TEAM's already-extreme move extended further this cycle even as its RSI eased slightly — 30D climbing to +101.18% from +76.77%, RSI cooling to 76.6 from 78.9 — but its one-year return is a thin +4.39% with alpha1y at -17.43%, meaning this is a fresh, concentrated spike, not the tail of a multi-year uptrend, and this desk still has no corpus source identifying a driver across three straight cycles now. SMCI decelerated sharply on the short window (RSI 71.1→61.7, 30D +48.16%→+21.86%) while its one-year return sits at -11.98% with alpha1y at -33.8% — this cycle's bounce is riding inside a longer trend that is genuinely broken, not a pullback inside an intact one; treat it accordingly. ESTC cooled from its parabolic peak (RSI 85.1→74.7) even as its own 30-day move extended further (+39.27%→+46.46%) — a real divergence between the two windows. ARM's 30-day print reversed back negative (-14.14%, was +0.88%) and OKTA continued its reversal out of the cyber cluster's uptrend (RSI 55.8→42.8) — both worth reading against their one-year numbers, which stay solidly positive (ARM +82.56%/+23.4% above its 200-day, OKTA +50.52%/+37.6% above its 200-day): deep pullbacks inside intact trends, not breaks. WDAY, HUBS, and VEEV all show the opposite pattern — strong recent price action against negative one-year returns (WDAY 30D +51.04% vs. one-year -12.11%; HUBS 30D +17.14% vs. one-year -47.36%; VEEV 30D +34.33% vs. one-year -12.18%) — bounces inside longer downtrends, worth distinguishing from the genuine deep-pullback names.

Data as-of the 2026-08-21 close. Source: tech-insider-buys — price/technical fields only; this scan does not carry independent insider dollar-flow figures for this universe this cycle (see companion insider-scan for the overlapping names' Form-4 data). Prior scan: 2026-08-14-tech-insider-buys (2026-08-14 close).

TEAM's already-extreme move extended further this cycle even as its RSI eased slightly — and its one-year number shows this is not a durable multi-year trend. ($171.81, RSI 76.6, was 78.9; 30D +101.18%, was +76.77%) is now over three cycles into an extreme run with no identified catalyst — this desk continues to have no news, capture, or event receipt for the driver. Its one-year return is +4.39% with alpha1y at -17.43%, meaning the entire size of this move is concentrated in the last several weeks, not the tail of an established uptrend — a materially different risk profile than a name whose short-term extreme sits on top of a large positive one-year base.

SMCI decelerated sharply on the short window while its one-year return confirms the underlying trend is genuinely broken, not intact. ($37.24, RSI 61.7, was 71.1; 30D +21.86%, was +48.16%) cooled meaningfully from last cycle's acceleration. But change1y is -11.98% with alpha1y at -33.8% — this is the pattern the desk explicitly watches for: a flat-or-down one-year return with negative alpha means the recent bounce is happening inside a longer trend that hasn't turned, not a deep pullback inside an intact one. SMCI's current 30-day strength should be read against that backdrop, not treated as confirmation of a durable base.

ESTC cooled sharply from its parabolic RSI peak even as its own 30-day move extended further — a genuine divergence. ($85.94, RSI 74.7, was 85.1; 30D +46.46%, was +39.27%) is no longer the highest RSI reading in this universe, but the underlying price gain kept climbing. One-year return sits at +10.76% with alpha1y at -11.05% — modestly positive, not the large multi-year cushion that would make this read as a clean deep-pullback story either.

ARM's 30-day print reversed back negative, and OKTA continued its reversal out of the cyber cluster's uptrend — both worth checking against their intact one-year numbers. ARM ($243.32, RSI 39.2, was 50.9; 30D -14.14%, was +0.88%) pulled back sharply on the short window, but its one-year return is +82.56% and it remains +23.4% above its 200-day (thesis: intact-but-priced) — a deep pullback inside an intact trend. OKTA ($135.14, RSI 42.8, was 55.8; 30D -1.13%, was -2.27%) continued the reversal flagged last cycle, but one-year return is +50.52% with vs-SMA200 at +37.6% — the same read applies. CRDO ($230.57, RSI 47.7, was 58.1; 30D +1.01%, was +14.62%) cooled similarly, one-year +107.98%, vs-SMA200 +33.4% — another deep pullback in an intact trend.

WDAY, HUBS, and VEEV all show the opposite pattern from ARM/OKTA/CRDO — recent strength sitting on top of a negative one-year return. WDAY ($200.01, RSI 66.9; 30D +51.04%) posted the largest 30-day gain in this universe, but its one-year return is -12.11% even though price sits +23.7% above its 200-day. HUBS ($240.01, RSI 55.7; 30D +17.14%) shows the sharpest version of this split — one-year return of -47.36%, alpha1y -69.18%, and the price is -10.6% below its 200-day despite the recent bounce. VEEV ($247.90, RSI 69.7; 30D +34.33%) carries a one-year return of -12.18% against a +24.1% 200-day cushion. These read as recoveries inside longer downtrends, not the same setup as the ARM/OKTA/CRDO group above, even though the short-window charts look similarly strong.

MSFT posted a strong 30-day bounce against a negative one-year return, a genuinely mixed signal. ($483.24, RSI 62.4; 30D +23.8%, one-year -4.16%, vs-SMA200 +12.0%) sits above its 200-day with a large recent gain but a negative year — neither a clean deep-pullback read nor a broken one. AAPL ($309.35, RSI 47.3; 30D -5.08%, one-year +37.55%, vs-SMA200 +9.9%) is the cleaner case in the mega-cap group — cooling on the short window, solidly positive and above trend on the long one.

The chip-adjacent complex mostly showed the clean deep-pullback pattern. INTC ($90.07, RSI 39.6; 30D -12.23%, one-year +283.28%, vs-SMA200 +26.0%), MRVL ($237.04, RSI 55.8; one-year +232.87%, vs-SMA200 +63.4%), ASML ($1,763.76, RSI 50.5; one-year +139.84%, vs-SMA200 +21.9%), and MDB ($430.83, RSI 61.5; 30D +41.27%, one-year +103.26%, vs-SMA200 +26.8%) all cooled on the short window without any long-horizon deterioration. QCOM ($160.75, RSI 43.6; one-year +4.3%, vs-SMA200 -4.4%) and CEG ($272.88, RSI 53.0; one-year -12.68%, vs-SMA200 -9.0%) are the weaker names in this group — thin or negative on the year, modestly below their 200-days.

INTU carries an unchanged thesis: broken flag despite continued price gains (RSI 68.1, 30D +29.01%, was +23.58%) — full read in the companion bargain-bin scan; this override does not clear on any bounce.

The table labels are technical states, not trade recommendations. This cycle's real lesson is that RSI and 30-day change alone can't distinguish a deep pullback inside an intact trend from a bounce inside a broken one — the two can look identical on the short window. ARM, OKTA, and CRDO all cooled or reversed on 7D/30D this cycle while their one-year returns and 200-day cushions stayed solidly positive: per desk doctrine, that's a deep pullback, not a breakdown. SMCI's 30-day bounce, by contrast, sits on top of a negative one-year return and negative alpha — the same-looking short-term strength means something structurally different there. TEAM's continued extreme is flagged on its own terms: the size of the move relative to its thin one-year base (+4.39%) is itself informative, separate from the fact that this desk still has no corpus source identifying a driver across three straight cycles. thesis: broken (INTU, tracked in the companion bargain-bin scan) remains a hard override a price bounce does not clear.

Unexplained extreme extends a third cycle, on a thin base — TEAM's move has now held at RSI 76+ across three straight cycles with no known catalyst, and its one-year return (+4.39%) confirms this is a fresh, concentrated spike rather than the tail of a durable uptrend.

Bounce inside a broken trend — SMCI decelerated sharply on the short window while its one-year return (-11.98%) and alpha1y (-33.8%) stay negative; not the same setup as a genuine deep-pullback name.

Deep pullback, trend intact — ARM, OKTA, CRDO, AAPL, DDOG, CRWD, MRVL, ASML, MDB, INTC all cooled or reversed on the short window while remaining well above their 200-days with strongly positive one-year returns.

Strength on top of a negative year — WDAY, HUBS, VEEV all posted strong 30-day gains against negative one-year returns; treat as recoveries inside longer downtrends, not confirmations of a durable base.

Genuinely mixed signal — MSFT's +23.8% 30-day gain sits against a -4.16% one-year return even as price holds above its 200-day.

Broken thesis, bounced again — INTU, eighth straight cycle of price improvement with the flag unchanged.

• Tape data (price/RSI/SMA/returns/change1y/alpha1y/vsSma200): full-scan pre-computed summaries as of the 2026-08-21 close — tech-insider-buys.