QEC Studio · Portfolio Operations · Confidential
Lodestar Capital · 北极星资本

Track C — Core

The compounding engine. Broad ownership, held for decades, tilted toward value — and bought hardest when everyone else is selling.

Operating Manual · Rev. LC-2607-C
The Mandate

55% of capital, compounding quietly

Track C is the largest and slowest part of the system — the compounding core that the catalyst sleeves (B, B-2) and the income sleeves (A, A-2) orbit. Its job is not to be clever. Its job is to own the market broadly, cheaply, and tax-efficiently, and to keep owning it through every drawdown the other tracks react to. The edge here is time and temperament, not timing.

Where Track B asks "what reprices this, and by when," Track C asks nothing of the day at all. It buys broad index exposure on a schedule, lets dividends and earnings compound, and is deliberately the most boring track you run. Two refinements make it more than a plain index: a systematic value tilt (Track C-2), and a VIX-triggered capitulation override that accelerates buying precisely when fear is greatest.

The Core

Broad, cheap, tax-placed

The anchor is broad-market beta — total US market (VTI) plus ex-US for global ownership. The discipline is in the boring parts: low cost, low turnover, and putting each holding where it's taxed least.

PrincipleRuleWhy
Broad ownershipVTI + ex-USOwn the whole market; don't bet on single names in the core.
Low turnoverbuy & holdTurnover is tax and cost leakage. The core barely trades.
HSA as stealth Roth→ Track CTriple-tax-advantaged; invest it in the core for max compounding.
VXUS locationtaxableForeign tax credit (FTC) is wasted in sheltered accounts.
Tax-loss harvesttaxable coreHarvest drawdowns in taxable; mind wash-sale windows.
Contribution cadencescheduledDollar-cost in; let new money do most of the rebalancing.
The Override

Buy hardest when fear peaks

The one moment Track C abandons its calm is a volatility spike. A VIX-triggered capitulation override accelerates core buying in staged steps as fear deepens — turning the market's panic into your accumulation. This is the core's only timing rule, and it points the opposite way to instinct.

VIX 45
First acceleration — deploy the initial reserve tranche into the broad core.
VIX 60
Deepen — second tranche. Disorder is now extreme; valuations are repricing down hard.
VIX 75
Maximum — final tranche. Historic fear; the core buys with both hands.
Note the inversion across tracks. At VIX 28–45 Track B sits out — catalyst trades fail in disorder. At those same levels and beyond, Track C accelerates in. That's not a contradiction: B is renting short-term repricing that panic destroys, while C is buying long-term ownership that panic discounts. The same volatility is a threat to one sleeve and a gift to the other. Keep the reserve for this override pre-designated so you're never selling one track to fund it.
Track C-2

The value tilt — factor, not selection

Value investing's first honest home in this system: a systematic value-and-quality tilt inside the core. No stock-picking, no judgment, no daily scan — a target allocation that leans toward value, rebalanced by drift bands quarterly. It harvests the documented value premium the way the rest of your machine works: rules held through discomfort, not discretion.

broad core
55%
US large value
20%
US quality
10%
intl value
15%

The tilt is a partial lean over broad-market beta, not a bet against it — you keep the anchor and add value and quality factor exposure alongside. Quality is deliberate: value-trap insurance at the index level, screening out the cheap-because-dying names a naive value screen would buy. Rebalancing is by bands — drift must breach both a relative threshold (>5%) and an absolute one (>3pp) — not the calendar alone, to limit turnover and tax drag, which matters because the core lives largely in taxable. International value is flagged for taxable specifically, for the same FTC reason as the broad ex-US holding.

The only hard part is doing nothing. A value tilt underperforms broad beta for long stretches — sometimes years. That underperformance isn't the strategy failing; it is the price of the premium, and the investors who earn it are the ones who hold the tilt when it's working against them. check_rebalance() makes the decision mechanical: it tells you when drift genuinely warrants a trade and otherwise says hold. No conviction input, no entry signal, no exit — because a tilt you second-guess is just active management with extra steps.
trackc2_tilt.py · drift-band rebalance check
def check_rebalance(holdings, targets, rel_band=0.05, abs_band=0.03):
    total = sum(h.market_value for h in holdings)
    for sleeve, tgt in targets.items():
        cur = by_sleeve.get(sleeve, 0) / total
        drift_abs = cur - tgt                 # percentage points off
        drift_rel = drift_abs / tgt           # relative to target

        # Trade ONLY if BOTH bands are breached — avoids churning
        # small sleeves on tiny relative moves (turnover = tax drag):
        if abs(drift_rel) > rel_band and abs(drift_abs) > abs_band:
            action, delta = ("BUY" if tgt*total > mv else "SELL"), tgt*total - mv
        else:
            action, delta = "HOLD", 0

# No conviction, no entry signal, no exit. The strategy IS the target
# weights; the discipline is holding the tilt when it underperforms.
# Prefer new-contribution rebalancing; sell last, and tax-aware.

Run on demand each quarter: python3 trackc2_tilt.py prints the drift table and the trades, if any. It is not a scanner and has no daily cadence — checking it more often than quarterly only tempts you to tinker.

Why this is the value home for most operators. Track A-2 is the system's other value sleeve — concentrated, single-name, discretionary — and it lives or dies on a judgment edge: the ability to correctly value a neglected small business, the one thing no code can confirm for you. C-2 is the opposite by design. It asks for no business analysis, no moat call, no falling-knife judgment. It harvests the same documented value premium mechanically, through factor exposure and a rebalance band, so the only thing required of you is the patience to hold the tilt through its droughts. That makes C-2 the right home for the value premium unless and until A-2's judgment edge is genuinely demonstrated — most of the premium, none of the discretion. The two are not redundant: C-2 is the factor, A-2 is the security. Run C-2 by default; earn the right to run A-2.
Track C-3

The thematic sleeve — own the trend, don't trade it

An optional, hard-capped sleeve inside the core for high-conviction secular themes — physical AI and robotics, the energy build-out behind compute, AI infrastructure. These names fail every value screen by construction, which is exactly why they cannot live in A-2. C-3 is a growth judgment expressed on C's temperament: bought slowly, held for a decade, and sized so that being entirely wrong is recoverable.

RuleSettingWhy
Sleeve cap≤20% of C at cost≈11% of total portfolio. Full failure costs a bad year, not the plan.
Hard ceiling30% of C by valueWinners run untrimmed until here; breach forces a trim back to 25%.
Positions3–5 maxSingle names or thematic ETFs. Conviction, not a mini-index.
Per-name cap≤8% of C at costNo single thesis can sink the sleeve.
Entrythirds / 6–12 moStaged in; thematic entries are the easiest to overpay for in one gulp.
Leverage & optionsnone, everThe sleeve's entire edge is surviving long enough to be right.
Holding period10-yr defaultThe thesis is a decade-scale supply/demand claim, not a quarter's story.
Exitthesis break onlyNever price. Each entry is filed with a written "what proves me wrong."
Review cadenceannual, writtenOne thesis memo per name per year. No daily quotes, no scanner.
Drift bands do not apply here. check_rebalance() governs the beta and factor sleeves, where the strategy is the target weight. C-3's strategy is the thesis, and thematic winners are power-law distributed — trimming them on a 5% band would systematically amputate the only positions that justify the sleeve. The 30% hard ceiling is the sole trim trigger. While C-3 is active, the value/quality/intl sleeves scale down pro-rata; the broad core's weight is untouched.
Boundary rules against the other sleeves. Against A-2: if a name passes a value screen, it belongs to A-2's discipline, not here — C-3 exists precisely for names value logic can never buy, and letting the two share a book pollutes both. Against B: a C-3 holding can never simultaneously be a B catalyst trade, and there is no "trading around" a core position — renting and owning the same name teaches your hands the wrong reflex. Against the override: VIX-triggered capitulation buying deploys into the broad core only, never into C-3 — panic is exactly when a discounted secular thesis and a busted story look identical, and the override must stay judgment-free.
The Place

What the core is for

Track C is the gravity the rest of the system orbits. It's why the alpha sleeves can afford to be aggressive and the income sleeves can afford to be patient: the bulk of your compounding is happening here, mechanically, regardless of what B, B-2, A, or A-2 do this week.

TrackRoleReacts to volatility by
C / C-2 / C-3Compounding core (55%) + value tilt + thematic sleevebuying MORE (override, core only)
A / A-2Income/landlord (30%) + value sleeveholding for yield / value
B / B-2Alpha: catalyst + trend (15%)sitting out / standing down

Keep the capital lines separate. The override reserve is part of C's allocation, pre-designated — never funded by selling A or B in a panic, which is exactly when those sleeves are worst to liquidate. The whole architecture only works if each track's risk and cash stay in their own lane.