Next-Day Movers

Your agent pulls the newest selection list, then measures the realized next-session moves on prior selections to score how often the model was right.

A daily short list of names about to travel.

Each trading afternoon a model ranks a universe of liquid optionable equities by how far it expects each name to move in the next session. The top names become that day's rows. The model reads implied volatility, realized volatility, and related volatility structure, and it is asked one question: which of these is about to move a lot.

Every past row also carries the answer. Once the next session closes, return and absolute_move are filled in and fixed, so the dataset is both a live screen and its own track record.

Trade movement without predicting direction.

The published market-effect study measured 3,005 selections across 157 names. Selected names realized a 4.8% median move against 2.4% on a normal day, and moves of 5% or larger hit 48% of the time against 24%. That is the whole thesis: the list runs about twice as volatile as the tape, on a schedule you can plan around.

Because the call is about size rather than direction, it fits non-directional structures first: long straddles and strangles, gamma-heavy day trades, or a movement filter in front of an event screen. It also works in reverse, as a list of names not to sell premium on tonight.

• Buy movement: run the newest date into a straddle or strangle screen before the open.
• Score the model yourself: pull a date range and count how many absolute_move readings clear your own threshold.
• Filter an existing watchlist down to the names the model expects to travel.
• Build per-name selection histories from date and ticker to see which tickers keep reappearing.
• Skip short premium in names flagged for the coming session.

Ranked before the session, graded after it.

Selections are generated daily at 3:30 PM in New York over the liquid optionable universe. The list is written once and never revised: a row published on a given date keeps the ticker it had that day, which is what makes the history usable for backtests.

Outcomes fill in on a one-session lag. On the newest observation date, return and absolute_move come back null, because the session they measure has not happened yet. When it closes, both are populated automatically and stay fixed. Pro includes full history and the latest session.

The pick, the outcome, and how to slice it.

The headline fields below are a subset. Every field, with exact types and semantics, is documented in the API reference.

One call returns the list and its scorecard.

One key works across the REST API, the hosted MCP server, and this dashboard. Every response is JSON with a { count, data } envelope.

import requests

url = "https://api.alphanume.com/v1/next-day-movers"
params = {
    "date_gte": "2026-08-01",
    "date_lte": "2026-08-25",
    "api_key": "alp_abc123"
}

r = requests.get(url, params=params)
print(r.json())

curl "https://api.alphanume.com/v1/next-day-movers?date_gte=2026-08-01&date_lte=2026-08-25&api_key=alp_abc123"

{
  "count": 2,
  "data": [
    {
      "date": "2026-08-25",
      "ticker": "NVDA",
      "return": null,
      "absolute_move": null
    },
    {
      "date": "2026-08-24",
      "ticker": "FIGR",
      "return": 21.29,
      "absolute_move": 21.29
    }
  ]
}

What this data does not claim.

• The dataset makes no directional claim. A large negative return is the model being right, not wrong.
• return and absolute_move are null on the newest date until the following session closes. Treat a null as pending, never as a zero move.
• Selection is limited to liquid optionable equities, so illiquid names and tickers without listed options never appear, however volatile they are.
• The published study measures raw price moves. It says nothing about what a straddle on those names cost, so a gross movement edge is not a net options result.
• Only the next session's move is stored. There is no path, no intraday high and low, and no volume on the row.