Options AI Screener Desk

Session 2026-10-08 · Thu rebuilt 11:05 ET

Two names today. The broad market reading held at 81, its best level in weeks, but every one of the five leading groups is slowing and only twelve of thirty-nine are rising. The strongest candidates clustered in energy again, and the record is emphatic that energy-only lists do badly, so one energy name is carried and one technology name alongside it. The record: fifty-seven tracked picks, 1.04% behind the index on average, with 25 of 57 beating it.

Today's call

Two names. One per sector, deliberately.

Constructive headline, narrow and slowing underneath

The broad market reading is 81 with volatility at 15.4, which on its own would allow a full list. Underneath it, all five strongest groups are decelerating and only 12 of 39 are rising. The screens agree the tape is thin: the signal screener's best composite today is 55.9, and nothing scores above 56.

Energy was the only rotation target with positive momentum, and it supplied four of the strongest candidates. That is exactly the cohort shape the record punishes. The two worst cohorts ever recorded here — 2026-09-02 at −12.8% and 2026-09-15 at −9.7% — were both all-energy. So one energy name is carried, not four, and a technology name fills the second slot.

Data freshness: all screening data is from this morning, and the news and analyst cross-check was available again today after three sessions without it.

Gates

Two fired. Leadership sets the cap; the energy record sets the shape.
1 · Regime81 · normal 10:26 ETVolatility 15.4, breadth reading 45%. Full list allowed, no mandatory confirmation. Note the screener's own sentiment gauge reads "fear" at 42 alongside this — the two disagree.
2 · LeadershipFIRED · cap 3 10:28 ETAll five highest-momentum groups decelerating: shipping +8.34/−25.99, cloud +6.45/−36.18, software +6.05/−27.68, energy +3.84/−13.37, commodities +3.60/−16.77. Cap 3.
3 · Intraday heatPassed 10:31 ETNothing on the shortlist is near +8%. MPC +3.8%, PAYS +0.2%. Two candidates were dropped for the opposite reason — see Dropped.
3b · EarningsPassed 10:34 ETMPC 26 days, PAYS 33 days. Nearest in the whole shortlist was VLO at 14.
3c · RSIMPC 74.7 · flagged 10:36 ETMPC 74.7, above the line. PAYS 67.5, below. A flag, not a veto. On MPC the oscillator block agrees with the RSI (stochastic 93.2, Williams %R −4.1, CCI 169.1, all Sell) while all twelve moving averages read Buy — short-term stretched inside an intact trend.
4 · BreadthFIRED · 12/39 10:28 ET31% of groups positive. The screener's own breadth reading is 45%, measured differently.
5 · Sector capDid not bind 10:55 ETTwo names, two sectors, cap 1 each. The energy concentration was resolved before the cap had to bind, by choosing one of four.

Where money is moving

Shipping and software lead on momentum; everything leading is slowing.

Shipping (+8.34), cloud (+6.45) and software (+6.05) top the momentum table, with energy fourth (+3.84) and commodities fifth (+3.60). Every one of them is decelerating, cloud most steeply at −36.2. Semiconductors are down 2.3% this morning and quantum 1.4%. Energy is the standout on the day — refiners are strongly bid, with Marathon +3.8%, Valero +3.2% and crude tankers +5.3%.

The only groups with positive acceleration are defensive or beaten-down: solar (+18.3 from −6.63 momentum), utilities (+12.2), electric vehicles (+11.7), building products (+11.0) and staples (+3.1). Staples is the single group that is rising on both measures, which is why the rotation engine scores Consumer Defensive at 95 and calls it BUILDING. None of its names cleared the relative-strength bar, so it is a rotation to watch rather than to buy.

Finalists

Two names, one per sector. Both clear the relative-strength rule.
PAYS$14.99 · Paysign · Payments technology

The highest relative strength outside energy, by a wide margin. Relative strength 95.5 against a next-best non-energy reading of 51.2, stage 2, perfect 10/10 stage scores on both the weekly and daily, and a Strong Buy analyst consensus. Momentum reading 7 of 10. Caveats: this is the only finalist on a single screen — it is not a rotation gem and did not make the signal screener's top band, so there is no second-source confirmation. It has run 178% since entering Stage 2 twenty-nine weeks ago, and its group (software) is the third-fastest decelerating on the board at −27.7. At $811M it is the smaller of the two names.

Rel str95.49
Stage2
SATA10 / 10
RSI67.5
Intraday+0.2
Earnings33d
Strong BuyRSI clearSingle screenGroup slowing
MPC$459.26 · Marathon Petroleum · Refining

The most corroborated name on the board, and the one energy slot. Relative strength 63.8, stage 2, 10/9 stage scores, and it appears on three screens — Stage 2, the rotation gems and the signal screener, where it rates buy at composite 52.0. Smart score 9 of 10, analyst consensus Buy, news sentiment 0.8, and the strongest institutional reading on the board at 0.86. A real catalyst today: Mizuho raised its price target to $457 from $304, keeping a Neutral rating, citing higher and longer-lasting refining margins — which is a margin thesis rather than a one-day spike. Caveats: RSI 74.7 is above the flag line; it has run 125% since entering Stage 2 thirty-six weeks ago; and it trades right at that raised target. Energy is the sector this record has punished most.

Rel str63.82
Stage2
SATA10 / 9
RSI74.7
Intraday+3.8
Earnings26d
Three screensCatalystSmart score 9RSI 74.7Energy

Options structures

No match for either finalist. That is the normal result.

Both income screens were rebuilt by this run: 725 cash-secured puts across 135 names and 146 bull put spreads (65 unique). Neither finalist has a matching structure. That is expected rather than a failure: the income screens want high-volatility large caps, and the Stage 2 funnel produces mid-cap momentum names — two systems sharing a dashboard. Today's spread screen is led by NBIS, MSTR, ORCL and CRWV; the put screen by MU, NVDA, IBIT, TSLA and SPCX. No LEAP call surfaced, which is as it should be given that strategy's record.

Dropped, and why

Two collapsed intraday, two cut by the energy rule, one whipsawing, one corrected.

Correction: ECO was cut on a sector claim that is wrong

This page first said ECO is really an energy name and that counting it as Industrials would have let a fourth energy name past the sector cap. That was asserted, not measured, and it is wrong. Daily-return correlation over 91 sessions puts ECO at 0.09 against energy, −0.09 against industrials and 0.60 against shipping — and at 0.10 against Marathon, the refiner it was supposedly doubling up with.

It is not a one-off. All seven tankers in this record cluster the same way: 0.66 against shipping, 0.15 against energy, 0.14 against the refiners, while the refiners sit at 0.55 against energy and 0.09 against shipping. Shipping is its own exposure. A tanker beside a refiner is not the concentration the cap exists to stop; two tankers filed under different sector labels is.

GKOS · VNCE collapsing today GKOS is down 13.9% on the session (opened 170.67, trading 148.32) and VNCE down 11.8%. VNCE carried the highest relative strength in the entire pool at 166.3 and topped the momentum list at 67.4. Both read as clean Stage-2 names in this morning's upload, which cannot see an intraday collapse. A breakout name down 12% on the day is not breaking out.
VLO · HPK energy, one slot Both qualify: VLO RS 60.3 on three screens, HPK RS 22.1 on two. With MPC taken, the sector is full.
ECO late stage — see correction RS 61.7, rated buy, up 5.3% today. It was cut this morning for the wrong reason — see the correction above. On the stated grounds it stays out: stage 2B, 61 weeks since breakout and 273% since, the most extended profile of any candidate. That is a weaker reason than the one given, and it is recorded as such.
CNC whipsaw The best technical summary on the board (the only Strong Buy, RSI 53.2 with room) and a fresh crossover. But the crossover history is seven alternating signals in six weeks, today's is a SELL at 63.89, and the price sits below yesterday's BUY at 64.99. It fails all three crossover noise filters at once.
FRD · SKK · TLIH · VAI too small or too thin FRD clears on relative strength at 51.2 but traded 1,214 shares this morning. SKK ($17.6M), TLIH ($16.2M) and VAI ($30M) are below any workable size floor.
RELL · BSET · HSTM · OGN below the bar RELL (RS 30.9) is stage 2B, 53 weeks out and down 2% today. BSET, HSTM and OGN sit at 20.2–20.3, barely over the line, each on one screen. OGN's 10-week range is 1.5% of price — too quiet to travel. None is strong enough to take a third slot on a day the leadership gate already capped the list.

Fifty-seven picks, nineteen flagging sessions

Alpha versus SPY from each flag-date close, re-priced 11:00 ET.
CohortMean alphaMedianBeat SPYNote
2026-08-31 · 7+4.56%+2.16%4/7CVI +40.1 carries it
2026-09-01 · 7−1.53%+2.94%4/7mean and median disagree
2026-09-02 · 3−12.83%−11.55%0/3all energy
2026-09-03 · 3−9.14%−8.44%0/3CRGY, ZYME, PFE
2026-09-08 · 3−2.96%−7.40%1/3ELPC the only winner
2026-09-11 · 3−1.78%−3.07%1/3all energy
2026-09-15 · 3−9.65%−9.63%0/3all energy
2026-09-16 · 2+2.50%+2.50%1/2MPC flagged here, +12.0%
2026-09-17 · 3+10.58%+6.52%3/3best cohort
2026-09-18 · 2−2.50%−2.50%0/2ABCL, CMBT
2026-09-21 · 3+8.06%+7.91%3/3RBRK, ALAB, NBIS
2026-09-23 · 3−3.32%−2.51%1/3RNG, MRNA, OMER
2026-09-25 · 2−0.64%−0.64%1/2SN, LCUT
2026-09-28 · 3−7.09%−4.87%0/3DGII, HUM, RVMD
2026-09-30 · 1−1.96%−1.96%0/1AMN
2026-10-02 · 2−0.38%−0.38%1/2SMTC, NAT
2026-10-05 · 3+3.12%+2.99%2/3IMOS, NTAP, DAC
2026-10-06 · 3−0.58%+1.05%2/3OKTA, FTNT, CRL
2026-10-07 · 1+0.30%+0.30%1/1AMD, one session old
All 57−1.04%−1.87%25/57trimmed −1.34% · −59.2 pts
2026-10-08 · 2———flagged today · first reading next session

The two worst cohorts on record are both all-energy

2026-09-02 at −12.8% and 2026-09-15 at −9.7%, neither with a single name beating the index. Both were three energy names flagged on one day. That is the single clearest pattern in nineteen sessions, and it is the reason today carries one energy name out of four that qualified.

The tails remain large and cut both ways. The best two names (CVI +40.1, TRMD +21.8) and the worst two (GEN −28.6, PRGS −21.4) are each worth about 60 points. Remove the best two and the book is 2.20% behind; remove the worst two and it is 0.17% behind. The trimmed mean of −1.34% is the honest figure, and the book is still behind its benchmark.