Regime-Lookout — MRI Market Interpretation Engine

Mobile-first risk-aware market interpretation system based on the MRI framework. 모바일 중심 MRI(Market Regime Interpreter) 기반 시장 해석 시스템.

Overview | 개요

Regime-Lookout is not a prediction model and does not provide trading signals. It interprets macro-financial conditions to describe the current market regime.

• Regime-Lookout은 예측 모델이 아니며 종목 추천 시스템도 아닙니다.
• 현재 매크로 환경을 분석해 지금 시장 상태를 해석합니다.

The goal is behavioral stability through structured interpretation of macro signals. (매크로 신호를 구조화하여 감정적 투자 판단을 줄이는 것이 목표입니다.)

The developer of this project is not formally trained in software engineering or economics. This system was developed for personal use with the assistance of AI tools. (본 프로젝트의 개발자는 소프트웨어 공학이나 경제학을 전공하지 않았으며, 개인적인 사용 목적을 위해 AI 도구의 도움을 받아 개발되었습니다.)

---

Core Principles | 핵심 철학

1. Interpretation over Prediction
   • 현재 시장 진단 중심.
2. Risk-Adjusted Confidence
   • 신호 충돌 시 자동 신뢰도 감소.
3. Behavior-First Design
   • 거래 신호가 아니라 의사결정 안정화 도구.

---

Methodology | 분석 방법

Feature Standardization | 변수 표준화

Public Yahoo Finance data are standardized using rolling z-scores.

• x = z( ln(QQQM / XLP) )
• y = z( VOO 20d return )
• rates = z( TNX change )
• usd = z( UUP change )
• vix = z( VIX change )
• goldFear = z(GLD) − 0.5*z(UUP)

✔ Gold-USD decoupling 적용 ✔ 달러 영향 제거 후 안전자산 신호 추출

Regime Classification | 시장 레짐 분류

Distance-based probabilistic matching to predefined prototypes:

• Risk-On
• Panic
• Stagflation
• Goldilocks
• Defensive
• Liquidity Crisis

Single prediction is not used. (확률 분포 형태로 출력됩니다.)

Confidence Engine | 신뢰도 엔진

Confidence is reduced when cross-asset signals conflict. Examples of penalty triggers:

• Risk-On + rising VIX
• Goldilocks + weak gold response
• Correlation surge across equities
• Liquidity stress + USD spike
• Divergent safe-asset behavior

Multiple triggers → larger penalty. (이벤트를 직접 탐지하지 않고도 지표 간 불일치로 신뢰도 조정.)

---

Output | 결과

• Today Score (0–100)
• Regime probability distribution
• Reasoning tags
• Confidence level
• 현재 시장 상태 설명만 제공.

---

Mobile-First Architecture | 모바일 중심 구조

Regime-Lookout is designed as a serverless progressive web app.

• GitHub Actions → data/latest.json update
• GitHub Pages → Web App
• Mobile Browser → PWA install

✔ 개인 서버 불필요 ✔ 자동 데이터 업데이트 ✔ 모바일 중심 UI

Tech Stack

• Frontend: React + Vite + PWA, IndexedDB snapshot caching, Swipe navigation without routing
• Data Pipeline: Yahoo Finance (yfinance), GitHub Actions scheduled updates, JSON snapshots
• Optional Backend: FastAPI (local testing only)

---

Why This Matters | 왜 중요한가

Investors face indicator overload and conflicting news. Regime-Lookout compresses macro complexity into a clear regime signal.

It helps users:

• Pause when risk rises
• Avoid panic selling
• Avoid reckless FOMO

(행동 안정화 도구로 설계되었습니다.)

---

Data Source and Disclaimer

• Data source: Yahoo Finance via yfinance.
• Data may be delayed or incomplete.
• Regime-Lookout is an analytical interpretation tool, not financial advice.

Language Support | 언어 지원

Language is automatically detected on first visit based on browser settings (ko/en), and user preference is stored locally for future sessions. (첫 방문 시 브라우저 언어를 감지하여 기본 언어를 설정하고, 이후에는 사용자 선택을 로컬에 저장하여 유지합니다.)

Confidence Clarification | 신뢰도 설명

Confidence represents overall interpretation reliability, not scenario-specific confidence. During market hours, the intraday interpretation is compared against the 60D daily distribution; divergence reduces confidence (consistency penalty). End-of-day latency alone does not automatically reduce confidence; freshness is evaluated using actual data update timestamps.

Project Status

Active development. Current focus:

• Mobile PWA stability
• Confidence calibration
• GitHub Actions data update
• UI simplification

Data Source & Usage Notice | 데이터 출처 및 사용 고지

This project fetches market data via the public Yahoo Finance endpoints (commonly accessed through the yfinance library) and publishes derived, lightweight JSON snapshots for the app UI.

• Data Source: Yahoo Finance (via yfinance)
• Intended Use: personal / educational / research use
• No guarantees: data may be delayed, incomplete, or inaccurate

Important: If you fork or deploy this project, you are responsible for complying with Yahoo Finance’s terms and any applicable local regulations.
Do not use this repository to resell, redistribute at scale, or offer paid data/API services based on Yahoo Finance data.

---

Disclaimer | 면책 조항

• This software is provided as-is without warranty of any kind.
• Nothing in this repository constitutes financial, investment, or trading advice.
• Use at your own risk.