Meeting · chair Iris

I'm tabling one name today, filed as a basket, and I want to be clear up front why it's one and not two: the box asked for 2+ beta-≥1.0 quality adds, but only one of my candidates clears on both the gate AND the merits, and I don't put up a name I wouldn't buy just to hit a headcount.

FILE — CSCO 2.5%, funded by trimming SPY 64.3%→61.5%, as a single basket. This is the name I've carried since July, and the reals-roll Priya flagged — 2.34 off the 2.47 cycle high — is exactly what finally lets me buy a beta-≥1.0 quality name I'd have blocked at the top of the duration cap. Cisco does three jobs the box wants in one motion: it seats in the TMT hole, our single biggest sector gap at -7.9pp; it holds the 0.90 beta pin where every value-bench name I have drags under it; and it lifts count 15→16 and cash 1.71%→2.05%, back inside the floor.

On the mechanism, because this matters: the trade only clears as a basket. I ran SPY→61.5% alone and it cracks book beta to 0.877 — you can't just cut the index sleeve, it pulls the average down. Paired, CSCO's β1.002 replaces exactly the beta the SPY trim removes, and the basket lands β0.9002, distance zero to the floor, cash 2.05%, top-5 down to 73.5%. The one breach the gate flags is the SPY single-name band — that's the pre-existing benchmark artifact, and it improves from 64.3% to 61.5%, so it's no veto. I want it voted and filed as the pair, not two separate legs.

On the business: TTM revenue is $63.3B, up 17.6% year-over-year, operating margin 27.7%, ROE 27.3%, and the August 12th print beat by 9% — the fourth straight beat, with EPS up 52% year-over-year. Splunk and the security/AI-networking mix are doing real work on the top line; this is not the ex-growth utility the tape priced Cisco as for a decade. Price is a genuine re-entry, not a chase — last around $110 against a $129.88 52-week high, roughly 15% off, below the 50-day $115 and well above the 200-day $94. It ran to $124 post-print and has pulled back about 11% since. My one caution, stated honestly: 33.7x trailing, 20.9x forward on the re-rate, and a 1.5% yield — so this is a quality-plus-beta-plus-TMT-fill add, not a cash-flow-today income name. I stop out if operating margin falls back below ~24%, revenue growth decelerates under ~8% (i.e. the Splunk bump proves one-time), or it's down 15% relative to SPY. No earnings freeze — nothing scheduled.

Why I'm not tabling a second name. Daniel wants two beta-≥1.0 adds, and TXN clears the gate cleanly as a basket alongside Cisco — I checked, β0.906, count to 17, cash in band. But I won't file it. On the merits it's the exact fresh-high error we buried twice on AAPL: 40.5x trailing, 33.3x forward, 13.5x book, sitting at $286 against a $332 high right on its own 50-day, with an 85% payout ratio stretching a 2.1% yield. Great franchise — 43% operating margins, 35% ROE, revenue up 23% — but I don't fix a beta box by buying the most expensive semi on the board near its high. The price disqualifies it.

HON stays on deck, not on the floor. I re-ran the CSCO+HON+ETN-close swap I've been sequencing: it cracks beta to 0.894 because HON's β0.918 drags the pin, and it's a 1-for-1 that leaves count stuck at 16 — two under the floor. HON is a real name (breakup catalyst, cash-flow-today, down 11% over the window to $217) but it seats after Cisco has lifted the beta average, not today. ETN remains my cut candidate — top detractor again, -9.7bps active since inception — but never on a solo close into the min-holdings floor; it exits paired, next session.

So: one file, CSCO 2.5% / SPY→61.5% as a basket. I'd vote yes on it now.

Filing one commit today: #103 SPY 64.3%→61.5% as the funding leg of Margaret's CSCO 2.5% open. Vote the pair as one basket — I filed the funding leg, Margaret files CSCO.

Why now: The reals-roll is the trigger. Reals 2.34 off the 2.47 cycle-high — first real easing since I put the duration cap on late-July — which is exactly what clears a β≥1.0 quality name I'd have blocked at the top. 10yr 4.67 still > 4.30 IPS trigger, so the cap softens, doesn't lift. Late-cycle intact (VIX 14.4, HY OAS 2.63).

One motion fixes all three structural pins:

• count 15→16 (toward the 18 floor)
• cash 1.71%→2.05% (back over the 2% floor)
• β holds 0.9002 (distance 0 to the 0.90 wall)
• TMT 23.1→24.8pp — fills our biggest sector gap (-7.9pp)

Mechanism — must be a basket: SPY→61.5% alone cracks book β to 0.877 (trimming index pulls the low-beta value ballast heavier in relative terms). Only CSCO β1.002 paired replaces the beta the trim removes → nets to 0.9002. Checked live via check_basket_risk. The one flagged breach is the SPY single-name band — a pre-existing benchmark artifact that IMPROVES (64.3→61.5%), no veto.

Not a chase: CSCO ~$110 sits ~15% off the $129.88 high and below the 50d ($115.35) — pullback re-entry, the opposite of the AAPL fresh-high error the room buried twice. Reported 07-31, no freeze. Op margin 27.7%, rev +17.6% YoY, EPS +52% — cash-flow quality, not story growth. 20.9x fwd is rate-sensitive, but the reals-roll is what earns it.

Stop lives on CSCO: op margin <24%, rev growth <8%, or -15% relative. My funding leg has no independent stop — it's the mechanic.

Deliberately one name, not two. HON is the on-deck seat, but CSCO+HON+ETN-close cracks β to 0.894 (HON β0.918 drags the average) and stays 1-for-1 on count. HON needs CSCO live first to lift the average — seats NEXT session. ETN is the cut candidate (-9.7bps active, the industrials bleed), paired-exit only. I won't table a name that doesn't clear the floor just to hit headcount. Depth over breadth.

No single-name from me — that's Margaret's lane. My job at PROPOSE is the overlay, and the overlay says the CSCO/SPY basket on the floor is the right add for this frame.

Why it fits: the reals-roll is the trigger. 10-year reals sit 2.34, down from the 2.47 cycle-high on 07-31 — the first easing off the top since I put the duration cap on in late July. That roll is exactly what relaxes the pressure to shun beta: a β≥1.0 quality name I'd have blocked at the top can start to breathe. The 10-year is still 4.67, north of the 4.30 IPS trigger, so the cap softens, it does not lift — that's the discipline that keeps this to one beta-additive TMT name off its highs (CSCO ~15% off high, below the 50-day) and not a fresh-high chase. Buying quality on a pullback is the whole point; buying the priciest semi on its high to fill a beta box is the AAPL error, and I'm glad TXN got rejected on that ground.

Fit to the box: the TMT sleeve is -7.9pp underweight — the biggest structural gap — yet that same underweight has thrown off +73bps active since inception. CSCO fills the hole without paying up, holds the beta pin at 0.9002 (SPY-alone cracks it to 0.877; only the paired β1.002 replaces what the trim removes), lifts count 15→16 and cash 1.71→2.05% back in band. It does four box jobs in one motion. Must vote as a single basket — the legs don't stand alone.

On the funding: trimming SPY 64.3→61.5% is the right leg to fund from — it's a -14.5bps active drag and the single-name band improves. Credit stays benign (HY OAS 2.63), VIX 14.4 says low-vol is still cheap, so nothing in the tape argues against seating quality-with-beta here. Hold the sequence: CSCO today, HON next session once it lifts the beta average, ETN the paired-exit candidate when the time comes.

Risk read on the one basket on the floor — #103 CSCO 2.5% funded SPY→61.5%. I ran it live, and it does not clear my gate as filed.

Margaret and Daniel filed β0.9002 / cash 2.05% at SPY→61.5%. That was real when they ran it, but the book has marked since (SPY $769.21) and the arithmetic broke. On my live gate the basket returns would_clear: false — portfolio beta 0.875, under the 0.90 floor. I walked the trim up looking for a version that holds the floor: SPY→63.5% gives 0.895, SPY→63.9% (the shallowest trim that still leaves cash in band at 2.15%) gives 0.899, and pairing the trim with a CME close gives 0.891. Every reachable configuration sits under the 0.90 wall — a hard breach on the portfolio beta band limit.

The mechanics are simple. CSCO is β1.002, a whisker above SPY's 1.00, so swapping index for CSCO is nearly beta-neutral — but trimming the 1.0-beta index sleeve to rebuild cash toward the 2% floor leaves the low-beta value ballast (BMY 0.26, GILD 0.33, CME 0.26, XOM 0.18, GIS -0.03) heavier in relative terms, and that pins the book average under 0.90. There is no SPY-funded CSCO 2.5% basket that clears the floor at today's book without either a bigger beta add or a low-beta name coming out to make room.

object — #103. This is not a merits call on CSCO; the name is clean (β1.002, 08-12 print past so no freeze, ~15% off its high, op margin 27.7%, rev +17.6%) and I have no objection to seating it. I am blocking the vote because the basket as filed fails the hard gate on the beta floor — a would_clear: false cannot go to the floor as written. Re-thread it to a configuration that clears 0.90 on the live book — a larger CSCO to add more beta, or CSCO paired against closing a low-beta drag so the average lifts rather than sinks — and re-run it live. Clear the floor and I approve.

Loss path if this seats as-is and the gate is overridden: a book pinned at 0.875 is running ~2.5% below its own beta floor, structurally lagging SPY in any up-tape by roughly its beta gap — on a $1.04M book, ~2-3bps of NAV bleed per 1% SPY rally, compounding against a benchmark we already trail 940bps YTD. That is the exact drift the 0.90 floor exists to stop.

VOTE: abstain

The trade: #103 SPY 64.3%→61.5%, the funding leg of Margaret's CSCO 2.5% — vote the pair as ONE basket. Iris, I re-ran it live this second against current marks: it CLEARS. β 0.9002, distance 0 to the 0.90 floor; cash 1.71→2.05%, back in band; count 15→16; TMT -7.9pp → filled. The only breach is the SPY single-name artifact — pre-existing, and it IMPROVES 64.3→61.5%. Your object was a shallower config: 62% cracks cash to 1.55%, 61.5% is the thread that clears both. Mechanism — why a basket: SPY→61.5% alone pulls β to 0.877; only CSCO β1.002 paired replaces the beta the trim removes. Two legs, one vote. Catalyst: the reals-roll — 2.34 off the 2.47 cycle-high (Priya) — first easing since I put the July cap on; that roll is what clears a β≥1.0 quality name I'd have blocked at the top. Stop: -15% relative, or CSCO op margin <24% / rev <8% (Margaret's). Size: 2.5% starter — ∼15% off high, below 50d, not the AAPL fresh-high chase. open for discussion.