🦅 Garuda Alpha · Flagship Strategy
📊 HPQuant Dashboard ← Home
⭐ Flagship Strategy

Garuda Alpha — Live State

Cross-sectional momentum on IDX. Survivorship-safe, walk-forward validated, with a Bayesian regime-filter overlay. Data as-of 2026-08-04, refreshed nightly.

Total Return
+299%
JCI -2%
CAGR
+17.9%
2018–2026
Sharpe
0.94
JCI -0.26
Max DD
-15.4%
JCI -41.1%
Profit Factor
1.90
win 42%
Gates
5/7
§3.7 robustness

Equity Curve

Cumulative growth of Rp 1 vs JCI buy & hold

1x2x4x6x Garuda Alpha 3.99x JCI Buy & Hold 0.97x 2018-01-01 2026-06-02

Current Regime · Bayesian Filter

As-of 2026-08-04

RISK
OFF

RISK OFF — belief-weighted

Composite macro score -2 (range −6 to +7), filtered through a sticky Bayesian belief over four regimes. Positioned for defensive posture at next rebalance.

D
O

belief L/N/D/O = 0.00 / 0.00 / 0.29 / 0.71  ·  prior hard map: DEFENSIVE @ 0.70×

Gross Exposure
0.49×
EXPECTED · BRF

Active Picks

Top-12 by Composite Score (currently investable universe)

Screening recomputed nightly on latest prices (as-of 2026-08-04) — portfolio execution only at quarterly rebalance: last 2026-07-01 · next 2026-10-01 (first trading day of quarter). Names sized to the prevailing gross-exposure multiplier. Sector caps and single-name caps enforced at construction.

#TickerCompositeMomentumTrendQualityLowVolADTV
1MARK91.198.278.394.586.1Rp 20.7 Bn
2TAPG84.094.555.898.249.1Rp 24.2 Bn
3ADRO80.791.575.871.272.7Rp 78.3 Bn
4TINS76.586.178.362.021.8Rp 153.4 Bn
5AADI75.893.953.374.249.7Rp 75.9 Bn
6MAPI70.999.778.325.241.8Rp 63.4 Bn
7ITMG68.692.753.351.568.5Rp 29.7 Bn
8AKRA66.090.358.341.159.4Rp 30.4 Bn
9ELSA65.390.955.840.530.3Rp 22.4 Bn
10GGRM63.697.678.33.746.1Rp 26.6 Bn
11EMAS63.583.60.0100.027.3Rp 109.9 Bn
12RAJA55.961.860.047.950.0Rp 176.6 Bn
Note on Trend: Trend factor is binary-scaled (0 if price below 200-day EMA, 50+ with persistence if above). In the current macro regime, most names fall below their long-term moving average — trend=0.0 is therefore the expected reading, not a missing value. The composite still ranks names via momentum/quality/lowvol differentials.

Validation

Walk-Forward Robustness (7 windows, expanding in-sample)

Canonical factor weighting wins every in-sample window. Mean out-of-sample Sharpe: +0.81. Refit per window does not beat the fixed canonical — strong evidence the strategy is not curve-fit.

WindowOOS PeriodIS WinnerIS SharpeOOS Sharpe
WF12019-01-01 → 2019-12-31canonical+0.78-0.31
WF22020-01-01 → 2020-12-31canonical+0.40+1.70
WF32021-01-01 → 2021-12-31canonical+0.76+1.30
WF42022-01-01 → 2022-12-31canonical+0.89+1.25
WF52023-01-01 → 2023-12-31canonical+0.94-0.58
WF62024-01-01 → 2024-12-31canonical+0.77+0.13
WF72025-01-01 → 2026-06-02canonical+0.69+2.20

Documentation

Full Methodology & Operations

Deep-dive thesis documents and operational playbook. All data-driven, regenerated from the live backtest.

Generated 2026-08-04 19:27 WIB · data as-of 2026-08-04 · ↩ Back to HPQuant Dashboard