesn't change the allocation. It doesn't trigger a rebalance. But it does say that the very last datapoint of the session decided not to press the cycle-low narrative any further.
Final Recommendation: 100% QQQ (EMA override holds, unchanged). Pure signal at 4.94 is still Extreme Risk-Off / 100% SQQQ on its own. QQQ marked $637.32 on the close, EMA 70 at $604.54, cushion $32.78. Next meaningful triggers are unchanged: a move through 5.15 would clear the override entirely and flip the book to a real bullish signal; a close below $604.54 would kill the override and turn the full 100% SQQQ signal live. Neither is close. From 4.94 to 5.15 is 0.21 points of ground the engine has not been willing to give up in this entire cycle.
📈 Two New All-Time Highs, One Unhappy Dow
Zoom out from the QQQ tape for a second, because the broader close deserves its own line. The S&P 500 closed at 7,022.95, up 0.80%, a fresh all-time high. The Nasdaq Composite ripped +1.59% to 24,016.02, also an all-time high. The Dow slipped -0.15% to 48,463.72, going the opposite direction on a day the tech-heavy indices were setting records. That divergence is a tell — not a scary one yet, but a tell. Money rotated into growth and out of the cyclical/industrial bucket on the same session, which is the opposite of what a broad, healthy, economy-is-fine rally looks like.
Leadership was chips and software. Microsoft and Oracle extended; ServiceNow and Salesforce closed green; the semis continued the "chips are back" story that has been quietly running under this whole override cycle. The single loudest mover was Robinhood +10% after the SEC approved a new proposal on day-trading rules for retail investors — which, read charitably, is pro-retail-participation policy hitting a tape where retail is already tilted risk-on. Bank of America and Morgan Stanley both gained post-earnings; BofA's beat was the headline number and MS's capital-markets line was the quality tell. For a day that was largely about the US-Iran peace-deal hope from the mid-day post, the breadth inside the green indices was narrower than the headline would suggest.
The part nobody is quite saying out loud: the Nasdaq is at an all-time high while the Dow is red, credit spreads haven't participated in this rally the way an all-clear would suggest, and the pure macro signal the engine reads is sitting at a cycle-low 4.91 print from earlier this afternoon. New highs with internal friction is the most common setup for the kind of session that gives back a week in three hours. It's also the most common setup for the kind of grind where the leaders just keep leading and the laggards stay parked. Both are on the table. Neither is the override's problem tonight.
🎯 My Take: The +0.03 Is Not a Reversal. It Might Be a Pause.
I want to be careful with the 4:19 PM tick. A +0.03 lift off a cycle low in forty-four minutes of post-close tape is not a reversal. It is the engine re-reading whatever piece of price-level math cracked at 3:35 PM and walking one step back from the extreme. I've been writing all afternoon that the divergence had never been wider; the last print of the day says at the margin the engine stopped making it wider. That's a meaningful distinction and a thin one, and I don't want to dress it up as something bigger than it is.
Here is what I actually think the post-close lift tells us. The score spent the afternoon marking QQQ overbought at every successive dollar of the push from $634 to $637. At some point after the close, a short-term component of the engine — probably a volatility read or a short-end rate input that refreshes on the 4 PM mark — came in a hair friendlier than the intraday extreme implied. The engine did what a calibrated model is supposed to do: it booked the improvement and lifted, one tick at a time, without pretending the underlying picture had changed. That is not the score "giving up and joining the rally." That is the score saying "I'm still in the same zone I was at 1:21 PM, the afternoon low was an overreach, and 4.94 is where the balance actually sits tonight."
Which is why I'm writing this as an evening wrap and not a victory lap. The override is still right. Day eleven is in the books. The close at $637.32 still marks the position at the high of the cycle and the widest cushion of the cycle. And the score, for all the noise of the afternoon's 4.93 → 4.95 → 4.91 → 4.94 chop, finishes the session back at the ref. Same position. Same allocation. Same trade. The pure signal retreated further during the session and then recovered back to flat at the close — which is the closest thing to "the divergence stopped growing" that I've been able to point to in over a week.
🔭 Thursday Setup: Five Sessions to the Ceasefire Clock
Tomorrow is April 16. The April 21 ceasefire date is exactly five trading sessions away (Thursday, Friday, Monday, Tuesday, Wednesday). That calendar is no longer background texture — it is now the single biggest variable between tonight's close and whatever this tape does next. If the Pakistan-talks framework from the 1 PM post moves from rumor to confirmed before the weekend, the $637 close stops being a local high and becomes a base for a push into $640+. If the week ends with no announcement, every dollar of Friday's gain has to survive a weekend where the headline risk cuts the other way.
Earnings-wise, the BofA and Morgan Stanley beats set a reasonably constructive tone into the rest of bank earnings this week. Netflix is on deck. The chip leadership under this whole cycle is going to get tested when the first megacap semi prints numbers, and the tape is not priced for a miss. Economic data is light tomorrow — nothing on the calendar with the heft to move the score's macro inputs by itself. Which means Thursday is likely to be, again, a price-level session, with the engine reading exactly how far QQQ is willing to push above $637.32.
Concrete scenarios I'm watching: (1) QQQ pushes to $640+ on continued diplomacy optimism, the score cracks below 4.90 into the 4.85-4.87 zone, and the divergence goes to genuinely unprecedented for this cycle. (2) QQQ chops sideways in a $635-$638 range, the score grinds in a 4.92-4.96 band, and the override extends to day twelve without fanfare. (3) A diplomacy headline disappoints, QQQ fades $5-$8 back toward $630, the pure signal catches a small bid, and the week ends with the cushion narrowing for the first time. The override is agnostic to which one happens. The story — the one this blog is actually trying to tell — cares a lot.
💡 Bottom Line
Day eleven of the EMA override closes green and sets a new cushion high of $32.78. The Nasdaq Composite prints a fresh all-time high at 24,016.02 and the S&P 500 sets a new record at 7,022.95, both on narrow, tech-led breadth while the Dow slips red — a classic leadership-narrow melt-up. The score finished the session at 4.94 after the 4:19 PM post-close lift, back at the ref for the first time since this morning, which is not a reversal but is the first session in over a week where the divergence stopped widening into the close. Ref unchanged at 4.94, override zone holds from anything below 5.15, and the final recommendation doesn't move: 100% QQQ. Five sessions to April 21. The trend filter is still the trade. The pure signal is finally, barely, taking a breath.