S&P 500 0-DTE Strike Band

Your agent reads today's band, then replays the historical bands against realized index ranges to report how often the level actually held.

Two strikes that describe the day's expected range.

0-DTE means an option that expires the same day it is traded. SPX lists an expiration every trading day, so traders are constantly picking strikes for a contract with hours left to live. This dataset answers the question that has to come first: how far is the index likely to go before the close.

Each row is one trading date with a lower strike and an upper strike. Together they mark the range the model expects the index to hold through the same-day close, rounded to strikes that actually exist on the board.

Put a number on how far the index travels today.

The published market-effect study ran 622 trading days and found band width correlates 0.80 with the day's realized range. The width is doing real work: when the model prints a wide band, the index tends to travel further, which is more useful than a containment rate on its own.

That makes the band a strike-selection input first and a sizing input second. Short strikes go outside it, spreads get structured around it, and on a day it prints unusually wide the honest response is often smaller size rather than wider wings.

• Place the short legs of condors and credit spreads outside upper_strike and lower_strike.
• Rank days by band width to separate quiet sessions from ones the model expects to move.
• Backtest containment on your own definition of a breach: any touch, or the close only.
• Compare the morning band against the realized high and low to build your own calibration curve.
• Use band width as an intraday sizing input rather than as a trade signal on its own.

Set in the morning, never revised after the close.

The band is derived from forward-looking implied probability distributions, realized volatility data, and broader market risk factors, and published each trading morning at 10:30 in New York. It reflects only what was available at calculation time.

Once published, a band is fixed. That is what makes containment testable: a study over the history scores the same two numbers a trader had at 10:30, not a range fitted after the day was over. Pro includes full history and the latest session.

The band, and the filters that pull it.

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

One call returns today's band, or years of them.

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/spx-0dte-strike-band"
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/spx-0dte-strike-band?date_gte=2026-08-01&date_lte=2026-08-25&api_key=alp_abc123"

{
  "count": 1,
  "data": [
    {
      "date": "2026-08-25",
      "lower_strike": 6740,
      "upper_strike": 6880,
      "instrument": "SPX"
    }
  ]
}

What this data does not claim.

• Containment in the published study was measured with SPX proxied by SPY times ten. The tests are run basis free, but the proxy is a rough edge and it is stated rather than hidden.
• Strikes are rounded to listed SPX increments, so the band is a tradable level, not a price forecast to the point.
• One row per trading day, with no intraday update. A band published at 10:30 does not react to a shock at 2 PM.
• The row gives a range, not a direction and not an option price. Whether selling that range pays depends on the premium, which is not in this dataset.
• Containment says nothing about path. The index can breach the band midday and close back inside it, which is why width and close containment are reported separately.