What is Glitch in the Credit, and can I trust its stress index?

Short answer: Glitch in the Credit (glitchinthecredit.com) is a free daily 0 to 100 stress index for the AI data-center credit boom, made by one person, Adam. Every morning it grades 17 signals in six layers from news, filings and data feeds, and each reading links its source and date, so you can check it instead of taking it on trust. It is new: the daily history begins Oct 25, 2026, so there is no track record yet, and we make no claim about how accurate it will prove. On Oct 30, 2026 it read 48, Elevated. It reports what the readings show. It isn’t a forecast, and it isn’t financial advice.

Glitch in the Credit is its own site, at glitchinthecredit.com. It is not part of any other tracker with a similar name, such as Compute Terminal’s weekly AI-Infra Credit Stress Index. How it compares with other trackers.

Who makes it?

The site is run by one person, Adam. Only a first name is published here: no surname, employer or credentials, so there is no résumé to rely on. That is why every reading shows its source: check the evidence, not the author.

An AI model (Claude) re-judges each signal every morning from news, filings and the data feeds, and links the source of each reading (how it works). Those judgments can be wrong, which is why the sources are linked.

To report a mistake, see the end of how it works.

Also on: X.

What does it measure?

The AI buildout is paid for largely with borrowed money. Glitch in the Credit watches that debt in the order a squeeze tends to travel, in six layers. The order is our working model, not a law. Each signal is graded calm (0), watch (1) or flashing (2); a layer averages its signals; the index averages the layers on a 0 to 100 scale (Quiet 0 to 19, Stirring 20 to 39, Elevated 40 to 59, Warning 60 to 79, Panic 80 to 100).

The alarm that matters most is the chain reaction: collateral, weakest borrowers and funds flashing together. One flashing light isn’t news; the combination is the signal. Method in full: how it works. Plain-language answer: how to measure an AI bubble with credit.

How can I check it?

1. Follow a reading to its source. Every signal has its own page with the status, the date of the evidence, a link to the source and past readings. Start with any of them, for example GPU rental prices.
2. Check the history. Each daily check has a permanent page at the daily archive, so you can see what it said on a given day.
3. Get the data. The latest reading is one file, api/latest.json, and the datasets are JSON and CSV on the data page. The collateral inputs behind the GPU numbers list a source and date on every line.
4. Compare it with official data. Several readings rest on public series and filings: the OFR Financial Stress Index (US Treasury), the Federal Reserve Board’s excess bond premium, Fed bank lending data on FRED, and filings on SEC EDGAR. If our reading and those sources disagree, believe the sources.

What can’t it tell you?

• It doesn’t tell you when anything will happen. It isn’t a timing tool.
• Some readings are judgment calls based on public reporting, some of it secondhand or unconfirmed. Each links its source so you can weigh it.
• The index averages graded statuses. It isn’t a probability of anything.
• The daily history begins Oct 25, 2026, with six daily checks so far. We have not shown that it gives early warning, and we make no accuracy claim.
• It watches AI data-center credit. It doesn’t score the whole market or any single stock.
• Prices from other providers (stock closes, GPU rents) are shown with credit and are not ours.

What does it cost, and can I reuse it?

It is free, including the full reading. The daily email goes out at 8 a.m. ET (newsletter). Our index, layer scores, signal readings and archive are CC BY 4.0: credit “Glitch in the Credit” and link to the page you used. Third-party data keeps its owners’ terms. See the free API and open data guide.

Follow the debt. Not a timing tool. Not financial advice.

Sources: Glitch in the Credit daily check, Oct 30, 2026 (api/latest.json); How it works; OFR Financial Stress Index; SEC EDGAR.