Name: Local Tail Variance Ratio (LTVR)
Brand: AlpineDataWorks.AI
SKU: ADW-105
Availability: InStock

Use cases

What it unlocks

⌁

For an agent

A hedge-timing agent monitors ADW-105 daily; when ltvr_raw (tail variance fraction) pushes the score above 75 and the trend field reads 'rising' — the current score is 74.3, 83rd percentile of a 2,454-day history, rising — it automatically initiates a protective-put spread on the underlying index, sizing the hedge proportional to the distance of the LTVR score from the 90th percentile, and records the IOM's confidence and source_lineage in the trade memo. The 60-day rolling tail-variance decomposition gives the agent a statistically validated reason to act before realized vol spikes, rather than after.

📈

For the business

A risk manager at a family office uses LTVR to distinguish between two superficially similar environments — moderately elevated realized vol with returns normally distributed around the mean (low LTVR) versus the same realized vol level but with variance concentrated in extreme return days (high LTVR). At a current 74.3 score the office knows variance is increasingly concentrated in the tails, which changes the protection strategy from simple delta hedges to convex payoff structures; this nuance was previously invisible in their standard deviation and beta reports.

Forward outlook

Prediction

Horizon: —
Recommended use: Track tail variance concentration to detect accelerating tail risk. Score > 65 = variance clustering in extremes — consider hedging.

Methodology

How it's built

60-day rolling window → tail (20th/80th pct) variance / total variance → LTVR ratio → percentile-rank → 0-100; high = variance concentrated in extremes

Stooq daily OHLCV (SPY, keyless)

Version ADW-105-live-1.0 · validated to beat a naive baseline · benchmark: none

Live response

The object an agent receives

One call returns the answer with its reasoning attached — the live Intelligence Object for ADW-105.

GET /v1/intelligence/adw-105

{
  "product_id": "ADW-105",
  "entity": "S&P 500 (SPY)",
  "score": 74.3,
  "trend": "stable",
  "confidence": 0.8,
  "top_drivers": [
    {
      "factor": "ltvr_raw",
      "contribution": 2.1791
    },
    {
      "factor": "ltvr_percentile_rank",
      "contribution": 0.7435
    },
    {
      "factor": "realized_vol_20d_ann",
      "contribution": 0.1622
    }
  ],
  "recommended_use": "Track tail variance concentration to detect accelerating tail risk. Score > 65 = variance clustering in extremes — consider hedging.",
  "methodology_version": "ADW-105-live-1.0",
  "freshness": "2026-06-26T20:00:16.189Z",
  "coverage": "S&P 500 ETF daily returns — 252 days, 60-day rolling window",
  "source_lineage": [
    "Stooq daily OHLCV (SPY)"
  ],
  "allowed_use": "informational",
  "ltvr_raw": 2.179142,
  "tail_window_days": 60,
  "tail_lo_pct": 0.2,
  "tail_hi_pct": 0.8,
  "realized_vol_20d_ann": 0.1622,
  "validation_status": "descriptive"
}

IOM schema