Structural Long–Short — AlphaBlock
α1phablock
← The 3N™ Methodology
// Structural Long–Short · SLS

The industry trades factors. We isolate the spread benchmark design leaves behind.

The one thing to know: a market-neutral strategy that earns the gap between a redesigned, scientific portfolio and the winner-biased, concentrated index it replaces.

Structural Long–Short is a systematic, dollar-neutral strategy: long the 3N™ probabilistic allocation, short its cap-weighted benchmark — same universe, opposite construction. The performance spread between the two allocation systems is the product: alpha isolated in a low-risk wrapper, without leverage or forecasts.

~30
structural spreads monitored
4–10
active spreads at any time
6–8%
target annualised volatility
$10B+
estimated strategy capacity
01 Two systems, one spread

Same universe. Opposite construction. The difference is the trade.

Every spread pairs two portfolios built from the same large-cap universe — so the differential reflects benchmark design, not stock selection.

LONG LEG · 3N™
The multi-bias allocation

3N treats the market as a closed statistical system of interacting biases — persistence, reversion, dispersion, concentration — and distributes exposure probabilistically across them.

SHORT LEG · MCAP
The winner-bias machine

Cap-weighting allocates by market value: as a price rises, its weight rises, and index flows buy more of it. Concentration is self-reinforcing — a single bias, compounding.

EXHIBIT A — Same universe, two weight curves
MCAP — a handful of names carry the index 3N™ — weight follows information, not size index weight constituents, ranked by market cap →
Illustrative. The short leg concentrates by price; the long leg distributes by probabilistic bias states. The gap between the curves is the structural spread.
Spreadi,t = R3N,i,t − RMCAP,i,t The spread is not a factor. It is the structural difference between two ways of building the same portfolio — and factor crowding cannot arbitrage it away.
02 The mechanic

Drawdowns in the spread are convergence opportunities.

When passive flows stretch the spread past a statistical threshold, the divergence has historically tended to normalise over a bounded horizon. The strategy trades exactly that — nothing else.

spread high-water · 3N − MCAP activation threshold · ≈−5% broad / −10% sector day 0 40 80 day 120 designed for convergence · ~120-day average, not a fixed window drawdown enter · dollar-neutral long 3N allocation · short MCAP ETF · equal-weighted exit · ~day 120 · on convergence spread reset
01 · MONITOR

~30 structural spreads, tracked continuously: sector SPDRs plus QQQ, OEF, TSX 60, Nikkei, FTSE, GDX, Indonesia 30 and more.

02 · TRIGGER

A spread becomes eligible only on a statistically significant drawdown — ≈5% for broad-market exposures, ≈10% for sectors.

03 · PAIR

Enter dollar-neutral: 3N allocation long, benchmark ETF short. Active spreads are equally weighted — no conviction sizing, no concentration.

04 · RESET

Hold until the spread converges — about 120 days on average, not a fixed window — then exit systematically. A rolling monthly refresh keeps the book an overlapping set of positions.

03 Risk architecture

Built to minimise risk — not to maximise return.

The framework's discipline is the risk control: every rule below is predefined, and none of it is overridden by judgment.

No leverage
Returns come from the spread, not from amplification.
Dollar-neutral at entry
Minimal directional exposure; beta drift is accepted and reset at expiry.
Equal-weight spreads
No single signal dominates the book.
Liquidity & borrow screens
Large-cap universes and liquid ETFs only; borrow feasibility confirmed before every short.
Sector-exposure monitoring
Continuous checks against unintended concentration.
Convergence-based exit
Every position is designed to exit as its spread converges — about 120 days on average.
What this strategy is not
Not discretionary stock-picking Not factor timing Not continuous hedging Not leveraged Not high-frequency arbitrage
04 Signal quality · June 2026

The Consistent Signal

Backtested for 10 years since January 2015. The headline measures:

76.3%
signal win rate across closed trades since 2015
~89%
peak win rate in the most-populated holding bucket (91–120 days)
+11.2%
median annualised return per closed trade
26.4pp
interquartile range of outcomes — the middle 50% of trades spans just 26.4 points
EXHIBIT B — Middle 50% of trade outcomes, annualised
0% −10% +20% +40% median +11.2% Q1 · +0.5 Q3 · +26.9
Box = interquartile range (middle 50% of closed trades), line = median, annualised per trade. The middle half of all outcomes sits inside a 26.4-point band — a signal that does not depend on extreme outcomes.

Returns build with duration rather than decaying — trades held beyond a year averaged +15.4% absolute. The signals identify structural trends, not short-term price noise.

Source: AlphaBlock internal signal research, June 2026; trade history from January 2015. Per-trade signal statistics on closed trades, gross of fees and costs — not portfolio returns, not a portfolio CAGR; position sizing and overlapping holds are not reflected. Model results have inherent limitations and do not represent the performance of any fund or account. Past performance does not guarantee future results.

05  Capacity & access

Institutional by design.

Large-cap universes and liquid ETFs keep market impact low at scale — estimated capacity exceeds $1 billion — and the universe is built to expand toward the ~100 most liquid global ETFs, STOXX50 and Australia next.

The strategy is available to institutions through licensing and managed deployment — and, like everything we run, it starts with proof: a sandbox on your mandate before any capital moves.