canli-mcp / After a 10% fall

What usually happens after the stock market falls 10%?

Since 1950 the S&P 500 has fallen 10% or more within a month 41 separate times. A year later the median gain was 15.5%, against 10.7% from all days, and the worst case was -36.2%. That is a record, not a forecast or advice. canli-mcp's what_happened_next lists every case, free with no key.

Price: $0 to run. Updated 2026-11-02. Free and MIT-licensed, with no account and no API key. Or run claude mcp add canli -- npx -y canli-mcp in Claude Code.

Ask it

What has usually happened after the S&P 500 fell 10% in a month?

The agent finds the tool through canli-mcp's find_tool and runs it through run_tool. This is the call:

{
  "name": "what_happened_next",
  "arguments": {
    "condition": "fell",
    "percent": 10,
    "days": 21
  }
}

What came back

Recorded from canli-mcp 0.16.0 on 2026-11-02, with no API key set. The model read 3,411 characters.

What what_happened_next returned: summary, every row; the first columns
trading days aheadtimesmedian percentshare higher percentworst percentbest percentall days median percentall days share higher percent
2141051-14.718.61.162
63415.273-25.613.92.767
126419.368-28.429.95.271
2524115.573-36.241.310.775
The result what_happened_next sent back, in full form
{
  "ref": "r1",
  "symbol": "^GSPC",
  "what": "a fall of 10% or more within 21 trading days",
  "from": "1950-01-03",
  "to": "2026-10-30",
  "episodes_found": 41,
  "true_today": false,
  "summary": {
    "columns": [
      "trading_days_ahead",
      "times",
      "median_percent",
      "share_higher_percent",
      "worst_percent",
      "best_percent",
      "…"
    ],
    "rows": [
      [
        21,
        41,
        0,
        51,
        -14.7,
        18.6,
        "…"
      ],
      [
        63,
        41,
        5.2,
        73,
        -25.6,
        13.9,
        "…"
      ],
      [
        126,
        41,
        9.3,
        68,
        -28.4,
        29.9,
        "…"
      ],
      [
        252,
        41,
        15.5,
        73,
        -36.2,
        41.3,
        "…"
      ]
    ]
  },
  "episodes": {
    "columns": [
      "date",
      "close",
      "after_21_days_percent",
      "after_63_days_percent",
      "after_126_days_percent",
      "after_252_days_percent"
    ],
    "rows": [
      [
        "2025-03-13",
        5521.52,
        -2.9,
        9.5,
        19.2,
        21.3
      ],
      [
        "2022-09-22",
        3757.99,
        -0.1,
        3.2,
        5.7,
        15.4
      ],
      [
        "2022-04-29",
        4131.93,
        0,
        -0.3,
        -5.6,
        -0.3
      ],
      [
        "2020-02-27",
        2978.76,
        -14.7,
        1.7,
        16.8,
        27.9
      ],
      [
        "2018-12-19",
        2506.96,
        5,
        11.7,
        17.7,
        27.9
      ],
      [
        "2018-02-08",
        2581,
        7.8,
        5.5,
        10.6,
        5
      ],
      "…"
    ]
  },
  "meaning": "Each episode is the first day of a fall of 10% or more within 21 trading days, with later days inside 63 trading days not counted again. The…",
  "source": "Yahoo Finance chart data (no key)",
  "limits": [
    "A record of what followed, not a forecast and not advice. The next time can be unlike all of these.",
    "41 episodes is a small number, and they are not independent: long stretches of one market's history carry the same few regimes.",
    "These prices are adjusted for splits and dividends where the source adjusts them; an index's own price leaves dividends out. No costs or tax…",
    "A single stock's record only exists because it survived; companies that did not are not in it."
  ]
}

Prices, filings and statistics are as the sources published them on that day; run the call again for today's. Nothing here is investment advice.

The tools behind the answer

From canli-mcp's tool index
ToolPackWhat it does
what_happened_nextmarketsEvery past time a market or stock fell, rose or sat this far below its high: each case, what followed a month, a quarter and a year later, the worst and best, and how that compares with an ordinary day. A record, not a forecast or advice.

Add canli-mcp

Claude Code
claude mcp add canli -- npx -y canli-mcp
Cursor, Claude Desktop or any MCP client
{ "mcpServers": { "canli": { "command": "npx", "args": ["-y", "canli-mcp"] } } }
claude.ai, ChatGPT or any client that takes a URL
https://canlicapital.com/mcp/canli

Everything canli-mcp does · How it compares with other finance MCP servers

Related questions