@ADZO
Macro investing, decoded
ADZO Research · Live track record

Track Record

Every regime change logged. Each call scored against benchmark return. Methodology before data — measurement is defined, not chosen after.

Statistical honesty
Track record begins 2026-05-11 (framework v4.0 inception). As of 2026-05-12, returns columns are empty — 30-day windows close at the earliest June 11. A claim of framework hit-rate requires 6+ data points per label or 20+ overall. That's a 6-month build, not 8 weeks. Aggregate hit-rate stats are deliberately not displayed until the threshold is met.

Measurement methodology

Defined before the data starts populating. Once locked, the methodology must not be retroactively changed to make calls look better.

What each label predicts

Constructive
Primary instrument expected to OUTPERFORM its regional benchmark over 90–180 days from the call date.
Mixed
Primary instrument expected to perform IN LINE with its regional benchmark; no directional claim. Hit if absolute relative return is within ±2pp.
Stressed
Primary instrument expected to UNDERPERFORM its regional benchmark over 90–180 days from the call date.
Breaking
Primary instrument expected to MATERIALLY UNDERPERFORM, with downside risk of 10%+ over 90–180 days.

Benchmark mapping per regime

RegimeInstrumentBenchmarkRationale
🇰🇷 Korea — MemoryEWYAAXJ (MSCI EM Asia)Single-country EM Asia exposure
🇹🇼 Taiwan — AI/logicEWTAAXJ (MSCI EM Asia)Single-country EM Asia exposure
🇨🇳 China — State recoveryMCHIEEM (MSCI EM)China is heavyweight of EM; ex-China benchmarks distort
🇮🇳 India — Domestic capitalINDAAAXJ (MSCI EM Asia)Single-country EM Asia exposure
🇯🇵 Japan — ReflationEWJIEFA (MSCI EAFE)Japan is developed-market; benchmark vs DM ex-US, not EM
🌏 ASEAN ex-MYASEAAAXJ (MSCI EM Asia)ASEAN composite vs broader EM Asia
🇲🇾 Malaysia — Data centresEWMAAXJ (MSCI EM Asia)Single-country EM Asia exposure
💰 Asia bonds — CarryKHYBHYG (US HY)Asia HY vs US HY captures the carry spread directly

Scoring rule

For each regime change-log entry, compute (instrument total return) − (benchmark total return) over each of the 30 / 90 / 180-day windows from the call date.

A call is HIT if the relative-return sign matches the label's directional claim. Mixed-label calls are HIT if the absolute relative-return is within ±2 percentage points (i.e. performed roughly in line, as predicted).

180-day is the primary measurement window. Regime changes operate on quarters-to-years; 30-day returns are mostly noise and should not be used as a framework-quality claim. The 30-day column exists to show whether the early read is directionally consistent, not to claim predictive power.

Statistical threshold for any hit-rate claim

Hit-rate aggregations require 6+ data points per label or 20+ overall before being reported as a track-record claim. Below that, the table below displays individual call returns without aggregating a hit rate. A small sample is not a track record.

Live calls (accumulating)

Every regime change since framework inception. +30d, +90d, +180d columns populate as the windows close. Empty cells are pending, not failed.

Call date Regime Label Instrument Trigger +30d rel. +90d rel. +180d rel.
2026-05-11 🇰🇷 Korea constructive · MEDIUM EWY vs AAXJ Memory cycle inflection + Korea exports +4.2% YoY pending (June 10) pending (Aug 9) pending (Nov 7)
2026-05-11 🇹🇼 Taiwan mixed · MEDIUM EWT vs AAXJ TSMC Q1 +13% YoY + ~$700B hyperscaler capex; valuation re-classified post v4.1 pending pending pending
2026-05-11 🇨🇳 China mixed · MEDIUM MCHI vs EEM Caixin 52.2 + NBS 50.3 expansion; property drag persists pending pending pending
2026-05-11 🇮🇳 India stressed · LOW INDA vs AAXJ FII outflows USD 22B YTD at 2008-pace; USD/INR record 95.26 pending pending pending
2026-05-11 🇯🇵 Japan constructive · MEDIUM EWJ vs IEFA Nikkei 62,833 ATH + Tankan Q1 17 + Shunto 5.3%; JGB 10y 1997-high pending pending pending
2026-05-11 🌏 ASEAN ex-MY stressed · LOW ASEA vs AAXJ Vietnam FDI sustained but Indonesia rupiah weak + Thailand political risk + Philippines PMI contracting pending pending pending
2026-05-11 🇲🇾 Malaysia constructive · HIGH EWM vs AAXJ 4-leg alignment: MYR 3.92 + PMI 51.6 4y high + foreign bond inflows MYR 16.5B + Q1 GDP 5.3% pending pending pending
2026-05-11 💰 Asia bonds mixed · LOW KHYB vs HYG v4.1 reclass per extended rule table (1 ✓ · 3 ◐); BNM stable + BOJ measured QT + RBI cut bias pending pending pending
Aggregate hit rate Withheld. Statistical threshold for a hit-rate claim is 6+ data points per label or 20+ overall. Current sample: 0 closed calls. Returns will populate as windows close (first 30-day window closes ~June 11, 2026). The framework will not report aggregate hit rates until the statistical threshold is met.

Realistic timeline

When does this page become an actual track record? Honest schedule below.

Week 0v4.2 scaffolding shipped (2026-05-12). 8 inception calls logged. Returns columns empty.
Week 44 weekly refreshes completed. First regime change log entries with real (post-inception) signals. Returns still empty.
Week 6First +30d returns populate for 2026-05-11 inception entries. Early read; mostly noise.
Week 14First +90d returns populate. First real signal of framework hit-rate per regime.
Week 26 (6 months)First +180d returns populate. Statistically meaningful sample begins to form. Possible peer-review responses returned by now.
Week 26+Aggregate hit-rate claim possible once 20+ closed calls accumulate (likely Q4 2026 / Q1 2027).

Audit grade moves from A− to A around Week 12-16 if the discipline holds (cadence met, peer reviews back, track record populating). Not Week 8 — that was the original under-estimate.