Ticker Nerd Review

How we rated Ticker Nerd

A membership here buys a portfolio rather than a piece of software: twenty US-listed stocks, chosen by rules, published beside the method that chose them. The factor families are named, the eligibility screen is printed in the order it is applied, the benchmark is stated, and the trade ledger includes the positions that lost money. The one thing a reader cannot check from outside is the record before the service began trading, and the page that prints the figures says which part of them was simulated. Wherever a page can be opened rather than a claim taken on trust, that page is the source used, and the list of them, each with the date it was read, closes the review.

Method and factor coverage holds the largest share of the five weights, on the reasoning that a rules-based service is worth what its rules look at, and that no answer to the other four questions rescues a model measuring the wrong things.

Method and factor coverage

One model scores 4,658 US-listed stocks every day across nine factor families, and the families are published: Revisions, Value, Momentum, Quality, Accruals, Issuance, Growth, Size and Volatility. Around fifty separate measurements feed them, none of them tuned against the result it is meant to predict. The investable pool is narrower than the scored one. Eligibility runs in a printed order: common shares with a primary US listing, standard sectors with real estate investment trusts excluded, a share price above $5, a median traded value above $10 million a day over the trailing quarter, positive trailing sales, no announced takeover, and at least a year of trading history. The 1,500 largest of the survivors make up the pool the portfolio draws from, which places size rank last rather than first.

What gives the coverage claim its weight is the list of things the model does not use. Price-to-book, market timing, analyst rating changes, insider buying, short-term reversal and seasonality are each named as rejected, with a reason beside them. A published rejection list is the harder disclosure of the two, since it shows which ideas were tried and dropped rather than only which ones survived, and the distinction is drawn finely here: analyst forecast changes are kept while analyst rating changes are not. One question is answered in two places. The screener page names Revisions, Value and Momentum as the families carrying the most weight at present, while the methodology describes the nine as sitting close to even, so the two pages give a reader different accounts of the same weighting.

4.6 for method and factor coverage. The nine families are named in full, the measurements behind them are counted, and the ideas the model rejected are published with their reasons, which is more than most services at this price say about what their model will not do.

Evidence and testing discipline

The history is split into three windows that were fixed in advance: 2005 to 2014 for building the model, 2015 to 2020 as a check on it, and 2021 to 2025 held back until the rules had been written down. Holding a window back is what turns a backtest into a test, because it is the only arrangement in which the person fitting the model cannot see the answers while fitting it. The data behind it is point-in-time Compustat and FactSet, which matters here: a factor study that reads restated company accounts is reading numbers nobody had at the time. The benchmark is the S&P 500 Equal Weight, chosen because it holds each member at the same size and so does not hand the model a free ride from a handful of very large companies. The S&P 500 is shown beside it, and the Nasdaq 100 is run as a check on whether the result is simply a growth index wearing a different name.

The discipline shows most clearly in what the page says about earlier attempts. Two strategies were built, carried to the held-back window, and failed there; neither was reworked afterwards. The rebuild that produced the current model is described as a post-development evaluation rather than a clean pass through a window nobody had seen, which is the accurate description of what it is. The returns are also regressed against the Fama-French five factors plus momentum, a test that asks how much of the result the known factors already explain, rather than being left as an assertion about the model’s originality.

4.3 for evidence and testing discipline. The windows were locked before the rules existed, the two strategies that failed the held-back test sit on the page with the admission that they were not revised, and the rebuild claims only what its own description supports.

The record and how it is reported

The figures are quoted against the equal-weight benchmark. Over the last twelve months the service reports 26.1% against 15.7%; for the year to date, 23.7% against 13.2%; annualised since 2005, 12.4% against 9.9%; and through 2022, a loss of 7.7% against a benchmark loss of 11.6%. The trade ledger carries 508 closed positions, and the names that lost money appear in it rather than being folded into a total.

The line that governs how all of it should be read sits on the same page: simulated before 20 July 2026, live since, with a note that backtested figures are hypothetical and were not traded. That single sentence decides what the long-run numbers are evidence of. The annualised figure reaching back to 2005 is the one a reader will weigh most heavily, and it is also drawn entirely from a simulation, so the period a subscriber can follow alongside the service is a matter of months. The disclaimer supplies the rest of the frame: the company is not a licensed adviser, the material is general and takes no account of anyone’s circumstances, subscribing creates no client relationship, and the owner’s own money is in the same twenty names members see.

3.8 for the record and how it is reported. The losses are published, the ledger is itemised and the label on the snapshot is unambiguous, while the headline track record runs back twenty years through a simulation and only months through traded money, which is the gap this score reflects.

What membership includes

Membership buys the Ticker Nerd 20: twenty stocks, equally weighted, long only, with no margin, no short positions, no hedging and no cash held back. The list is reviewed every four weeks, and a typical review replaces a name or two. Working from the public ranking is a different exercise from holding the portfolio, because the rank measures a company while the rules decide whether the portfolio owns it and when it lets it go.

The free side is real rather than a sample. Every stock page is open to anyone, and the Market Radar arrives on a Monday, computed off the Friday close and sent before the US market opens, reporting the largest rank changes across roughly 4,600 names with the factor behind each one. A reader can follow the ranking for nothing, and what the subscription adds is the twenty names plus the rules that select and replace them.

4.3 for what membership includes. The portfolio rules are published in full and the review cycle is stated plainly, while the reader who wants to see which twenty names are held today has to subscribe before finding out.

Price and terms

The membership costs $39 a month or $199 a year, charged through Stripe, with a reminder email sent a week before each renewal. The annual rate is framed on the page as under $4 a week and more than half off the monthly equivalent, which comes to $468 across twelve months. There is no free trial and no free slice of the portfolio; the thirty-day window is the trial.

That window runs from the original date of purchase and covers the full amount. Beyond it, the terms describe purchases as final unless the law requires otherwise. Stopping the subscription can be done from the account settings or by writing to support, and it takes effect at the end of the billing period already paid for rather than at the moment it is requested. The licence is personal and non-commercial, cannot be transferred to anyone else, and the terms name the State of Victoria in Australia as the law governing the agreement. Liability is limited so that indirect and consequential damages are excluded.

4.3 for price and terms. Thirty days is long enough to watch a full review cycle turn over and judge the method rather than a first impression, and the renewal reminder arrives before the charge rather than after it, while a subscription closed part-way through a paid period is not refunded for the remainder.

Who Ticker Nerd is for

This suits a reader who wants a process rather than a tip, and who is content to hold twenty names that will not be the ones leading a finance site’s front page. The published method is the reason to subscribe, and the reader the service is built for is the one who intends to follow the rules through a poor year, since the drawdown is part of what the method costs rather than a fault in it.

A reader who wants recommendations shaped around their own situation, or who wants a person to tell them when to sell, will not find either here: what is published is impersonal, the firm is not a licensed adviser, and the exits are decided by the rules rather than by anyone reading an individual’s circumstances.