Questions
institutional allocators
ask us.

The maximum drawdown across all 17 backtested periods was 36.7% in 2011 (European debt crisis), followed by 30.5% in 2022 (Ukraine/inflation). Crisis periods tend to produce the largest drawdowns, but also the highest ultimate returns.

Crucially, crisis-period drawdowns were followed by the strategy's strongest absolute returns: +41.33% in 2008 and +45.26% in 2020. The strategy benefits from volatility rather than being harmed by it — consistent with the positive convexity property documented in the research.

Low-volatility, unidirectional regimes are the BLT Engine's weakest environment. The strategy's mean-reversion mechanism requires sufficient price oscillation around structural equilibrium levels. In sustained trending markets (like 2016 or 2019), deviation from equilibrium may be persistent, triggering stop-losses before reversion occurs.

The four negative years in the backtest (2009, 2016, 2019, 2025) were all characterised by either post-crisis FX normalisation or sustained low-volatility regimes. Mean losses in these years averaged approximately −7.3%.

The BLT Engine operates with a mean win rate of 43.7% across all periods — well below 50%. This is a structural feature, not a weakness. The strategy generates a mean profit-loss ratio of 3.69×, meaning winning trades are on average 3.69 times larger than losing trades.

This asymmetric payoff profile is characteristic of convex strategies and is confirmed by the Sortino premium: in 11 of 13 profitable years, the Sortino ratio exceeds the Sharpe ratio, indicating that a significant fraction of total volatility is upside volatility.

Standard Sharpe-ratio-based allocation frameworks will systematically underestimate the strategy's true risk-adjusted return profile for this reason.

No. The BLT Engine uses fixed percentage position sizing with no leverage. Each trade represents a fixed percentage of equity, and per-instrument and portfolio-level position limits are strictly enforced. This design choice is deliberate: leverage would distort the structural decorrelation properties and introduce a risk profile inconsistent with the strategy's theoretical foundations.

The BLT Engine is a low-frequency systematic strategy, averaging 63 trades per year across the backtested period (range: 37–111 trades/year). This low trading frequency has two important implications:

• Transaction cost efficiency: Mean fees represent only 3.8% of gross profits across all periods
• Overfitting resistance: Low frequency substantially reduces the parameter space available for in-sample overfitting relative to high-frequency strategies

Most alternative strategies claim uncorrelated returns. In practice, their low measured beta arises from statistical averaging (long/short equity), regime-dependent behaviour (trend-following), or active hedging (options strategies) — and typically fails precisely when decorrelation is most needed: during market stress.

Category IV structural decorrelation is qualitatively different: the BLT Engine's near-zero market beta (β̄ₘ = 0.037, p = 0.540) arises from the instrument universe itself. FX pairs and precious metals ETFs carry no structural dependence on equity market direction. No hedging is required to maintain β ≈ 0 — it is a property of the instrument construction.

The mean-variance optimal allocation to a near-zero beta, positive-Sharpe asset is given by:

At the BLT Engine's observed Sharpe and volatility profile, this formula typically implies a 10–30% satellite allocation, consistent with conventional institutional satellite sizing. Adding any uncorrelated, positive-Sharpe strategy strictly improves the combined portfolio Sharpe ratio regardless of its magnitude: SR_combined = √(SR_P² + SR_A²).

OrgaX LLC is currently operating in a research and development capacity. The company is not currently registered as an investment adviser, fund manager, or broker-dealer in any jurisdiction.

Any future fund management or advisory activities will be conducted in full compliance with applicable regulatory requirements in the relevant jurisdiction, including obtaining all necessary registrations, licences, and approvals prior to accepting investor capital.

Prospective investors and their legal advisors should conduct full due diligence on the regulatory framework applicable to any contemplated investment structure.

All seven working papers are independently published on SSRN (Social Science Research Network), a preprint repository widely used in academic finance. They have not been formally peer-reviewed by an academic journal — a distinction we maintain with full transparency.

The empirical validation within the papers uses four complementary statistical methodologies:

• Bootstrap confidence intervals (n = 10,000 resamplings)
• Permutation tests — confirms results are not due to chance ordering
• Monte Carlo equity curve simulation — actual performance at 50.7th percentile, ruling out lucky sequencing
• Regime-conditional analysis — crisis vs. normal period beta comparison

All four methodologies independently confirm the zero-beta property (p ≈ 0.540 across methods). The identical-parameters design across all 17 periods substantially mitigates overfitting concerns.