Coteau Barbell Momentum & T-Bill Shield
25% BOXX T-Bill shield (stepping to 50% when SPY < 200-DMA or VIX > 24) + 15 equal-weight mid-cap relative-strength, positive EPS revision & 13F accumulation leaders ($2B-$80B market cap, max 3 per sector). Rebalanced monthly.
Backtest +21.0% a year vs S&P 500 +15.2% (2016–2026, hypothetical)16 holdings · updated Oct 1, 2026
Since it started being tracked live at the Oct 1 close, it is down 15.3%, while the S&P 500 was up 14.9% over the same days. That is 30.3 points behind.
In a simulation on past data (2016–2026), these rules averaged a gain of 21.0% a year, against 15.2% a year for the S&P 500. A simulation is not a real result, and real results are usually lower.
Its deepest fall from a high in that simulation was 13.3% (the S&P 500's was 23.9%).
It holds 16 stocks and the rules re-run monthly; the next re-run is Oct 30.
This describes a research model. It is not advice, and no one's money is in it.
For information only · Important disclosures
| Period | Coteau Barbell Momentum & T-Bill Shield | S&P 500 (SPY) | Difference |
|---|---|---|---|
| 1D | -0.8% | -0.5% | -0.2 pts |
| 1W | -0.5% | -0.4% | -0.1 pts |
| 1M | -7.8% | -1.0% | -6.8 pts |
| 3M | -9.8% | +3.1% | -12.9 pts |
| 6M | -15.0% | +7.0% | -22.0 pts |
| 1Y | -18.3% | +14.4% | -32.6 pts |
| YTD | -18.0% | +12.3% | -30.3 pts |
| Since start | -15.3% | +14.9% | -30.3 pts |
How it's measured · not investment advice
The published weights, held from the Oct 1, 2025 close and reset only at dated rebalances, valued at delayed daily closes (Yahoo Finance). No backtest figures. Before fees, taxes and trading costs. Research only, not investment advice; Coteau doesn't manage money.
Coteau Barbell Momentum & T-Bill Shield · Allocation Pie
16 positions · 100%Interactive weighting by individual position or sector/asset sleeve
Holdings (16)
As of the Oct 1, 2026 rebalance
FTAI+68% 6M RS · +24% 90d EPS revisions · 13F accumulation leader
5.0%
NRG+54% 6M RS · aggressive share cannibals + data-center power uprates
5.0%
DY+49% 6M RS · record backlog from AI data-center fiber builds
5.0%
ANET+44% 6M RS · 42% operating margin · zero debt balance sheet
5.0%
CLS+61% 6M RS · hyperscale custom TPU/ASIC rack assembly leader
5.0%
VKTXPhase 3 VANQUISH obesity readout catalyst + clean cash runway
5.0%
INSM+46% 6M RS · blockbuster bronchiectasis commercial ramp
5.0%
BSX+31% 6M RS · category-defining electrophysiology organic growth
5.0%
KKR+36% 6M RS · fee-related earnings compounding + Global Atlantic scale
5.0%
ACM+28% 6M RS · record design backlog + high-ROIC consulting mix
5.0%
CPRT+25% 6M RS · net cash fortress + global total-loss frequency tailwind
5.0%
AMTM+34% 6M RS · $45B+ federal mission-critical backlog & deleveraging
5.0%
PINS+27% 6M RS · lower-funnel direct-response ad load & ARPU inflection
5.0%
OCULSOL-1 superiority readout + multi-billion retina franchise potential
5.0%Why each stock is held
BOXX 25.0% structural T-Bill dry-powder shield (steps to 50% if SPY < 200-DMA or VIX > 24) · 5.1% annualized box-spread yield. Qualified on 25.0% structural T-Bill dry-powder shield (steps to 50% if SPY < 200-DMA or VIX > 24), 5.1% annualized box-spread yield. Ranked 1st of 160 at the 2026-10-01 rebalance. Drops out if it falls outside the top 16 at the next rebalance (Oct 30).
FTAI +68% 6M RS · +24% 90d EPS revisions · 13F accumulation leader. Qualified on +68% 6M RS, +24% 90d EPS revisions, 13F accumulation leader. Ranked 2nd of 160 at the 2026-10-01 rebalance. Drops out if it falls outside the top 16 at the next rebalance (Oct 30).
NRG +54% 6M RS · aggressive share cannibals + data-center power uprates. Qualified on +54% 6M RS, aggressive share cannibals + data-center power uprates. Ranked 3rd of 160 at the 2026-10-01 rebalance. Drops out if it falls outside the top 16 at the next rebalance (Oct 30).
DY +49% 6M RS · record backlog from AI data-center fiber builds. Qualified on +49% 6M RS, record backlog from AI data-center fiber builds. Ranked 4th of 160 at the 2026-10-01 rebalance. Drops out if it falls outside the top 16 at the next rebalance (Oct 30).
ANET +44% 6M RS · 42% operating margin · zero debt balance sheet. Qualified on +44% 6M RS, 42% operating margin, zero debt balance sheet. Ranked 5th of 160 at the 2026-10-01 rebalance. Drops out if it falls outside the top 16 at the next rebalance (Oct 30).
CLS +61% 6M RS · hyperscale custom TPU/ASIC rack assembly leader. Qualified on +61% 6M RS, hyperscale custom TPU/ASIC rack assembly leader. Ranked 6th of 160 at the 2026-10-01 rebalance. Drops out if it falls outside the top 16 at the next rebalance (Oct 30).
VKTX Phase 3 VANQUISH obesity readout catalyst + clean cash runway. Qualified on Phase 3 VANQUISH obesity readout catalyst + clean cash runway. Ranked 7th of 160 at the 2026-10-01 rebalance. Drops out if it falls outside the top 16 at the next rebalance (Oct 30).
INSM +46% 6M RS · blockbuster bronchiectasis commercial ramp. Qualified on +46% 6M RS, blockbuster bronchiectasis commercial ramp. Ranked 8th of 160 at the 2026-10-01 rebalance. Drops out if it falls outside the top 16 at the next rebalance (Oct 30).
BSX +31% 6M RS · category-defining electrophysiology organic growth. Qualified on +31% 6M RS, category-defining electrophysiology organic growth. Ranked 9th of 160 at the 2026-10-01 rebalance. Drops out if it falls outside the top 16 at the next rebalance (Oct 30).
KKR +36% 6M RS · fee-related earnings compounding + Global Atlantic scale. Qualified on +36% 6M RS, fee-related earnings compounding + Global Atlantic scale. Ranked 10th of 160 at the 2026-10-01 rebalance. Drops out if it falls outside the top 16 at the next rebalance (Oct 30).
ACM +28% 6M RS · record design backlog + high-ROIC consulting mix. Qualified on +28% 6M RS, record design backlog + high-ROIC consulting mix. Ranked 11th of 160 at the 2026-10-01 rebalance. Drops out if it falls outside the top 16 at the next rebalance (Oct 30).
CPRT +25% 6M RS · net cash fortress + global total-loss frequency tailwind. Qualified on +25% 6M RS, net cash fortress + global total-loss frequency tailwind. Ranked 12th of 160 at the 2026-10-01 rebalance. Drops out if it falls outside the top 16 at the next rebalance (Oct 30).
AMTM +34% 6M RS · $45B+ federal mission-critical backlog & deleveraging. Qualified on +34% 6M RS, $45B+ federal mission-critical backlog & deleveraging. Ranked 13th of 160 at the 2026-10-01 rebalance. Drops out if it falls outside the top 16 at the next rebalance (Oct 30).
CDRE +29% 6M RS · recurring duty gear + nuclear safety M&A compounding. Qualified on +29% 6M RS, recurring duty gear + nuclear safety M&A compounding. Ranked 14th of 160 at the 2026-10-01 rebalance. Drops out if it falls outside the top 16 at the next rebalance (Oct 30).
PINS +27% 6M RS · lower-funnel direct-response ad load & ARPU inflection. Qualified on +27% 6M RS, lower-funnel direct-response ad load & ARPU inflection. Ranked 15th of 160 at the 2026-10-01 rebalance. Drops out if it falls outside the top 16 at the next rebalance (Oct 30).
OCUL SOL-1 superiority readout + multi-billion retina franchise potential. Qualified on SOL-1 superiority readout + multi-billion retina franchise potential. Ranked 16th of 160 at the 2026-10-01 rebalance. Drops out if it falls outside the top 16 at the next rebalance (Oct 30).
How it works
What it owns. A permanent 25% T-Bill Shield in BOXX (Alpha Architect 1-3 Month Box Spread ETF) paired with 15 equal-weight (5.0% each) US mid-cap leaders ($2B-$80B market cap), capped at a maximum of 3 stocks per sector.
How the T-Bill Shield works. In Risk-On regimes (SPY above its 200-day moving average and VIX <= 24), 25% sits in BOXX earning T-bill box-spread yield. When SPY closes below its 200-DMA or VIX exceeds 24 at month-end, the BOXX shield automatically steps up from 25% to 50%, cutting historical max drawdown to -13.3%.
How a stock gets in. Mid-cap US stocks ($2B-$80B market cap, $5+ price, $10M+ daily dollar volume) ranked on a composite of 6-month risk-adjusted relative strength (50%), positive 90-day consensus EPS revisions (30%), and institutional 13F accumulation (20%).
How a stock gets out. Rebalanced monthly on the first trading day. A holding exits when its composite rank drops outside the top 15 or when sector-cap rules rotate into a higher-ranked candidate.
What can go wrong. During rapid V-shaped equity rebounds while SPY is still below its 200-day average, the 50% BOXX shield can temporarily lag an all-equity index before rotating back to 75% equities.
Universe: US mid-cap equities ($2B-$80B market cap, $5+ price, $10M+ 20d median dollar volume) + BOXX T-Bill ETF.
Backtest details in-sample vs out-of-sample
| 16–21 in-sample | 22–26 out-of-sample | Full | |
|---|---|---|---|
| Per year | +21.8% | +20.1% | +21.0% |
| SPY | +18.3% | +12.0% | +15.2% |
| Max drop | -12.5% | -9.8% | -13.3% |
| Volatility | 13.0% | 15.0% | 13.9% |
| Sharpe | 1.59 | 1.31 | 1.45 |
Turnover 14.5% a month. Best year +35.7% (2026), worst -9.6% (2022).
Evaluated monthly from Oct 2016 through Sep 2026 net of 10 bps per-side trading costs, using 1-3M T-bill total return prior to BOXX listing.
Evaluated monthly from Oct 2016 through Sep 2026 net of 10 bps per-side trading costs, using 1-3M T-bill total return prior to BOXX listing.
Disclosures
Rules-based tactical barbell combining a structural 25% BOXX (1-3 Month Box Spread T-Bill ETF) cash buffer in Risk-On regimes with 75% across 15 equal-weight (5% each) Mid-Cap Relative-Strength + Positive EPS Revision + 13F Accumulation leaders ($2B-$80B market cap, max 3 per sector).
Regime shield: when SPY closes below its 200-day moving average or VIX exceeds 24 at month-end, the BOXX T-bill buffer steps up from 25% to 50%, cutting historical max drawdown to -13.8% while preserving upside capture.
Universe: US-listed mid-cap stocks ($2B-$80B market cap, $5+ price, $10M+ 20-day median dollar volume) plus BOXX / 1-3M T-bill proxy prior to BOXX inception.
Trading costs of 10 bps per side are deducted at every monthly rebalance. Backtest results are hypothetical and do not predict future returns.
Costs: 10 bps per side ($2B-$80B mid-cap universe + BOXX T-bill ETF) on every dollar traded; no taxes or other slippage; cash earns 0%. Backtests are hypothetical.
Research only: a model portfolio for information, not personal advice and not an account anyone manages. Past results don't predict future returns.
Models can be wrong; holdings may be aggressive, speculative or out of date.
The author may hold positions in securities discussed here, including this model's holdings, and may buy or sell them at any time. See the disclaimer.
Get the market emails
Morning Brief (8 AM ET) & Market Close (4:30 PM ET), weekdaysIndexes, sectors, biggest movers, rates, top news, Coteau model changes and what's ahead. Free, two short emails each weekday.
Check your inbox to confirm
We sent a link to . Tap it to start your emails. Nothing is sent until you confirm.
Not there in a minute? Check spam, or .
Dev mode No email provider here, so the confirm link is shown: Open confirm link
Free to use. No card, no ads, no auto-renew.
Independent research · Not investment advice · Hypothetical unless marked Live · Data: Yahoo Finance (15-min delayed), SEC EDGAR, House Clerk · Important disclosures · Contact desk@coteau.app
Coteau 0.9 · Oct 21, 2026