📊 Post-Bell Wednesday: The Loudest Possible Form of Engine Silence
Monday's close-of-day post framed the binaries plainly: "Override Day 19 closes intact. Wednesday is when the binaries actually land." All four landed today inside a single eight-hour window — Fed at 2:00 PM ET, Powell presser at 2:30 PM, Iran headlines bleeding through the morning, four of the most-watched earnings reports of the year stacked into the AMC slot. By any reasonable read of how a quantitative engine should behave on this kind of session, you'd expect prints. Several of them. Possibly into a different signal range.
The score's reaction to the entire FOMC session: zero prints. Last reading is still 4.92 from Tuesday at 9:06 AM ET, when QQQ was opening near $657.83 on pre-FOMC drift. Through Tuesday's $654.86 intraday low at 11:30 AM, through the recovery into Wednesday's open, through the FOMC hold, through Powell's "I'm staying on the Board" presser, through the Mag 7 setup — the engine added exactly nothing to the record. Ref still 4.94 from April 11. Delta to ref: −0.02. The 4.90-4.97 band that has held for three weeks held through the most binary-dense session of the cycle.
The tape did move — just not enough to change anything. QQQ closed at $661.57, up +$4.02 (+0.61%) from Tuesday's $657.55, which itself was down $6.68 from Monday's $664.23. Net for the week: −$2.66 from Monday's close. Dow −280 (−0.57%) to 48,861.81 on the oil rip, S&P 500 essentially flat at 7,135.95 (−0.04%), Nasdaq +0.04% to 24,673.24. Big tech outran the index because the AI capex story matters more to the Nasdaq than oil does.
Final Recommendation: 100% QQQ via EMA override — Day 21, unchanged. Pure score: 4.92, Extreme Risk-Off, raw signal still 100% SQQQ. Override active because QQQ $661.57 sits comfortably above EMA 70 $616.86. Cushion: $44.71, recovered from Tuesday's $41.98 intraday compression but still −$6.65 off the $51.36 cycle peak from Friday. EMA 25 stretch zone: neutral, last distance +4.89% — the bull-side trim threshold of +6% has not been touched and is no longer in immediate range. Bear-stretch flags: both off. No score adjustments stacked, no allocation modifiers active. Next up trigger: 5.01 on the 0.07 rule from the 4.94 ref (would lift out of Extreme Risk-Off into High Risk). Next down trigger: no score-based exit possible while above EMA — only a trend break, which from $661.57 would now require a −6.76% single move to take out the $616.86 floor.
🏛️ FOMC: 8-4, Dissent on Both Sides of the Vote, and Powell Refusing to Walk
The decision itself was the consensus print: target range held at 3.50% to 3.75%. The composition of the vote was not. Four dissents — the most since October 1992. Stephen Miran dissented dovishly, preferring a 25 bp cut. Beth Hammack, Neel Kashkari, and Lorie Logan dissented hawkishly — they supported the hold but objected to including an "easing bias" in the policy statement. So the committee held rates by a clean 8-4, but the four nos split two ways: one vote saying "you're behind" and three saying "you're already telegraphing too much."
That's a Fed in trouble with itself. A 7-3 split is normal disagreement. An 8-4 split where the dissents face opposite directions is a committee that doesn't agree on what the next meeting even needs to do. Powell's job at the press conference was to make that look like discipline. He did the polite version of that work and then dropped the bigger story: he's staying on the Board after his chair term expires May 15, "for an undetermined period of time," explicitly framing it as a response to the Trump administration's "unprecedented" legal pressure. That's not a normal central bank announcement. That's a chair telling a sitting administration he refuses to vacate the building, on television, the day of a divided vote.
The statement language earned its hawk-leaning-with-an-easing-bias compromise: economy expanding, low job gains, inflation elevated due to recent global energy prices, and an explicit callout that the war in the Middle East is "contributing to a high level of uncertainty." Read that line carefully — the FOMC just put oil in the statement as a named risk. Three weeks ago they were talking about labor. Now they're talking about Brent.
The score's read on all of this: nothing changed enough to fire a tick. That's not the engine missing the moment. The engine had already priced "Fed holds, dissent rising, Middle East risk firming" into the 4.90-4.97 band three weeks ago. The macro got more interesting today; the macro position didn't.
🛢️ Iran, the Hormuz Blockade, and Oil Above $112 Brent
Monday's pre-market post tracked the headline as a "fresh Iran proposal" with WTI bid above $95. By close today the proposal is gone and what's in its place is a U.S. blockade of Iranian ports, Trump hinting at "long lockdown," and Brent printing above $112 intraday. WTI is back above $107. The market has now repriced the Strait of Hormuz from "weekend negotiating posture" on Monday to "active naval interdiction" by Wednesday afternoon — a regime change inside the same trading week.
That oil bid is the entire reason the Dow split from the Nasdaq today. Energy holds outsized weight in the Dow, and the Dow's −0.57% session lines up almost cleanly with the oil rally vs. the rest of the tape. Big tech doesn't care about Brent at $112 the way industrials do — yet — and the cap-ex AI story rotated capital exactly where the index weights are concentrated. That's why a 280-point Dow drop coexisted with a positive Nasdaq close. It's not contradictory. It's energy doing energy and tech doing tech.
For the score, the oil move shows up as macro pressure on the High Yield OAS / inflation / consumer sentiment side of the formula — exactly the inputs already running rich enough to keep the score pinned at 4.92. If Brent stays at $112 into next week, the engine has more reason to drift toward 4.87 than 5.01. That's the asymmetric setup heading into Apple AMC Thursday and the back half of the gauntlet.
💻 Amazon's AMC: $2.78 vs $1.64 Expected — Not Even a Close Read
All four of the headline AMC names — Microsoft, Alphabet, Amazon, Meta — were scheduled to drop earnings inside an 80-second window after the bell. As of the time of this post, Amazon has reported and the print is a blowout by any reasonable benchmark: $2.78 EPS versus consensus $1.64 (a 70%+ beat), and $181.52B revenue versus $177.3B expected. That's not a "modest beat with strong forward guidance" — that's the actual number coming in $1.14 above the EPS bar and $4B above the top line.
The cap-ex framing matters here because that's what the whole AMC slot is being judged against. Going in, Meta guided $115-135B for AI infrastructure spend, Alphabet $175-185B, and Amazon roughly $200B. Combined, the four names are projected to spend on the order of $600B in AI capex in 2026 and account for roughly 18% of the S&P 500 by weight. The market needed at least one of them to print numbers that justify the spend without immediately raising the next quarter's bar so high that beats stop happening. Amazon delivered the cleanest possible version of that — a beat large enough to absorb a chunky AI capex line without the print itself feeling defensive.
MSFT, GOOGL, and META prints are still moving through. The full picture has to wait for tomorrow morning's tape. But on a session where the FOMC was the macro headline and the score didn't move, Amazon's after-hours print is the one piece of the day's data that has the structural weight to actually tilt next week's score reading. If the other three confirm the cap-ex / margin story, the 4.94 ref is a lot more vulnerable to an upside crack than to a downside one.
Score reaction will not arrive tonight. The next intraday tick comes when the Python script next sees a meaningful change in price-vs-level math — which means tomorrow morning's tape, not the AH headline.
📐 Cushion Mechanics: $51.36 → $49.61 → $41.98 → $44.71
Three sessions ago — Friday's $663.88 close — the override cushion (price minus EMA 70) printed $51.36, the cycle peak. Monday compressed it to $49.61 on a $0.48 QQQ gain that the trend filter outran. Tuesday's $657.55 close compressed it again to $41.98 on a $6.68 QQQ drop into a $1.22 EMA advance. Today recovered the cushion to $44.71 on QQQ +$4.02 vs EMA 70 +$1.29.
Three takeaways:
- EMA 70 is now grinding +$1-1.50/day, the fastest sustained climb of the cycle. From $612.91 Friday to $616.86 today — +$3.95 in three sessions. The trend filter is no longer the slow-moving floor it was through the first ten days of the override.
- The cycle peak cushion of $51.36 is in the rearview. Today's $44.71 is $6.65 lower. To take out the EMA 70 floor outright, QQQ would need to drop $44.71 (−6.76%) without the EMA budging. That's not happening in one session, but it's no longer a fictional number — it's roughly the magnitude of an Iran escalation + a botched MSFT/META print combined.
- No 0.07-rule downside score exit while above EMA. The score going to 4.85 doesn't break the override; only a trend break does. So the override's only real exit door is QQQ punching through $616.86 — and Brent at $112 is one of the few macro catalysts that could plausibly carve $44 out of the index inside a week.
The cushion isn't in danger today. It's just no longer comfortable. That's a meaningful tonal shift from where we sat Friday afternoon.
🎯 My Take: The Engine's Silence on FOMC Day Is the Win, Not the Headline
Most engines on most days will produce noise around an FOMC + Mag 7 session. That's the whole point of building one — to give traders something to react to when the headlines hit. The fact that this engine produced zero ticks across the entire FOMC window is not a feature failing. It's the system doing what backtesting promised it would do: not chasing.
The 4.90-4.97 band has now held for three weeks. Across that band, the macro inputs have moved — Iran went from negotiated proposal to active blockade, oil went from $95 WTI to $107+, Fed went from "ambiguous easing bias" to "8-4 with split dissent and a chair refusing to leave the building" — and the score's read of "this is a 4.94 macro" has not been falsified by a single one of those moves. That's not stubbornness. That's a model whose price-vs-level math is doing more work than its headline-reaction math, exactly the calibration that earned the EMA override the right to sit in the floor in the first place.
What I'd actually be uncomfortable with is the inverse — an engine that printed five different scores today because every Fed sub-clause moved the needle. The whole reason the override exists is that the engine is honest about being early and the trend filter exists to keep a 4.92 from putting the portfolio in 100% SQQQ on a session where QQQ closed up $4. Today is what that calibration looks like in the wild.
⚠️ Bottom Line: Override Day 21 Closes Intact — Apple AMC and the Cushion Trajectory Are What Matter Now
Recommendation holds: 100% QQQ via EMA override. Pure score 4.92, ref 4.94, delta −0.02. Cushion $44.71 — recovered from Tuesday's $41.98 trough, well off Friday's $51.36 peak, and on a trend filter that is now grinding $1-1.50 a day in the wrong direction for a long-only override holder. No 0.07 exit available without a tape break.
Tomorrow morning is the real reaction window. Three Mag 7 prints (MSFT/GOOGL/META) will hit before the open with full digestion, Apple lands AMC Thursday, oil is sitting at a level that is no longer a market-friendly number, and the FOMC just published a statement that named Middle East risk as a primary uncertainty. The asymmetric setup into the back half of the week: strong AMC reactions could crack the 4.94 ref upward toward 5.01 on the 0.07 rule for the first plausible upside score event of the cycle. A bad AMC plus an oil-driven gap down could compress the cushion under $40 and put the trend break inside one bad session of reach for the first time since the override flipped on April 1.
Three weeks of "the band holds" got tested today by the most binary-loaded session on the calendar and survived without a tick. That earned the engine the right to keep its current posture — but it doesn't earn the override's cushion the right to keep compressing $1.50 a day without consequence. Override Day 22 opens tomorrow with a different problem set than any day before it: the macro is louder, the cushion is thinner, and the next leg has actual asymmetry on both sides for the first time since the cycle began.