What Lighthouse is
Lighthouse is an educational macro-signal platform. It tracks the most important indicators in markets — credit spreads, liquidity flows, yield curves, regime classifications — and explains what they mean in plain language.
It exists for one reason: most people who want to understand markets are stuck choosing between sources that are either too technical or too sensational. Bloomberg Terminal costs £20,000 a year. Macro Twitter is mostly noise. Mainstream financial media is mostly performance. There's a missing middle: rigorous data, honest interpretation, no hype, no tribal politics, no "altseason in 30 days" predictions.
That missing middle is what Lighthouse is trying to be.
What Lighthouse is not
This part matters more than the previous part.
Lighthouse is not a trading tool. It will not tell you what to buy or sell. The data we track describes regimes and conditions — it does not predict short-term price movements, and anyone who claims otherwise is selling something.
Lighthouse is not financial advice. Nothing on this site is personalised to your situation, your goals, or your risk tolerance. If you need financial advice, talk to a qualified, regulated financial adviser. The author is not one.
Lighthouse is not a paid signal service or trading bot. It tracks publicly available macro data and explains it. That's the entire product.
Lighthouse is not trying to be Bloomberg, Refinitiv, or any institutional terminal. Those tools have their place. Lighthouse is for the curious individual who wants to understand the macro environment without needing a finance degree or a six-figure data subscription.
Who built it
Lighthouse is built by Graham Kennedy, the founder of Draxiq — a small AI-and-web agency based in Southport, UK.
Graham is not a finance professional. He is not a CFA, not a hedge fund manager, not an institutional trader. He's a self-taught builder who got interested in macro through years of trying to understand what was happening in markets — and got frustrated that most "educational" content in this space was either gated behind expensive subscriptions or watered down to the point of uselessness.
Lighthouse is being built in public, deliberately. The build process — including mistakes, course-corrections, and design decisions — will be documented as the platform grows. This isn't a polished product launched with marketing money behind it. It's a working tool, shipped iteratively, by someone learning alongside the people using it.
If that approach makes you trust Lighthouse less, that's fine — there are plenty of macro newsletters with credentialed authors and slick brand identities you can subscribe to instead. If it makes you trust Lighthouse more, you're probably in the right place.
Where the data comes from
Every data point on Lighthouse traces back to a primary source. No exceptions.
Federal Reserve Economic Data (FRED)
The bulk of our quantitative data comes from FRED, operated by the Federal Reserve Bank of St Louis. FRED is the gold standard for free, citation-quality economic data. It's what every academic paper, central bank research note, and institutional research desk uses as a reference. Source URL: https://fred.stlouisfed.org
We track 16 FRED series across four categories:
Credit spreads:
- ICE BofA US High Yield OAS (BAMLH0A0HYM2)
- ICE BofA HY BB / Single-B / CCC tier spreads (BAMLH0A1HYBB / BAMLH0A2HYB / BAMLH0A3HYC)
- ICE BofA US Investment Grade OAS (BAMLC0A0CM)
- Moody's Baa minus 10-Year Treasury (BAA10Y) — long-history, daily since 1986
- Moody's Aaa minus 10-Year Treasury (AAA10Y) — daily since 1983
Treasury yields:
- 10-Year Treasury (DGS10) — daily since 1996
- 2-Year Treasury (DGS2) — daily since 1996
- 30-Year Treasury (DGS30) — daily since 1996
- 10-Year Treasury monthly (GS10) — monthly since 1953 for long-history depth
- 2-Year Treasury monthly (GS2) — monthly since 1976
- Federal Funds Rate (FEDFUNDS)
Liquidity:
- Fed Total Assets (WALCL) — the Fed balance sheet
- Treasury General Account (WTREGEN) — the US Treasury's checking account at the Fed
- Reverse Repo balances (RRPONTSYD) — money parked at the Fed by money market funds
- Bank Reserves (WRESBAL)
Market context:
- Trade-Weighted US Dollar Index (DTWEXBGS)
- VIX volatility index (VIXCLS)
- S&P 500 (SP500)
Pulled, stored, refreshed
Data is pulled once daily at 7am UK time via the FRED API and stored in a local SQLite database. We also did a one-time historical backfill from FRED's CSV downloads to give us deep history where available. As of late April 2026, the database contains 71,580 daily observations across all 16 series.
A note on the ICE BofA series
The ICE BofA OAS series we track (high yield, investment grade, tier breakdowns) are licensed indices that FRED limits to a 3-year rolling window across both API and direct CSV access. This means our high-frequency credit spread data goes back to April 2023.
For longer historical context, we use the Moody's Baa10Y and Aaa10Y series — the same credit spread proxies finance academia has used for over 50 years. These give us 39 years of daily data going back to 1986. When you see references to historical episodes (1987 Black Monday, 1998 LTCM, 2008 GFC, 2020 COVID), the values come from this series, queried directly from our database.
This is documented openly because transparency on data sources is more important than pretending we have data we don't.
Editorial sources for analysis and articles
The interpretation layer — articles, glossary explanations, regime framing — draws on a curated library of named-author sources. Our knowledge base currently contains 56 catalogued sources from these authors:
Howard Marks (Oaktree Capital co-chairman) — 8 memos covering Sea Change, Easy Money, The Impact of Debt, Gimme Credit, On Bubble Watch, and others. URL: https://www.oaktreecapital.com/insights/memos
Lyn Alden (independent macro analyst, founder of Lyn Alden Investment Strategy) — 8 articles covering fiscal/monetary frameworks, QE/MMT primers, Treasury bond risks, and Bitcoin liquidity dynamics. URL: https://www.lynalden.com
Ray Dalio (founder, Bridgewater Associates) — 24 catalogued resources including books (Big Debt Crises, How Countries Go Broke, Principles for Dealing with the Changing World Order), LinkedIn essays, animated explainers, and PDFs from economicprinciples.org. URLs: https://www.economicprinciples.org and https://www.principles.com
Sources are tiered by editorial quality: Tier 1 sources can be cited as primary evidence for empirical historical claims. Tier 2 sources are cited as framework or perspective ("Marks argues..." or "Dalio's framework suggests..."). Tier 3 sources are catalogued but used selectively, with empirical claims separated from advocacy.
How articles are built
Every factual claim on Lighthouse traces to a primary source, with the citation visible at the bottom of every article. This isn't ornamental — it's a discipline that catches errors before they ship.
When we write that "the Moody's Baa10Y credit spread peaked at 6.16% on 4 December 2008," that claim has a fact ID, a verifier method (the SQL query that produced the number), and a last-verified date. If the data later updates or the methodology changes, every article using that fact updates automatically on next render.
The technical architecture, in plain terms:
- Facts live in JSON files with full provenance
- Articles cite facts by ID, not by re-typing the claim
- A renderer reads the article Markdown, expands each citation into its claim text plus a numbered footnote, and produces the final HTML
- Time-sensitive values (like "today's spread reading") are pulled live from the database at render time
This means we can't accidentally drift from the data. Every claim is anchored.
Honest limitations
Some things Lighthouse cannot do, and won't pretend to:
Predict short-term moves. Macro signals describe regimes — calm versus stressed versus crisis. They don't predict whether the S&P 500 closes up or down tomorrow. Anyone using regime data for short-term trading is misusing the tool.
Replace personalised advice. A 30-year-old with no debt and a 35-year-old with a young family and a mortgage have different needs. Lighthouse describes the macro environment. It doesn't decide whether the macro environment is good or bad for your specific situation.
Catch every regime change in real time. Credit spreads are leading indicators, but they're not perfect. False alarms exist. Some real warnings are missed. The tool is meant to make you better-informed, not infallible.
Cover all asset classes equally. The current focus is US macro, credit, and rates. We track equity volatility (VIX) and the dollar (DXY) for context, but we do not track individual stocks, crypto fundamentals, real estate, commodities, or international markets in any depth. That may change. For now, US macro is the focus.
Replace doing your own work. Lighthouse compresses a lot of complex data into readable form. That compression is valuable, but it's also imperfect. Any decision that involves real money should involve your own additional research.
Editorial principles
Six principles guide what gets shipped:
Citations or it didn't happen. Every numeric claim links to a primary source.
Beginner-first writing. Most macro content assumes the reader already understands the jargon. Lighthouse assumes you don't, and explains as we go.
Anti-hype. No "this could be the next Lehman" framing. No countdown clocks. No urgency-marketing. The data is interesting enough on its own.
Honest about uncertainty. When we don't know something, we say so. When data is stale, missing, or contested, we flag it.
Transparent about methodology. This page exists because a tool that doesn't show its working can't be trusted.
No tribal politics. Macro analysis has been politically poisoned for years. Inflation isn't left or right. Liquidity isn't a partisan question. We try hard to keep tribal framings out of the analysis.
Roadmap
What's live now:
- This about page
- The /lighthouse landing page
- The first foundation article: Credit Spreads — The Smoke Detector
- A working data collector pulling FRED daily
- A knowledge base of 56 sources, 32 facts, and 23 glossary terms
What's coming, in rough order:
- A polished dashboard showing all live signals with regime classification
- Eight more foundation articles covering bond yields, the yield curve, liquidity mechanics, regime change, and the major historical episodes in detail
- An AI assistant trained on the knowledge base and live data
- Weekly market briefings via email
- Telegram alerts for regime changes
- Quizzes to reinforce the foundation curriculum
Lighthouse is being built incrementally. Each piece ships when it's actually good — not on a marketing timeline.
Contact
Questions, corrections, source suggestions, or feedback are all welcome. Reach Graham at hello@draxiq.com.
If you spot a factual error in any article, please flag it. The fact-citation system is designed to make corrections easy: tell us which fact ID is wrong, what should be there instead, and what your source is. We'd rather fix it than be right by accident.
Last updated: 30 April 2026. This page will be updated as Lighthouse evolves.