How DividendGuard's Trap Classifier Works

A dividend cut rarely comes out of nowhere. By the time a company announces it, the warning signs were usually visible in the financials for months — sometimes years. The Trap Classifier exists to catch those signs early.

The High Yield Problem What It Checks Why It's Deterministic How It's Funded Data Sources What It Isn't

The problem with "high yield"

A high dividend yield can mean one of two things: a great value opportunity, or a company whose stock price has fallen because the market already expects a cut. Telling these apart from a yield number alone is impossible. It requires actually looking at the underlying business — and most free tools don't do that.

What the Trap Classifier checks

Rather than relying on a single metric, the Trap Classifier evaluates each company across several independent dimensions of dividend safety:

Leverage — how much debt the company carries relative to its earnings power, and whether that debt load is increasing.
Payout sustainability — whether the dividend is actually covered by free cash flow, not just accounting earnings (which can be manipulated by non-cash items).
Earnings quality — whether reported profits are backed by real cash generation, or inflated by one-time items, asset sales, or aggressive accounting.
Dividend growth trend — whether the pace of dividend increases is decelerating, which often precedes a freeze or a cut.
Historical context — how the company's current financial position compares to its own multi-year trend, not just a snapshot.

Each of these is evaluated independently and combined into a single safety classification, so no single weak metric is hidden by an otherwise strong one.

Why the score itself is deterministic

The Trap Classifier's safety score is generated by fixed, auditable logic — not by an AI model. Every score is the output of the same rules applied consistently to every company, every time. That means the same inputs always produce the same output: no randomness, no model drift, no black box. A score means the same thing today as it did six months ago.

We do use AI elsewhere in the product — for example, to surface relevant news context around a stock you're researching. But that layer is kept entirely separate from the classifier score itself, clearly labeled as a distinct signal, and never allowed to influence the underlying safety rating. The score you see is always the deterministic one.

How it's funded

DividendGuard is free to use. There's no subscription — the product is supported by minimal, non-intrusive advertising instead.

Where the data comes from

All financial data is sourced from SEC EDGAR filings — the same primary-source data public companies are legally required to disclose — not third-party estimates. Prices are sourced separately and cross-validated for accuracy.

What it isn't

Not investment advice. The Trap Classifier is a screening tool designed to flag risk early so you can investigate further — not to make buy or sell decisions for you.

Always do your own research before making investment decisions.