What we publish

Meadow Investment Journal publishes free editorial and informational articles on one subject: how artificial intelligence is used to monitor the long-term allocation of assets. We write for readers who hold portfolios for years — pension reserves, endowment-style pools and patient households — and who want a clear, sourced account of what the models actually do.

The journal is editorial only. We do not sell subscriptions, paid articles, signals, tools, seats, deposits or execution. There is no premium tier and no waiting list.

Three monitoring threads

Every article we publish returns to one of three threads. They are narrow on purpose, because long-term allocation monitoring is a narrower discipline than most AI-in-finance coverage suggests.

1. Drift from policy weights

A long-term allocation is a set of target weights chosen to fit an owner’s horizon and risk. Markets move those weights away from policy. We cover the AI methods that measure drift, score it against cost and tax friction, and flag when a rebalance is worth doing rather than reflexive.

2. Regime and correlation shifts

The diversification an allocation relies on can break when long-run correlations and volatility regimes change. We follow the models that detect those shifts early, because a portfolio that stops diversifying has stopped doing its job even if its weights look unchanged.

3. Horizon and liability matching

The monitoring question changes as the horizon does. We follow the research on horizon-aware monitoring — how models test whether the assets still cover the liabilities and the spending a pool was built to fund, and how that test is refreshed as years come off the horizon and the window shortens.

What we do not do

We do not give personal investment advice. Nothing here is tailored to your situation.
We do not connect to your accounts, ingest your holdings or run a model on your portfolio.
We do not publish trading signals, price targets or “best AI stock” lists.
We do not accept payment for placement, coverage or a favourable reading of any research.

How an article is built

Each article starts from a published source — a peer-reviewed paper, a regulator note or a named practitioner document. We summarise the method in plain language, mark where the authors themselves flag limits, and link the source so you can read it yourself. Where a claim is contested in the literature, we say so rather than picking a side.

Read with us

If a thread above is the one you have been trying to follow, the journal is free to read and updated as the research moves. If you want to ask about a specific article or suggest a source, send an inquiry and we will reply by email.