Model classification: Below Safe Price
Today's completed-day market price is at or below our calibrated Safe Price, and the company's filings qualify.
Methodology
We build an Estimated Fair Value for a company from its own SEC filings, then publish a single Safe Price that has stayed below the company’s later fair value in 94.91% of out-of-sample tests. No analysts. No price targets. No guessing.
A business is worth the cash you can pull out of it over its life, discounted for the fact that you could have earned something else with the money. Growth only counts if the company has demonstrated it can reinvest earnings at a decent return. You never pay full price for a business — you wait for a margin of safety.
Three ideas Buffett and Munger have talked about publicly for fifty years. We took them seriously and built an engine around them.
Every input is a public, dated, verifiable source:
We do NOT use:
For every company we cover, the page shows six things and nothing else:
Every classification renders under the fixed prefix Model classification: — the frame that says this is an output, not personal advice.
Model classification: Below Safe Price
Today's completed-day market price is at or below our calibrated Safe Price, and the company's filings qualify.
Model classification: Undervalued, Above Safe Price
Price is below our Estimated Fair Value but above our Safe Price.
Model classification: Above Fair Value
Today's completed-day market price is at or above our Estimated Fair Value.
Model classification: No Rating — Insufficient Filings
The company's filings do not show enough consistent history to make a classification.
Model classification: No Rating — Industry Not Covered
This industry requires specialized filings we don't yet include.
Model classification: No Rating — Data Unavailable
We could not build a complete valuation for this ticker today.
Model classification: No Rating — Price Unavailable
We could not read a completed-day market price for this ticker.
One additional classification, “Thesis Intact,” renders only inside the iOS app when a user has personally saved a reading below the Safe Price and the model still shows price below Estimated Fair Value. It never appears on the public site or in tweets.
Before we ever published a Safe Price, we ran every historical version of every Safe Price against what the company's later filings said the fair value turned out to be. The rules of the test are strict, and they are the same for every company:
Four validation numbers we are willing to publish (exact wording, denominators, and caveats attached — as required by our public contract):
“Our Safe Price stayed below the company’s later fair value at the next SEC filing in 94.91% of cases.”
Denominator: 746 out of 786 observations across 96 issuers.
Caveat: This is Safe Price coverage against later filing-derived fair value. It is not a probability of stock profit and not exact fair-value accuracy.
“On the financials-qualified subset, the 90% Safe Price coverage rose to 95.91%.”
Denominator: 328 of 342.
Caveat: This is Safe Price coverage against later filing-derived fair value. It is not a probability of stock profit and not exact fair-value accuracy.
“For qualified safe-entry decisions, later filings confirmed the entry in 96.88% of cases.”
Denominator: 31 of 32.
Caveat: This is Safe Price coverage against later filing-derived fair value. It is not a probability of stock profit and not exact fair-value accuracy.
“In the calendar-year lockbox — where our calibration cannot see any outcome from the year it is scoring — qualified safe-entry support was 96.55% at the next filing and 96.15% at two filings out.”
Denominator: 28/29 and 25/26.
Caveat: This is Safe Price coverage against later filing-derived fair value. It is not a probability of stock profit and not exact fair-value accuracy.
We recheck this valuation when a new SEC filing lands, when market price crosses our Safe Price, or when the safe horizon expires — whichever comes first.
For companies covered in our frozen research universe the page renders instantly from a precomputed row. For a typed-in ticker outside that universe we run a fresh valuation against the latest filings.
One dollar number representing what the business is worth today based on that company’s own SEC filings. Derived per company — not from an industry-wide multiple or a fixed growth rate.
A conservative discount to the Estimated Fair Value. We publish four confidence tiers (80%, 85%, 90% default, 95%) that give the reader progressively more cushion. Higher confidence — a lower number — means a thicker margin of safety.
Across 786 out-of-sample historical valuations spanning 96 companies, our Safe Price stayed below the company’s later filing-derived fair value in 94.91% of cases. It is Safe Price coverage against later filing-derived fair value — not a probability of stock profit, and not exact fair-value accuracy.
No. Inputs are limited to the company’s own SEC filings (10-K, 10-Q, 8-K), U.S. Treasury yields, and completed-day U.S. equity closing prices. We do not use Wall Street analyst earnings estimates or price targets.
We recheck this valuation when a new SEC filing lands, when market price crosses our Safe Price, or when the safe horizon expires — whichever comes first.
The filings do not show enough consistent history to make a classification, the industry needs specialized filings we do not yet cover, we could not build a complete valuation, or we could not read a completed-day price. It is the engine telling you it does not know.
No. This is a filing-based estimate of what a business is worth. It is not a recommendation to buy or sell any security, and it is not personalized investment advice.
The app applies this model to every U.S. filer, on demand.
Get the app — $99.99/moBilled through the App Store. Cancel anytime.
Model output · Education, not advice. Full disclaimer. Last reviewed 2026-08-26.