Article published Jul 1, 2026. Prices below use latest available snapshots.
Question: NVDL MSTU spread structural tell wrapper-collapse terminal Verdict: resolved: structural step-function tell not tactical mean-revert
What this resolves
The 2026-05-26 weird-ETF consolidation filed a ~20-min followup: pull the 1y daily NVDL/MSTU ratio and read the chart shape — mean-revert (→ tactical signal) or trend/step-function (→ the structural-regime tell the investigation described). The literal chart-render needs a daily OHLC series for both legs, which the spot summaries don't carry. But the fresh 2026-07-01 spot windows plus the 5/26 baseline already answer the question decisively, and no daily series would change the call.
Price truth: all levels/returns from the ETF scan summaries, as-of 2026-07-01 close.
The spread did not mean-revert — it step-functioned
| Leg | Price | 30d | 3m | 1y | vs SMA200 | RSI | Regime | From 52wk high |
|---|---|---|---|---|---|---|---|---|
| NVDL (2x NVDA) | $29.04 | -24.49% | +18.14% | +25.25% | -3.6% | 42 | downtrend | -32.88% |
| MSTU (2x MSTR) | $2.00 | -63.03% | -51.22% | -97.66% | -85.9% | 34.8 | collapse | -98.15% |
| MSTX (2x MSTR, sib) | $9.13 | -63.10% | -51.56% | -97.69% | -86.0% | 34.9 | collapse | -98.16% |
5/26 baseline for reference: NVDL ~+115% 1y / MSTU ~-92% 1y, spread flagged "wide."
Verdict on shape: structural step-function collapse, not mean reversion. MSTU went from -92% to a terminal $2 (-98% from its 52wk high, regime "collapse"), and its MSTX sibling is identical (-97.69%). A 2x daily-reset wrapper at $2 down 98% is functionally dead — there is no mean-revert path back to the NVDL leg. This is exactly the "daily-reset leverage works on trends, dies on chop → NAV bleeds to zero" mechanism the investigation described, now run to its terminal conclusion. The "spread becomes a tactical mean-revert signal" branch is dead. MSTU's headline +23.46% one-day pop is the residual bounce-on-any-directional-MSTR-move the 5/26 note predicted — pure leverage mechanics noise, not recovery.
Instrument implication: MSTU is now uninformative on the downside (it can't collapse much further from $2/-98%). Keep it only as a dead-wrapper marker; the informative leg has moved to NVDL.
The standing-watch trigger is now ambiguous — split it into two legs
The 5/26 trigger read: "When the spread narrows (MSTU recovers / NVDL stalls) → regime-shift signal." That conflates two mechanisms that now point opposite directions:
- Narrowing via MSTU recovery (the 5/26 mental model) — would mean crypto-leverage money returning. Not happening; MSTU is floored.
- Narrowing via NVDL cooling (what is actually happening now) — NVDL is 30d -24.49%, regime flipped to
downtrend, 3m still +18.14% and structurally intact (1y +25.25%, 3y +609.76%, only -3.6% vs SMA200). NVDY (the single-share NVDA-income sibling) echoes it: +25.10% 1y but regime "pullback." This is the AI leg de-levering, not the crypto leg reviving.
These are opposite tells. Narrowing-via-NVDL-weakness is an AI-side crowded-trade unwind read (ties to ai-capex-digestion), whereas the 5/26 framing treated any narrowing as a crypto/regime-rotation signal. A single NVDL/MSTU ratio can't tell them apart. Recommendation: the standing watch (and the deferred weird-etf-watch instrumentation followup) should track two legs separately — AI-leg NVDL trend health vs crypto-leg MSTU floor — not one ratio.
Trend-hold lens
Neither NVDL nor MSTU is a hold candidate — both are decay wrappers, useful only as regime tells. NVDA the underlying remains the structural name; NVDL's -24.5% 30d is a leveraged echo of NVDA's own pullback, not a break (vs SMA200 only -3.6%). No action on the wrappers.
Feasibility note (why no chart)
A rendered 1y daily ratio chart with exact step-break dates needs daily OHLC for NVDL + MSTU, which the spot summaries don't provide, and which a read-only producer can't fetch or compute. But per the feasibility gate: the chart no longer changes the call — the tactical-vs-structural question is settled by the spot windows above. This followup is retired; only the two-leg trigger refinement carries forward into the instrumentation decision.
Sources
- Price truth: etf-monster-universe, as-of 2026-07-01 close.