Frequently Asked Questions

A daily research service. Every trading day, machine-learning models estimate the probability that each of about 300 stocks and ETFs will close higher five trading days later. Those names are then narrowed through four gates into a short list of Four-Gate selections: the ML probability, a fundamental screen, current-event analysis and final human review. Company-specific fundamental criteria do not apply to ETFs.

Swing traders who want probability-ranked selections, and cash-secured put and Wheel traders who want a stock-specific 90% CSP/Wheel level for each selection.

Every trading day. The models run after the close. The Full ML Table and ML + Fundamentals list are normally posted before the market opens.

The Four-Gate selections and Daily Pick follow the current-event analysis and final human review, so they may be published or updated after the open. Every table shows its posting time.

The Daily Pick: one current Four-Gate selection each trading day, with its 5-day ML probability, Four-Gate status and commentary. Sign up with an email and password; no credit card.

We begin with a universe of approximately 300 stocks and ETFs. Candidates must have a machine-learning probability of at least 75% and pass the applicable fundamental screen.

Two AI reviewers then examine the remaining candidates for relevant news, company actions, SEC filings and other current information, alongside the ML probability and analyst opinions. A candidate moves on only when both reviewers agree that current information supports a higher price over the next 30 days.

James Krider personally examines that shortlist and chooses one selection that he considers a useful illustration of the system, while avoiding excessive repetition. The free Daily Pick is not necessarily the highest-probability stock. PTP Pro members see every Four-Gate selection.

No. The published probability remains the original machine-learning estimate of a higher close five trading days later.

The probability threshold is one gate. Fundamentals, current-event analysis and human review determine which candidates move forward; they do not recalculate or increase that probability.

One plan with everything: the Full ML Table, the ML + Fundamentals list, and all daily Four-Gate selections with their 90% CSP/Wheel levels. The founding-member rate is $39/month, locked in for as long as you stay subscribed. There are no other tiers.

Cancel anytime: click your account icon, choose Manage billing, and cancel in the billing portal. Your access continues to the end of the current monthly billing period. Payments are non-refundable except where required by law.

The model estimates a 78% chance that the stock will close higher five trading days later. It does not mean the stock will gain 78%.

If forecasts at this level are well calibrated, approximately 78 out of 100 comparable forecasts would close higher and 22 would not. That describes a pattern across many forecasts, not a promise about an individual stock.

The historical statistics shown in the tables come from testing during model development, on held-out data: data that came after the training data and was not used to build the model. They describe historical forecasts in the relevant probability band and are not a record of returns earned by subscribers.

Testing the underlying ML forecasts, evaluating the final Four-Gate selections and measuring returns from a defined trading strategy are separate questions. Historical model statistics alone do not establish the performance of the complete selection process or a trading strategy after costs.

Results on newly published forecasts are being collected.

These columns summarize historical test observations in the same probability band. They do not describe that individual stock's own trading history or a subscriber's account performance. Max Drawdown refers to the five-day holding period.

The daily ML probability, morning discount and 90% CSP/Wheel level are stock-specific outputs.

It is a stock-specific reference level derived from historical price behavior for the applicable Friday window. The published level uses that morning's opening price and remains a fixed morning snapshot. Selections posted Monday through Wednesday use that week's Friday; Thursday and Friday selections use the following Friday.

In roughly nine out of ten comparable historical windows, the stock closed above the calculated level at the end of the window; in roughly one out of ten, it closed below. This does not mean the stock stayed above the level throughout the period.

The calculator lets members apply the posted discount to a later price. That calculation does not change the published morning snapshot.

The level is a historical measurement, not a guarantee or a recommendation to sell a put at a particular strike. See the Method page for expiration-window details.

No. PTP publishes probabilities, selections and reference levels. It does not give personalized trade instructions or investment advice. Whether and how anything fits your account is your decision.

James Krider, MD, a physician and quantitative trader. He built and tested every model, reviews every Four-Gate selection before publication, and trades the system daily.

Use the Contact page, or reply to any PTP email. Questions are answered by James Krider, MD.