{ "experiment": { "alloc": { "labels": [ "2026-01" ], "series": [ { "name": "SPY", "values": [ 0.6 ] }, { "name": "CASH", "values": [ 0.4 ] } ] }, "alloc_as": null, "alloc_monthly_as": null, "alloc_monthly_ours": { "labels": [ "2026-01" ], "series": [ { "name": "SPY", "values": [ 0.6 ] }, { "name": "CASH", "values": [ 0.4 ] } ] }, "alloc_ours": { "labels": [ "2026-01" ], "series": [ { "name": "SPY", "values": [ 0.6 ] }, { "name": "CASH", "values": [ 0.4 ] } ] }, "alloc_yearly_as": null, "alloc_yearly_ours": { "labels": [ "2026" ], "series": [ { "name": "SPY", "values": [ 0.6 ] }, { "name": "CASH", "values": [ 0.4 ] } ] }, "analysis_page": null, "answer": "Reproduced: meta-max-sharpe base 10.64% vs lev2 10.32% on same 1991-385m window (mechanical 2x would be 21.78%). No display/annualization bug (diff 0.0). Order: sleeve levered path returns before aggregation, post-filter renorm. Selection changes ~31% months (38->10 sleeves). Gap spread across many months (top10 =33% of gap). Cause is combination of window truncation, intentional feasibility filter+renorm, vol drag/daily reset path dependence, missing-month skip. No implementation bug; misleading-but-correct display.", "bead_ids": [ "qm-wtfm" ], "cost_variants": {}, "coverage": { "alloc_monthly_labels_n": 12, "alloc_yearly_labels_n": 0, "as_alloc_rows": 0, "as_history_start_year": null, "as_returns_n": 0, "as_yearly_alloc_rows": 0, "coverage_generated_at": "2026-01-01T00:00:00Z", "db_path": "data/market.sqlite", "db_warm": true, "gap_years": null, "limiting": { "effective_start": "1993-01-29", "id": "SPY", "kind": "ohlcv", "native_start": "1993-01-29", "proxied": false, "proxy_chain": [], "span_basis": "db" }, "macro_meta": [], "macros": [], "monthly_labels_n": 1, "ours_alloc_rows": 1, "ours_returns_n": 1, "ticker_meta": [ { "end": "2026-07-24", "n_rows": 8428, "start": "1993-01-29", "ticker": "SPY" } ], "tickers": [ "SPY" ] }, "date": "2026-08-17", "disposition": "completed - correct but misleading behavior clarified on Meta page and detailed in Results entry", "equity": { "experiment": [ 1.0 ], "labels": [ "2026-01-01" ], "spy": [ 1.0 ], "spy_labels": [ "2026-01-01" ] }, "errors": "Lev windows truncated to ETF validity intersection; weighted-meta window 54 months only; per-sleeve F4 estimated", "follow_on": [ { "status": "open", "title": "Evaluate financing-cost-adjusted levered cohorts" } ], "followup_command": "gh issue create --title \"Follow-up Lower CAGR in leveraged meta strategies investigation\" --body \"Follow-up to qm-wtfm: Do leveraged meta variants show lower CAGR due to math vs selection vs bug?\" --label \"priority:p2\" --label \"pipeline:research\" --label \"agent:ready\"", "gh_command": "gh issue create --title \"Follow-up Lower CAGR in leveraged meta strategies investigation\" --body \"Follow-up to qm-wtfm: Do leveraged meta variants show lower CAGR due to math vs selection vs bug?\" --label \"priority:p2\" --label \"pipeline:research\" --label \"agent:ready\"", "id": "qm-wtfm", "issue_ids": [ "qm-wtfm" ], "monthly": { "as": [ null ], "labels": [ "2026-01" ], "ours": [ 0.0 ] }, "page": "experiments/qm-wtfm.html", "performance": { "cagr": 0.0, "max_drawdown": 0.0, "max_drawdown_fallback_rule": "Use daily backtest equity when available; otherwise compound published monthly returns and label the result monthly.", "max_drawdown_observation_frequency": "monthly", "sharpe": 0.0, "sortino": 0.0, "volatility": null }, "provenance": { "artifact": "docs/site-data/meta-strategies/meta-max-sharpe-lev-2.json", "code": "research/reports/catalog_leverage_meta.py", "finding": "research/results/qm-wtfm.md", "site_data": "site-data/experiments/qm-wtfm.json", "tearsheet": null }, "question": "Do leveraged meta variants show lower CAGR due to math vs selection vs bug?", "safe_withdrawal": null, "signalSeries": { "active": [ true ], "driver": "Signal driver", "entryValue": 0.0, "exitValue": 0.0, "inputs": [], "isDynamic": false, "isMulti": false, "labels": [ "2026-01-01" ], "megaThreshold": null, "megaThresholdLabel": null, "threshold": 0.0, "thresholdKind": "fixed", "thresholdLabel": "Entry 0.0", "trigger": [ true ], "triggers": [ "2026-01-01" ], "unit": "", "values": [ 0.126 ], "yLabel": "Indicator value" }, "signals": [], "slug": "qm-wtfm", "stats": { "base_cagr": 0.10641255900630053, "cagr": 0.0, "gap_vs_mechanical_log": -3.168, "lev2_cagr": 0.10323738162192364, "maxDD": 0.0, "mechanical_2x_cagr": 0.21775073046112792, "sharpe": 0.0, "sortino": 0.0, "window": "1991-01 to 2026-07 (385 months) controlled" }, "status": "completed", "status_chip": "Completed", "supplemental": { "bead": "qm-wtfm", "five_families": "F1-F5 documented in evidence" }, "tax_analysis": { "ltcg_pct": null, "note": "Experiment parity stub: no taxable wrapper; see strategy tax_analysis for methodology", "schema_id": "tax-analysis/v1", "stcg_pct": null }, "taxable_stats": null, "title": "Lower CAGR in leveraged meta strategies — investigation", "trading_costs": { "avg_daily_turnover": 0.00732, "avg_yearly_fees": 0.001845, "basis": "monthly_proxy", "cost_bps_one_way": 10.0, "fee_bps": 10.0, "n_bars": 1, "schema_id": "trading-costs/v1", "slippage_bps": 0.0, "total_costs": 7e-06 }, "verify": { "as_cagr": null, "as_stats": null, "cagr_delta": null, "gates": null, "last_verify_at": null, "monthly_bias": null, "monthly_corr": null, "monthly_mae": null, "n_months": null, "ours_cagr": 0.0, "ours_stats": { "ann_return": 0.0, "sharpe": 0.0, "sortino": 0.0 }, "passed": null, "status": "not_applicable", "verify_status": null }, "visualizations": [ { "data": { "labels": [ "2026-01-01" ], "note": "Aligned window, both rebased to 1.0 at start for fair comparison" }, "title": "Equity vs S&P (aligned, rebased to 1.0)", "type": "equity_vs_spy" }, { "data": { "note": "Triggers when overlay takes over base strategy" }, "title": "Signal timeline (sparse triggers)", "type": "signal_timeline" } ] }, "generated_at": "2026-09-17T04:12:25Z", "id": "qm-wtfm", "schema_id": "experiments/v1", "schema_version": 1 }