← Back to Flight Club

CONOPS · NO. 001

Alpha Convergence Strategy

Systematic Alpha Through Independent Agreement

INSTITUTIONAL WHITE PAPER

Strategy inception: June 27, 2025·Data through: September 30, 2026·October 2026 Edition·Stage 1 Validation Complete

Executive Summary

Alpha Convergence is an actively managed, rules-based equity strategy designed to identify and exploit periods of temporary mispricing through the convergence of multiple independent signals. The strategy seeks to generate excess returns while minimizing unintended factor concentration.

The investment process combines two investment models, each developed to capture distinct informational and behavioral inefficiencies in public equity markets. Portfolio construction emphasizes independence across positions, equal weighting, and disciplined rebalancing in order to reduce reliance on any single signal, factor, or market environment.

The strategy currently maintains a concentrated portfolio of exactly 22 equities and is benchmarked against the S&P 500 Equal Weight Index, measured through RSP, the Invesco ETF that tracks it, reflecting its equal-weight design, with no permanent strategic sector tilts. Risk is managed through position sizing discipline, diversification across independent signals, and continuous monitoring of correlation dynamics, volatility, and drawdown behavior.

No hypothetical back tests or simulated performance are presented, as such analyses are unlikely to capture the qualitative judgment, model evolution, and macro-economic decision-making embedded in the process. Additionally, a back test scores companies on an average of what they used to be rather than what they are today.

Early live results across a varied market backdrop have included sharp factor rotations, elevated volatility, and episodic risk-off conditions. While these observations provide useful insight into how the strategy behaves under stress relative to its design objectives, longer operating history across full market cycles is required before drawing durable conclusions.

Alpha Convergence is best viewed as an emerging strategy with a clearly articulated process, explicit risk awareness, and a commitment to transparency. Ongoing evaluation will focus on the stability of signal behavior, correlation dynamics, drawdown control, and consistency relative to benchmark over time. The record is short by calendar. The measurement around it is not: the return series is reconciled to a tracking account at the custodian, each rebalance period is tested as a separately constituted cohort, every session is classified by market regime, and the record is run through formal significance tests, so a good deal can already be observed, tested, and, if the thesis were wrong, falsified within 317 trading days.

A note on claims: statements in this paper are of four kinds, and we keep them visibly separate. Observed live results carry dates and figures. External research is cited. Design intentions describe what the process is built to do. Forward expectations are labeled as expectations.

Performance Summary

The following metrics reflect 317 trading days (June 27, 2025 through September 30, 2026). On June 29, 2026 the strategy completed 252 trading days, the firm’s predetermined first live-validation milestone. Figures are shown at two horizons with equal prominence, 1 Year (the trailing 252 trading days) and Since Inception (the full record). Returns are presented both gross and net of a 2% annual model advisory fee, the top of the firm’s 1-2% fee schedule, applied monthly; transaction costs are deducted from both. Net figures reflect the model fee only; actual advisory fees vary by mandate and will affect realized returns. See Form ADV Part 2A for the applicable fee schedule. Two comparisons are shown. RSP, the equal-weight S&P 500, is the evaluation benchmark: it shares the strategy’s equal weighting and its predominantly large-cap universe, so excess return against it isolates stock selection. SPY, the cap-weighted S&P 500, is shown as the market reference because it is what most investors experience as the market; returns relative to SPY also reflect the spread between equal-weight and cap-weight investing and are not by themselves a measure of selection skill.

Asymmetric Return Profile

The strategy exhibits favorable asymmetry: on a since-inception basis capturing 98.1% of benchmark gains while experiencing 75.6% of benchmark losses. This 1.30:1 capture ratio reflects the quality bias inherent in our dual-screen approach. Both capture legs softened this period: downside capture rose from 74.7% to 75.6% while upside capture eased from 101.7% to 98.1%, moving the ratio from 1.36:1 to 1.30:1 (prior-edition figures restated on the total-return benchmark basis). The asymmetry remains favorable, and capture figures continue to move materially on small numbers of sessions at the current sample size.

Drawdown and Recovery Expectations

The observed maximum drawdown of -8.46% occurred on March 30, 2026 from the February 11, 2026 peak, during the late-March 2026 stress event (VIX peaked at 31.1). This figure is informative but not necessarily representative of our forecast under more severe adverse conditions.

We expect downside capture to rise toward or above 100% during acute stress (a short-term downward move) as correlations spike and quality factors provide incomplete protection. During sustained bear markets, we also expect downside capture to run above the current since-inception level, with the quality tilt providing partial rather than full protection. During recovery phases, we expect upside capture to compress during early-cycle “junk rallies” when low-quality securities outperform.

The Strategy Within the Defensive Allocation

This section describes a client-level allocation decision that uses Alpha Convergence as its equity sleeve; it is not a measure of the equity strategy’s performance. Alpha Convergence is an equity strategy, and this paper evaluates it as one: every return figure above is for the equity strategy alone. Fixed income allocations are set client by client according to each client’s needs, so this paper does not evaluate them. During the current structural event, however, the allocation itself has been part of how we manage risk, and an equity strategy is best judged in that context as well as on its own. Since the Phase 1 defensive action in March 2026 (see Crisis Management Framework), the flagship client allocation has held approximately 55% in the strategy and 45% in short- and intermediate-term fixed income. The table below covers the 132 trading days of that defensive posture. It sets the allocation against the equity strategy alone, against SPY and RSP alone, and against the same 55/45 allocation built with SPY or RSP in place of the strategy and the identical fixed income sleeve, so the only difference between the allocation and the blends is the equity sleeve. The allocation reduced both return and drawdown relative to the equity strategy, as designed. It trailed SPY held alone, as a 45% fixed income position would be expected to in a rising equity market, and it led RSP held alone and both allocation-matched blends. Its maximum drawdown was smaller than that of the equity strategy, SPY, RSP, and the RSP blend, but deeper than that of the SPY blend.

The strategy within the defensive flagship allocation, March 24 through September 30, 2026, the structural-event period in which the defensive posture has been in force. The allocation return is the flagship allocation as reconciled to the Schwab tracking account. The blends pair SPY or RSP total returns at 55% with the identical fixed income sleeve return at 45%, so the allocation and the blends differ only in the equity sleeve. Net of the 2% annual model advisory fee, the allocation returned +10.04%. Fixed income allocations are otherwise set client by client, and actual accounts vary with allocation, the timing of contributions and withdrawals, and fees.

Why We Do Not Present Long-Term Back Tests

This paper excludes multi-year simulated back tests, for two reasons. First, a back test applies today’s selection rules to companies as they used to be. Point-in-time data can prevent look-ahead bias, but it cannot make the companies, technologies, competitive structures, management teams, or investment problems of earlier years economically the same as those the process faces today. Across a multi-year simulation each company is, in effect, scored on a time average of its past self, while our process evaluates what a company is today and where it is heading, so that average is a poor proxy for our selection process and can be actively misleading. Second, portfolio construction incorporates qualitative and macroeconomic judgments made in real time, and those judgments cannot be reconstructed afterward without hindsight contaminating them. A long back test of this evolving process would therefore communicate a degree of historical replicability and precision we cannot honestly claim. We present live results only, with their confidence intervals, and let the record lengthen.

Information Advantage

Dual Architecture as Edge Source

Consider how a bank evaluates a loan application: two independent systems examine the same borrower from fundamentally different angles.

These systems have low correlation by design. When both systems independently approve the same borrower, the loan gets funded.

The Alpha Convergence Parallel

Alpha Convergence applies the same logic to securities selection: two independent analytical systems measuring fundamentally different phenomena.

Correlation as Market Regime Indicator

We monitor correlation between our models weekly as an indicator of how momentum and fundamental signals are aligning. We interpret these movements alongside external indicators like VIX, SPY, and IWM to identify potential regime (market conditions) changes.

The Agreement Filter

Our dual screen requires agreement.

Information Edge Cycling: The Di Mascio Framework

Di Mascio, Lines, and Naik (2017) studied transaction-level data from 752 institutional portfolios and documented a robust finding: newly purchased stocks earn positive risk-adjusted returns that decay predictably over time. Alpha averages 36 basis points in the first month, follows a power-law decay with a half-life of approximately four months, and reaches zero by month twelve with no subsequent reversal. The shape of this decay has direct implications for portfolio design.

The critical finding is the horizon mismatch: the average institutional manager holds positions for 2.2 years (approximately 554 trading days), which is roughly double the 12-month window in which alpha exists. Once alpha is exhausted, continued holding adds zero marginal return while tying up capital that could be redeployed into fresh signals. Many active managers give up value not through poor stock selection but through holding positions long past the point where their informational advantage has expired.

How AC Addresses the Horizon Mismatch

Alpha Convergence is designed to harvest alpha during the steep portion of the Di Mascio curve and rebalance before marginal alpha declines to the flat zone. The information cycle operates as follows:

Empirical Validation: Periods 2 Through 7

Each rebalance period is an independently constituted test of the information-edge thesis. At each rebalance the portfolio is rebuilt from fresh dual-model agreement signals, so each period’s cohort of names is a new, prospective test of whether the edge reappears and is harvested within the intended holding window. Independent here means independently constituted; it does not mean the periods’ returns are statistically independent, since they share one market, one process, and one manager. In this paper a period is the holding window between two rebalances, and its cohort is the portfolio selected at the start of it. Results are shown alongside the alpha an average institutional manager would be expected to earn over the same holding window. P6 closed on September 21, 2026 at 63 trading days, and P7 opened on September 22 and is open at 7 trading days.

Because each cohort is selected afresh, consistency across periods is unlikely to rest on a single position or a single rally. The closed P2-P6 periods delivered between +3.34% and +7.38% of excess versus RSP, materially larger in magnitude than the Di Mascio reference curve. The two are not the same statistical object: Di Mascio measures stock-level, risk-adjusted alpha in event time for an average institutional manager, while these figures are portfolio excess returns versus RSP, so the comparison indicates scale rather than a precise ratio. The most informative period is P4: AC returned +1.63% while RSP fell -3.29%, generating +4.92% of excess in the only closed period of the record in which RSP fell, rather than in a rally. P5 closed at the top of that distribution after a Month 1 softened by Phase 1 defensive positioning, a complete absorb-and-recover lifecycle. P6 opened soft, recovered to close Month 1 at +2.41%, absorbed a negative Month 2 of -0.09%, the first negative month of the record, traced to a single-name event of the kind that recurs from time to time in a 22-position equal-weight book, closed Month 3 at +1.02%, and finished at +3.34% over 63 trading days, the lowest of the closed periods but positive and ahead of both benchmarks. P7 opened on September 22 and stands at -0.30% over seven trading days, a reading too short to carry any information about the new book. Future periods may land anywhere in this range depending on conditions, but the repeatability across independently constituted cohorts is consistent with the design.

Per-period excess versus RSP. Each period’s cohort is newly selected; P4 produced +4.92% excess while RSP fell -3.29%. Period boundaries are the authoritative firm attribution boundaries; each rebalance day belongs to the closing period, so the March 23, 2026 session is attributed to P4 and the period rows sum to the 317-day total. Net figures deduct the 2% annual model advisory fee applied monthly; because the fee is charged at each calendar month-end, per-period net figures can differ from an exact daily accrual by up to about one month’s fee (roughly 17 basis points) for periods that begin or end mid-month; the since-inception total is unaffected.

Portfolio Construction

We build portfolios with exactly 22 concentrated positions through rigorous correlation management. Each portfolio represents high-conviction positions designed, through the correlation management described below, to approach the diversification of a larger randomly selected portfolio; the research below supports the base case that about 20 equal-weighted holdings remove most idiosyncratic risk. Positions are drawn from a universe of more than 1,300 liquid securities; holdings are predominantly large-capitalization U.S. equities and also include ADRs and, occasionally, mid-cap companies where a sector such as real estate lacks suitable large-cap candidates. We estimate strategy capacity at roughly $500 million to $2 billion.

High-Conviction Diversified Portfolios

Antón, Cohen, and Polk (2021) demonstrated managers’ highest-conviction positions typically outperform the market by 2.8% to 4.5% annually, while positions beyond a manager’s best ideas contribute little or no alpha. The authors concluded, “investors would benefit if managers held more concentrated portfolios.”

Sufficient Diversification for Risk Reduction

Foundational research on equal-weighted portfolio diversification demonstrates risk reduction benefits diminish rapidly beyond approximately 20 securities. Raju and Agarwalla (2021) confirmed that a 20-stock equal-weighted portfolio diversifies away approximately 90% of idiosyncratic risk on average, while the CFA Institute’s “Peak Diversification” study (2021) found that for large-cap equal-weighted portfolios, peak diversification occurs between 15-26 stocks, with minimal incremental benefit beyond this range.

Enhanced Diversification Through Correlation Management

We target:

Academic research on ensemble methods demonstrates that near-zero correlation between predictors, whether slightly positive or negative, maximizes variance reduction while ensuring models address related investment problems (Dietterich (2000); Brown et al. (2005)).

Signal-Driven Reconstitution. Portfolio rebalancing is signal-driven, aligned to the Di Mascio alpha decay lifecycle and bounded by a maximum holding period of 90 calendar days. Di Mascio et al. (2017) documented that alpha on newly purchased securities decays with a 4-month half-life and reaches zero by month 12. On those reported parameters the marginal rate is highest at purchase, and close to half of a position’s twelve-month information alpha is expected to accrue in its first quarter. A holding window of about 90 calendar days is therefore set to harvest the steepest part of the curve and reset the portfolio before most of the decay. The design sequence is fresh information, the strongest marginal edge, early harvesting, and a reset before substantial decay. Our rebalance cadence harvests alpha during the steep-slope portion of the decay curve and redeploys into fresh signals before marginal alpha diminishes.

Rebalance timing can be adjusted for execution conditions within the 90-day maximum. In September 2026 the scheduled P7 rebalance fell just ahead of the Federal Reserve’s September meeting, and we deferred execution by five trading days for greater clarity, rebalancing on September 21, the 90th calendar day of P6. Over the five deferred sessions the P6 book returned -1.63% against -0.71% for RSP; the cost relative to an earlier P7 entry cannot be observed directly. We regard the deferral as prudent risk management and expect such adjustments to be infrequent.

Crisis Management Framework

Precedents

Although Alpha Convergence launched in June 2025, Life UnLocked Partners has navigated significant market dislocations for existing clients using protocols that inform our current approach. These responses demonstrate our risk management philosophy in practice.

2022 Down Market Response

During the 2022 market decline, we executed a defensive reallocation for client portfolios:

Defensive Shift: Reallocated equity exposure to volatility-impaired, bond-like instruments (primarily fixed-to-floating rate preferred securities) to preserve capital while maintaining income generation.

Damage Path Assessment: Waited for clear damage paths to establish before seeking re-entry opportunities while earning significant alpha: for 2022 the firm-wide client record returned -4.20% net of actual fees against -11.20% for its segment-blended benchmark (supplemental firm-wide record, not a single-strategy composite).

Adjacent Opportunity Focus: Sought recovery positions “three degrees from the damage path”: companies that benefit from dislocation without direct exposure to damaged sectors.

April 2025 Volatility Response

The April 2025 market volatility provided a more recent test of our risk management approach:

Immediate Action: Smaller reallocation to preferred securities during the initial volatility spike.

Model Enhancement: Extended defensive positioning to complete reengineering of our fundamental model to its current Alpha Convergence architecture.

Controlled Re-entry: Client rotation to the enhanced Alpha Convergence model on August 19, 2025, following its first seven weeks of live validation, in which the strategy returned +4.62% against +3.12% for RSP (P1).

March-April 2026 Hormuz Stress Response

The March-April 2026 stress event provided the most recent and most severe live test of our crisis management framework. We executed a staged defensive response in the face of an unbounded oil shock and structural risk to private credit and tariff-exposed equities. Our framework distinguishes three phases of escalating defensive action, plus a reserved tail-event contingency; only the first two phases have been triggered to date.

Phase 1 (March 20-23, 2026, Capital Preservation): Following the Storm Center structural-event flag on March 19, we initiated defensive action on March 20 and completed it by March 23. The action trimmed both the equity sleeve and the VCIT (intermediate-term corporate bond) holding to fund a 15% SCHO (short-term Treasury) position, repositioning the portfolio for capital preservation ahead of the scheduled P5 rebalance. The P5 rebalance executed on March 23, with the new portfolio in place for its first session on March 24, in the context of this Phase 1 posture: equity selections diversified away from private credit exposure and tariff-vulnerable names. We deliberately retained selected oil and energy positions whose operations sit largely outside the Strait of Hormuz transit risk, providing diversification without taking on direct geopolitical exposure to the choke point.

Phase 2 (April 20, 2026, Credit De-Risking): As the Hormuz risk failed to bound and the duration of the stagflation regime extended, we executed a second defensive shift on April 20, selling the remaining VCIT position and rotating the proceeds into additional SCHO. This trade reduced credit duration and credit spread risk inside the fixed income sleeve while preserving income generation. The objective was to insulate the bond allocation from any further widening in corporate credit spreads driven by the oil shock filtering through to investment-grade issuers.

Current Allocation State: Since Phase 1 in March 2026, client portfolios have held approximately 55% equity / 45% fixed income, and since the Phase 2 trade in late April roughly 30% of overall portfolios has been allocated to SCHO. This is materially more defensive than the standard positioning of approximately 65% equity, and reflects our judgment that the Hormuz event has structural rather than transitory characteristics. With the ceasefire MOU signed in June, we have backed off our oil sector avoidance as of the June 22nd rebalance while maintaining the same fixed income posture and allocation against continued uncertainty. With the September 21 rebalance into P7, we reinstituted reduced exposure to oil and to 10-year rate sensitivity for defensive purposes; this is a reduction, not a full avoidance. We also stayed away from home- and auto-related risks, even though names in those areas scored well within our models. The two-phase design preserves optionality: the SCHO allocation can be unwound rapidly when the Storm Center signal de-escalates, and the equity sleeve remains positioned in dual-model agreement names, currently with defensive tilts toward lower oil and interest-rate sensitivity that we expect to ease as the structural event resolves.

Phase 3 (Conditional, Not Yet Triggered): The framework defines a third escalation step that has not been required during this event: an additional equity-to-fixed-income trim that would bring portfolios to approximately 45% equity / 55% fixed income. This would be triggered if the Storm Center signal escalates further or if the four-week material-change rule confirms a deeper regime shift. The framework also reserves a fourth contingency, mostly cash, for a true tail-event scenario; that contingency has never been triggered and is not described in detail here.

Framework Implication: The staged response demonstrates a key feature of our crisis management approach. We do not treat defensive repositioning as a single binary decision. Phase 1 addressed the initial structural-event signal with the most easily reversible defensive shift. Phase 2 escalated only when the underlying risk drivers failed to resolve within their initial estimated bounds. Phase 3 remains in reserve. This staged approach reduces the cost of false positives while preserving the ability to escalate when warranted by the evolving evidence.

The “Three Degrees from Damage Path” Framework

Our recovery positioning philosophy avoids both the risks of bottom-fishing in damaged securities and the opportunity cost of complete sector avoidance. Instead, we seek adjacent beneficiaries:

This framework provides asymmetric exposure: participation in sector recovery without direct exposure to damaged balance sheets or uncertain restructuring outcomes.

Hormuz Application (March 2026): First-degree positions (tanker operators and producers with direct Strait of Hormuz throughput exposure) were excluded from the P5 portfolio. Second-degree positions (refiners and petrochemical companies dependent on crude flows through the Strait) were treated with caution. Third-degree targets (companies with mild energy input sensitivity whose prices declined from broad risk-off sentiment rather than operational disruption) formed the core of the rebalanced portfolio. The P5 portfolio accepted mild oil exposure as structurally unavoidable while maintaining deliberate distance from the primary damage path.

Defensive Repositioning Triggers

Our risk management framework includes quantified triggers for defensive repositioning. These thresholds inform, but do not mechanically determine, allocation decisions:

Regime Performance Analysis

Stress-Test Periods

Alpha Convergence launched June 27, 2025 after our model completion achieved operational readiness. What followed was an early stress test in which the strategy encountered multiple conditions historically adverse to quality-focused mean reversion strategies:

Trash Rally conditions (small cap speculation outperforming quality);

Risk-Off Selloff conditions (broad market decline); and

Momentum-led recovery (sentiment-driven broad rally)

Market Regime Timeline

Markets have cycled through several distinct regimes, each presenting a different challenge for a quality-focused strategy:

Performance by Regime

The following table shows Alpha Convergence performance against benchmarks across each regime:

AC = Alpha Convergence. RSP = S&P 500 Equal Weight ETF (primary benchmark). SPY = S&P 500 ETF. IWM-RSP Spread measures small cap speculation vs. quality; positive spread indicates “trash rally” conditions hostile to quality strategies. Hostile and Favorable describe market conditions, specifically whether the factor environment has historically rewarded or penalized quality-focused selection; they are not labels for the strategy’s results. Four early sessions lack a full 5-day spread window and are unclassified, so regime days total 313 of 317, and the total hostile share in the timeline is from this classification (241 of 313 classified days). Regime figures are calculated gross of fees; net-of-fee total returns are presented with equal prominence in the Performance Summary, and net excess by rebalance period appears in the period table. Regime classification uses sigma bands recomputed from the full classified sample at each update; as the sample grows, band movement can reclassify a small number of historical sessions, so regime day counts and per-regime statistics may shift between editions without any change in the underlying return data. Methodology and current band parameters are documented in the firm Regime Performance Report.

Preliminary Regime Conclusions

Performance in Hostile Regimes: Across 241 trading days classified hostile to quality strategies, Alpha Convergence returned -4.01% while RSP fell -18.09%, generating +14.08% cumulative alpha vs RSP. The two trash-rally regimes were the largest contributors (Severe Trash Rally +6.38% over 38 days, Momentum/Trash Rally +6.53% over 57 days). A quality-focused strategy should suffer when small cap speculation dominates; instead, we kept pace with the rally and added alpha while maintaining quality discipline.

Performance in Favorable Regimes: Across 72 trading days classified favorable, Alpha Convergence returned +52.70% while RSP returned +40.60%, generating +12.10% cumulative alpha vs RSP. The Broad Market Rally bucket produced +11.30% alpha over 54 days as the strategy participated in upside. The Quality Rally bucket (18 days) sits at effectively zero (0.00%) on a thin sample and is discussed below.

VIX Regime Consistency: Across VIX regimes, the strategy generated positive alpha in two of four buckets: Normal Vol +39.40% (228 days) and High Vol +1.50% (15 days). Low Vol (38 days) produced -2.27% and Elevated Vol (36 days) produced -1.72% cumulative alpha. The Normal Vol band remains the dominant source of alpha.

Cumulative Alpha: +33.75% alpha vs. RSP across 317 days of varied market conditions. This represents an annualized information ratio of 1.98.

Expected Soft Spots: The two regimes without positive alpha, Quality Rally (0.00%, effectively flat) and Risk-Off/Quality Flight (-1.13%), share a common structure: both compress the cross-sectional dispersion the strategy is built to harvest. The selection edge operates against a quality baseline, owning quality names the market has temporarily mispriced relative to that baseline. In a Quality Rally the market is already rewarding the baseline itself, so the gap between quality and the crowd narrows and there is less dispersion left to capture. In Risk-Off conditions correlations spike toward one and cross-sectional dispersion collapses outright, taking the strategy’s raw material with it. We therefore expect thin or negative alpha in these regimes rather than treating it as anomalous. The same mechanics explain where alpha concentrates: trash-rally conditions, where the gap between quality and speculation is widest, supplied the largest contributions. This is a structural trade-off of the design, not a calibration error, and it is disclosed as such.

Regimes Not Yet Tested

We seek continued stress testing. Current evidence shows positive cumulative alpha in three of four hostile regime categories, with a documented exception in Risk-Off/Quality Flight. We also caution against extrapolating this pattern to untested regimes until we have live data.

Implementation

Conclusion

Alpha Convergence is a deliberately transparent, live-data-driven strategy. Its advantage emerges from structurally independent models and disciplined rebalancing. The core insight: two analytically independent screens measuring genuinely different phenomena provide more reliable signal when they agree than either screen alone.

Contact

Life UnLocked Partners
Mark Tennenbaum, Chief Investment Officer
mark@lifeunlocked.partners

Statistical Appendix: Hypothesis Testing

The following tests evaluate whether observed metrics are statistically distinguishable from zero. All tests use 317 trading days (June 27, 2025 through September 30, 2026). n = 317.

Interpretive Notes

The Information Ratio standard error follows Lo (2002), "The Statistics of Sharpe Ratios," Financial Analysts Journal. The approximation SE = sqrt((1 + ratio^2/2)/n) is standard but assumes iid returns. Daily financial returns violate this assumption through autocorrelation, volatility clustering, and non-normal tails, so the reported t-statistics and p-values should be understood as indicative rather than exact. The Sharpe Ratio is assessed under a Newey-West HAC standard error (q=4) per Lo (2002) to account for that autocorrelation.

The Information Ratio of 1.98 eased over the added month. IRs of this magnitude remain characteristic of early-stage strategies with favorable sequencing, and almost always compress toward the 1.0 to 1.5 range over full market cycles as the strategy encounters a broader distribution of market conditions. We present the current figure as a point-in-time observation, not a long-run expectation.

The Win Rate 95% confidence interval of [49.1%, 60.1%] includes 50.0%, and the binomial test is not significant at the 5% level (p = 0.103, two-sided). As the sample grew, the proportion of winning days moved toward a coin flip even as the average size of winning days stayed favorable. Win rate measures how often the strategy beats the benchmark, not by how much.

All hypothesis tests assume returns are drawn from a stationary process. To the extent that the strategy's signal quality, market conditions, or portfolio composition have changed materially during the observation period, the assumption of stationarity is approximate. The firm’s 252-day first live-validation milestone has been reached; formal tests for structural breaks are deferred until additional post-validation sample accrues.

Three of the four metrics are statistically significant at the 5% level (Excess Return, Information Ratio, and Sharpe versus zero). These are different diagnostics of the same live return history, not four independent experiments; the Win Rate is non-significant (p = 0.103) as the proportion of winning days sits near a coin flip even though the average size of winning days remains favorable. Sharpe inference is assessed under the Newey-West HAC standard error, whose 95% interval [0.44, 3.70] excludes zero; the separate test against the firm’s 1.10 Sharpe hurdle reported in the Statistical Note is not yet met. These results are directionally supportive but should be interpreted with appropriate caution given the limited sample size, non-iid return dynamics, and the expected compression of early-stage performance ratios. The firm’s 252-day first live-validation milestone was completed on June 29, 2026.

Statistical Reference

Statistical reference. Each row is a different diagnostic of the same 317-day live return history, not an independent experiment, and the rows answer different questions: whether the strategy beat RSP on average, after fees, after controlling for beta, on most days, and per unit of risk. Gross and net refer to the 2% annual model advisory fee. Annualized active-return estimates and intervals are arithmetic (daily mean times 252). Where both a one-sided and a two-sided p-value are in use across the firm’s documents, both are shown.