Mag-7 / NVDA Peak Window 2026-2027: Per-Name Test on the Magnificent Seven

A per-name forward call on the Magnificent Seven cohort. The same calibrated stack as the AI bubble window (US rahu over stellium 3.0x and Markets saturn natal slow 2.75x) tested at name level rather than index level.

What this window typically looks like

Between 1 October 2026 and 30 June 2027, the seven names that together represent roughly twenty-eight percent of S&P 500 market capitalisation enter the structural-pressure phase of the AI-led concentration cycle. The per-name test sharpens the index-level AI bubble call by asking which of the seven names absorbs the pressure. A cohort-level pass at the index can hide dispersion. The per-name test makes the dispersion legible.

Per-name expected ranges inside window

Cohort behaviour expected through the window

• Dispersion widens. The cohort traded as one block through 2024 and 2025. Inside a structural-pressure window the dispersion across names typically widens as fundamentals separate. Earnings exposure to AI capex, balance sheet quality, and existing operating leverage all become individually relevant.
• The deepest drawdowns sit at the concentration centre. Historical cohort-leadership cycles show the single names most exposed to the dominant narrative (NVDA for AI infrastructure, TSLA for AI robotics) carry the deepest drawdowns. Cohort tail names (AAPL, GOOGL with broader revenue mix) carry the shallowest.
• Peak month clusters. Across the dot-com 2000 cohort the leadership names peaked within a sixty-day band. Cisco peaked late March 2000, Intel peaked late August 2000, Microsoft peaked December 1999, Oracle peaked September 2000. Peak clustering inside a stacked-signature window is a feature of the historical cohort-leadership pattern, not a guarantee.

Macro and earnings calendar inside the window

Section 1. The call

This article makes a dated, per-name testable prediction about the Magnificent Seven cohort. Tempora's reading of the US 1776 founding chart, run through the calibrated event corpus, says structural pressure on the seven concentration leaders rises sharply between 1 October 2026 and 30 June 2027. The window opens roughly eight weeks before Saturn's direct station on 25 November 2026 and closes thirty-eight days after Saturn ingresses sidereal Aries on 23 May 2027.

The discipline Tempora applies here is identical to the discipline in the AI bubble forward call (Q4 2026 to Q1 2027 index-level test), the Bitcoin halving cycle bottom call (1 July 2027 to 31 January 2028), the US recession window call (1 July 2027 to 30 June 2028), and the Iran 2027 tension window (1 April to 30 September 2027). We are not picking a target price for any single name. We are not saying NVDA prints a specific number on a specific date. What we are doing is testing whether the cohort-leadership concentration unwinds through the stacked-signature window in the way the historical record on this signature implies.

The body walks through three things. First, what the per-name test asks, in language that does not assume background in Vedic astronomy. Second, how the dot-com 2000 cohort-leadership cycle resolved at name level, and what the 2026 to 2027 cohort-leadership setup looks like by comparison. Third, the per-name thresholds that decide whether the call lands or fails. We commit, on the public record, to publishing the verdict in either direction.

Section 2. The mechanism, walked through

2.1 Same chart, same stacked signatures, sharper test

The mechanism on this Mag-7 article is identical to the AI bubble forward call. The chart is the US 1776 founding chart computed for 4 July 1776 in Philadelphia under the canonical Sibly birth time. The dasha schedule places the chart inside its Venus major period (mid-2010s into the 2030s) with a Rahu sub-period currently active. The transit data, taken from Swiss Ephemeris with True Pushya Paksha ayanamsa, shows Saturn at 19.65 degrees sidereal Pisces in mid-September 2026, Jupiter at 23.57 Cancer (exalted), and Rahu near 5.44 Aquarius opposite the US natal Cancer stellium (Sun, Venus, Jupiter, Mercury within roughly fifteen degrees).

The two stacked signatures both fire across Q4 2026 into Q1 2027. US rahu over stellium carries a calibrated lift of 3.0x across n=8 historical US events. Markets saturn natal slow carries a calibrated lift of 2.75x. The boundary is the same as on the AI bubble call: these are chart-level signatures whose calibrated lift is built against national and market event corpora respectively, not against personal charts. Per-name applicability follows from the cohort-leadership concentration pattern visible in the historical record (see Section 3), not from a separate per-stock calibration.

2.2 Why per-name on top of the index-level test

An index-level test passes when at least one of the test conditions fires anywhere in the index. The AI bubble Condition A (NASDAQ or S&P 500 single-session move of -3 percent or worse) and Condition B (Mag-7 average drawdown exceeding 20 percent) can both register MET while the per-name dispersion stays mixed. For example: if NVDA prints -45 percent and AAPL prints -3 percent and the other five sit between, the Mag-7 average might land at -20 percent and pass Condition B, while only one or two names individually meet the per-name threshold. The index test passes; the per-name test does not.

The per-name test is therefore stricter. It asks the harder question. Does the structural-pressure window actually compress the cohort across names, or is one or two names doing all the work while the rest carry the basket? The reconciliation answer published 14 July 2027 will be a per-name table with peak date, peak price, trough inside window, drawdown and pass-or-fail per name. The verdict aggregates: four or more passes equals MET, six or more equals stronger MET, three or fewer equals FAILED.

2.3 Why this window is wider than the AI bubble window

The AI bubble window opens 25 November 2026 (Saturn direct station) and closes 31 March 2027 (Rahu releases the stellium opposition by ingress to Capricorn around 1 January, but the call extends through Q1 to capture the post-station tail). The per-name Mag-7 window opens 1 October 2026 to capture peak-month detection before Saturn station and closes 30 June 2027 to capture the post-station drawdown completion. The wider window is appropriate because per-name peaks typically lead the index peak by weeks to months (the dot-com cohort showed peaks spread across a six-month window from December 1999 to August 2000), and per-name troughs often lag the index trough by weeks to months for the same reason.

Pre-window peak inclusion (2 August 2026 onward, within sixty days before window open) is by design. The structural-pressure signature begins to build during Saturn's approach to its direct station, not at the exact station date. Allowing the peak detection to reach back sixty days before window open captures the early-leading peaks the historical record shows.

Section 3. The dot-com 2000 cohort, name by name

The cleanest historical analog to the 2026 Mag-7 concentration setup is the dot-com 2000 leadership cohort. The same calibrated signature (US rahu over stellium) fired through 2000, with Rahu transiting Cancer directly over the US natal stellium and Saturn moving slowly through late Aries to early Taurus and aspecting US natal Saturn. The leadership names of that cycle were concentrated in tech and telecoms. The drawdowns from cycle peak to cycle trough, name by name, are below.

The cohort-leadership average drawdown for the dot-com 2000 cycle was -83 percent over the peak-to-trough span of roughly two years. The peaks clustered within a nine-month band from December 1999 to September 2000. The troughs clustered within a five-month band from May 2002 to October 2002. The shape of the cycle was extreme, but the cohort cohesion was high: the names rose together and fell together. The per-name test on that historical cycle would have resolved as a stronger MET (five of five at the leadership level, with thresholds well below historical outcomes).

The 2026 Mag-7 cohort is the structural analog. The seven names sit at roughly twenty-eight percent of S&P 500 market capitalisation as of late 2025, comparable to the dot-com 2000 concentration measured at roughly twenty-five percent across the comparable leadership basket. The capex concentration in five of the seven names exceeds the 1999-2000 dot-com peak by several measures (full table in the AI bubble forward call article). The calibrated signatures firing across Q4 2026 and Q1 2027 are the same.

What the per-name test is not claiming: that the 2026 to 2027 drawdowns will match dot-com 2000 magnitudes. The per-name thresholds (NVDA -25 percent, AAPL -15 percent, MSFT -15 percent, META -20 percent, GOOGL -15 percent, AMZN -20 percent, TSLA -30 percent) sit materially below the historical cohort-leadership average of -83 percent. The thresholds are set at the level where the structural-pressure signature should fire if it fires at all. A cycle that bottoms at -25 percent on NVDA is a structural correction without being a dot-com magnitude event.

Section 4. The test condition, what would invalidate this call

Every Tempora forward call carries a binary test condition with a stated threshold and a window close date. The call is wrong if three or fewer of the seven names meet the joint test by window close on 30 June 2027. The joint test per name is:

Condition per name. The name records its cycle all-time high price (intraday or close, whichever is higher) on a date between 2 August 2026 and 30 June 2027 inclusive, AND records a subsequent intra-window trough that produces a peak-to-trough drawdown of at least the per-name threshold (NVDA -25 percent, AAPL -15 percent, MSFT -15 percent, META -20 percent, GOOGL -15 percent, AMZN -20 percent, TSLA -30 percent) by window close on 30 June 2027.

Aggregation across the seven names: at least four passes resolves MET; six or seven passes resolves stronger MET; three or fewer passes resolves FAILED. There is no middle ground. If exactly four pass, the call is MET. If exactly three pass, the call is FAILED. The threshold cliff is at four. This binary cliff is the discipline; partial credit is not on offer.

This is the discipline that separates a calibrated per-name forward call from a general sector forecast. A sector forecast lets the forecaster claim partial credit for a dispersed outcome. A per-name call commits to a per-name table with a binary aggregation, and the forecaster takes the public verdict either way. Same discipline as the AI bubble call (Q4 2026 to Q1 2027), the Bitcoin halving cycle bottom call (Q3 2027 to Q1 2028), the US recession window (Q3 2027 to Q2 2028), the Iran 2027 window (April to September 2027), the Russia February 2028 sudden-action window, and the Gold post-exaltation reset call (rolling 3-month Oct-Dec 2027 average versus baseline).

Section 5. Reconciliation commitment

Window closes 30 June 2027. Reconciliation, in Tempora's discipline, means the article body is updated with the outcome no later than 14 July 2027, fourteen days after window close. The update lands in a Section 2 reconciliation block on this article, with a per-name table (peak date, peak price, trough inside window, drawdown, pass-or-fail per name), the aggregation result (count of passes out of seven, verdict MET or stronger MET or FAILED), and the chart-side recomputation showing the engine output for the window with full hindsight.

The same dual-loop reconciliation applies as on every Tempora forward call. The first loop is the outcome: did at least four of the seven names meet the joint test? The second loop is the mechanism: when we re-run the engine after window close, does the chart-side reading we published before the window match the chart-side reading the engine produces with hindsight? Both loops get a public verdict. The published Section 2 reconciliation states the directional outcome (passes out of seven) and any mechanism corrections required (for example if a calibrated signature input changed, or a transit date in the calendar was off by days).

The reconciliation also lands on the public tracker at tempora.ltd/tracker and stays there indefinitely. The brand position Tempora is staking out is not that the call lands. The brand position is that the call is testable and that the verdict is published either way. Misses stay on the tracker indefinitely as part of the audit trail. The audit holds the testability.

Frequently asked

Read next

• AI bubble Saturn-Rahu cycle. The index-level forward call this article extends. Same chart, same stacked signatures, index-level test condition.
• US recession window 2027-2028. The next-cycle extension of this call into broader-economy territory.
• Saturn return in Aries 2027-2030. Methodology bridge piece showing how this call sits among the five active forward calls inside Saturn-Aries.
• Bitcoin halving cycle bottom 2027 to 2028. Crypto-cohort companion forward call inside the same transit backdrop.
• Gold 2027 post-exaltation reset. The asset-cohort companion forward call running on a different planetary backbone.
• Methodology: how Tempora forward calls work. The calibrated-tier framing, falsifier discipline, and reconciliation commitment.

This article was prepared by Tempora Research as a per-name extension of the calibrated AI bubble forward call. It is a Tier 1 tracker article. The chart positions and transit dates are reproducible against the public engine using True Pushya Paksha (PVRN Rao) ayanamsa via Swiss Ephemeris. Calibrated weights and engine output are reproducible from calibrated_weights.json and market_calibrated_weights.json. Internal audit log maintained. This article does not constitute medical, financial, legal or professional advice. First published 2026-05-20 by Tempora Research.