Systematic Equity
Research Portfolio
> Long-horizon systematic equity research with out-of-sample validation and institutional-style risk analysis. A disciplined factor approach to US equities.
Performance Summary
| Period | Strat CAGR | SPY CAGR | Strat Sharpe | SPY Sharpe | Strat DD | SPY DD |
|---|---|---|---|---|---|---|
| 2000–2007 | +12.2% | +1.8% | 0.394 | -0.076 | -25.4% | -46.7% |
| 2008–2015 | +15.2% | +6.6% | 0.612 | 0.294 | -43.4% | -51.8% |
| 2016–2023 (OOS) | +14.5% | +13.3% | 0.591 | 0.545 | -28.1% | -33.7% |
| Full 2000–2023 | +14.0% | +7.0% | 0.535 | 0.253 | -46.5% | -54.4% |
> The 2016–2023 segment is true out-of-sample — all parameters were frozen before this data was observed. Max drawdown is lower than benchmark in every sub-period.
Cumulative Growth (2000–2023)
Drawdown Profile (2000–2023)
Methodology
A systematic research engine.
The strategy uses systematic signal-driven analysis to identify equities. The approach targets return premia with clear theoretical rationale. Holdings are structured for extended periods, and the approach is designed to capture persistent return sources through a disciplined, repeatable process.
> Performance varies across market regimes. Returns are driven by identifiable return sources — not unexplained alpha.
Signal Construction
A signal-driven approach with clear theoretical rationale. Constructed from market and company data. No look-ahead, no data-mining.
US Equity Universe
Diversified US equity universe with portfolio-level construction constraints.
Portfolio Construction
No optimization overfit. Portfolio construction follows pre-specified rules designed to be resilient under conservative transaction cost assumptions.
pie_chart Factor Attribution
> Returns are primarily consistent with identifiable return sources. No statistically significant alpha detected — the signal effectively harvests multiple return drivers simultaneously. Evaluated via multi-factor model.
> Portfolio construction characteristics naturally result in balanced risk distribution. Signal profile supports the current construction with structural reasoning.
Research Process
Mechanism First
Every signal should be grounded in structural reasoning. Empirical patterns without return-side implications are measurement artifacts, not strategies.
Lock Before Search
No factor research until a specific identified deficiency exists. "Increasing CAGR" is not a valid reason. Correct question: what risk does this reduce?
Pre-Specified Exits
Every strategy ships with yellow/red flags defined before deployment. No post-hoc rationalization. If breached, stop first — then investigate.
Documented Regime Behavior
> Performance varies across market regimes, with documented periods of relative weakness. Rather than hiding weak periods, every regime where the strategy lags its own universe has been identified, quantified, and classified by evidentiary strength.
Classified by evidentiary strength
Alternative hypotheses tested and rejected
No further search until trigger condition met
"The measurement is real. The interpretation it suggested is wrong." — Candidate mechanisms were tested directly and falsified when they failed to improve robustness.
block Research Branches
> Multiple alternative specifications were evaluated and removed when they did not improve expected behavior. No factor is added to an existing strategy without a specific identified deficiency it solves.
Risk Framework
warning Investigate, Do Not Stop
dangerous Consider Stopping
check_circle Expected Behavior
Disclaimer
Past performance is not indicative of future results. All performance data shown is based on historical backtesting and live out-of-sample tracking. Investment involves risk, including the potential loss of principal. The strategy described is not investment advice. This is a personal research portfolio, not a registered investment advisory.
Strategy by Anak Agung Ngurah Alit Mahendra Putra, Quantitative Researcher at Qwark | Last reviewed: July 2026