We Spent Years Building Leading Indicators. Now You Can Trade Them on Autopilot.

By AltIndex Research · 9 min read · June 9, 5:33 pm

The big institutions have always had an edge: teams of analysts, satellite imagery, credit-card panels, and data feeds that cost more per year than most people invest in a lifetime. Alternative data, the digital footprint a company leaves before it reports earnings, is one of the few places a regular investor can level that field. A hiring spree shows up on LinkedIn before it shows up in revenue. A product catches on in app stores and web traffic before the market reprices the stock. A small-cap starts getting talked about on Reddit before the funds notice. We have spent years turning those signals into one number per stock, the AltIndex AI Score, and the only way to act on it was to log in, run the screens, and place every trade by hand.

That changes today. We have taken three of our strategies, built entirely on those signals, and made them available on Autopilot, the trading app. You follow a portfolio, the positions copy into your own brokerage, and it rebalances every month as the signals move. The screening, ranking, and rebalancing happen for you. Your money never leaves your own account, and you can stop following at any time.

Here is what each portfolio is built to capture, how it has performed, and the risks that come with it.

Portfolio #1: AltIndex AI Top 10

This is the broadest expression of the model. It holds the 10 stocks with the highest AltIndex AI Score at each rebalance, equal-weighted. The AI Score blends four families of signal into one number: social audience growth, hiring momentum, customer traction from web traffic and app downloads, and fundamentals like revenue growth, insider buying, and analyst targets. A stock rises to the top only when several of those are pulling in the same direction, so the portfolio ends up holding the companies firing on all cylinders at once, with a six-month-plus horizon in mind.

Over the past 24 months, the strategy returned a hypothetical 30.2% over six months, 72.6% over twelve, and 58.1% over twenty-four, against 47.5% for the S&P 500 and 45.5% for the Russell 2000 over the full window. It beat both benchmarks across all three periods in the backtest. It also fell about 14% in early 2025 when the whole market sold off, then recovered within a few months. These are backtested, hypothetical figures, not the results of a real account.

The live test, which began on May 4, 2026, returned a simulated 17.47% through June 5 against the S&P's 2.43%. One month is far too short to read anything into, and the number is simulated rather than realized, but it is the only forward-looking data we have so far.

A sample of the current holdings: Marvell (MRVL), Astera Labs (ALAB), and Nebius (NBIS). The full ten are on the portfolio page.

Portfolio #2: Reddit's Favorites

Reddit can move a stock, but the most-mentioned list is almost always the same megacaps: Nvidia, Tesla, Apple. The opportunity is not there. It is in the smaller names that suddenly get talked about far more than usual, the moment retail attention starts building before the funds catch on. The problem with chasing that signal raw is that it also surfaces pure meme plays with nothing underneath them. This portfolio is the signal with a quality gate bolted on.

It looks for companies with a market cap between $1 billion and $100 billion and bullish Reddit sentiment, then requires a high AI Score, revenue growing year over year, and positive cash flow. The market-cap band cuts out both microcaps and mega-cap noise. The financial filters are what separate this from a meme basket: you get the retail momentum, but only in companies with a real business behind the attention.

It was the strongest performer of the three in the backtest, returning a hypothetical 20.0% over six months, 73.5% over twelve, and 89.9% over twenty-four. It also had the smoothest ride, with the smallest peak-to-trough drop at 7.5% and the highest share of positive months at 65%. One thing to weigh against those numbers: the quality filters were designed in 2026 and applied backward to historical data, so they show how today's strategy would have behaved, not what a live portfolio actually did. Treat this one as illustrative rather than realized. The live test, from May 7 to June 5, 2026, returned a simulated 14.24% against the S&P's 2.43%.

A sample of the current holdings: Rocket Lab (RKLB), AST SpaceMobile (ASTS), and Robinhood (HOOD).

Portfolio #3: Hyper-Hiring Growth

Companies do not open hundreds of job listings on a hunch. Aggressive hiring usually means signed contracts, approved budgets, and growth that has been committed to internally but has not yet reached the income statement. Research from JPMorgan and Two Sigma has found that job postings tend to lead reported revenue by one to three quarters. This portfolio buys that lead time, holding the companies adding headcount fastest on LinkedIn, then applies a quality gate so the growth comes with substance rather than just a hiring page. Public Storage (PSA), one of the current holdings, grew its job postings roughly 88% over a recent 90-day stretch, the kind of move the strategy is built to catch.

Its record is the most mixed of the three, and the timing matters. Over six months it returned a hypothetical -2.1%, trailing both benchmarks. Over twelve it returned 10.5%, still behind. Only over the full twenty-four months did it pull ahead, at 57.3% against the S&P's 47.5%. A weak early-2026 stretch, dragged down by financial-sector names, is the main reason recent figures look soft. The live test, from May 7 to June 5, 2026, returned a simulated 3.56%, narrowly ahead of the S&P's 2.43%.

A sample of the current holdings: Okta (OKTA), Public Storage (PSA), and Western Alliance Bancorporation (WAL).

Volatility sits between the other two, higher than the S&P 500 but lower than Reddit's Favorites, and the quality screen tilts it toward profitable, growth-oriented companies. It fits an investor who wants a forward-looking growth signal without meme-stock exposure, and who is comfortable holding a strategy through a stretch where it has lagged.

What the backtests show

The two-year figures are a backtest. We applied each strategy's current rules to point-in-time historical data, asking on the first of every month which stocks ranked highest by that strategy's signal, then measuring the following month. A backtest is a legitimate way to study a strategy, but it is calculated with the benefit of hindsight and is not the same as having run the strategy live. It can be flattered by hindsight optimization, survivorship bias, and look-ahead bias, and as noted, Reddit's Favorites uses filters that did not exist during the historical window.

The one-month figures are a live test, but a simulated one, covering roughly May 4 to June 5, 2026. Neither set of numbers represents actual client trading. Both are gross of trading costs, which run an estimated 1% to 2% a year in spread friction, and both ignore taxes. Monthly rebalancing creates taxable events in a taxable account, though tax-advantaged accounts like IRAs avoid that. All three strategies are long only, hold US-listed equities, and use no shorting, options, leverage, or crypto.

Twenty-four months is also a limited sample. It covers one significant drawdown and one strong rally, but it has not been tested through a full bear market or recession. Past performance, whether live, simulated, or backtested, does not predict future results, and these are concentrated portfolios that carry real risk, including the possible loss of principal. The full methodology, monthly results, and a candid section on the biases above are in the backtest white paper.

Why this matters

The thesis behind AltIndex has always been that alternative data is the closest thing a regular investor has to the edge the institutions take for granted, and that the edge shows up earliest in signals like hiring, web traffic, and social attention. Following a portfolio on Autopilot does not change that thesis. It removes the work between the signal and the trade, and it lets you see exactly how each strategy has behaved over time before you commit anything to it.

Which portfolio fits, and how much of your money belongs in any of them, depends on your goals and how much volatility you can live with. These portfolios are general information, the same for every reader, and are not tailored to your situation, objectives, or risk tolerance. They are not investment advice. What we can give you is a precise account of what each one holds and how it has performed, which is more than most products will show you.

Sources: AltIndex AI Score and alternative data signals, AltIndex Autopilot backtest (24-month historical, hypothetical) and live test (May 4 to June 5, 2026, simulated), benchmark data for the S&P 500 (SPY) and Russell 2000 (IWM). Performance figures are gross of trading costs, fees, and taxes, which would reduce returns. Autopilot is operated by Autopilot Advisers, LLC, a separate, SEC-registered company that AltIndex is not affiliated with and does not control. The links to follow each portfolio are paid referral links, and Invested Inc. may earn a commission. This article does not constitute investment advice.

Disclosure: This article is for informational purposes only and does not constitute investment advice. AltIndex is a research and analytics publisher operated by Invested Inc. It is not a registered investment adviser or broker-dealer, and nothing here is personalized investment advice or a recommendation to buy, sell, or hold any security. The model portfolios are impersonal, general information, the same for every reader, and are not tailored to your financial situation. The performance figures shown include backtested, hypothetical results and simulated live-test results, neither of which represents actual client trading. Backtested performance is calculated with the benefit of hindsight and may be affected by hindsight optimization, survivorship bias, and look-ahead bias. The Reddit's Favorites quality gates were designed in 2026 and applied to historical data, so those figures are illustrative rather than realized. Past performance, whether real, simulated, or backtested, does not guarantee future results. The portfolios are offered through Autopilot, operated by Autopilot Advisers, LLC, a separate, SEC-registered company AltIndex is not affiliated with and does not control. The links to follow a portfolio are paid referral links, and Invested Inc. may earn a commission when you sign up or follow through them, which is a conflict of interest that does not change the price you pay. AltIndex does not manage your account or execute trades. Investing involves risk, including the possible loss of principal. Always do your own due diligence and consider consulting a licensed financial, legal, or tax professional before investing.