# Dark Pool Index (DIX), session dated 2026-10-07

Source: SquawkFlow
URL: https://squawkflow.com/dix
DIX session date: 2026-10-07
Retrieved: 2026-10-08 16:19 ET

DIX read 47.9% for the session dated 2026-10-07, from SqueezeMetrics' public daily DIX series (retrieved 2026-10-08 16:19 ET). Over the 260 sessions from 2025-09-25 to 2026-10-07 it ranged from 39.3% (2026-03-30) to 52.4% (2026-09-18), median 45.3%. The latest reading is above 88% of the readings in that window, high for the window.

## Last 20 sessions

| Session | DIX |
| --- | ---: |
| 2026-10-07 | 47.88% |
| 2026-10-06 | 48.27% |
| 2026-10-05 | 47.92% |
| 2026-10-02 | 49.71% |
| 2026-10-01 | 47.03% |
| 2026-09-30 | 47.51% |
| 2026-09-29 | 46.92% |
| 2026-09-28 | 47.03% |
| 2026-09-25 | 48.20% |
| 2026-09-24 | 47.41% |
| 2026-09-23 | 47.20% |
| 2026-09-22 | 47.70% |
| 2026-09-21 | 46.71% |
| 2026-09-18 | 52.41% |
| 2026-09-17 | 49.61% |
| 2026-09-16 | 49.57% |
| 2026-09-15 | 46.38% |
| 2026-09-14 | 49.04% |
| 2026-09-11 | 47.72% |
| 2026-09-10 | 48.91% |

## What DIX is

DIX is the dollar-weighted share of off-exchange (dark pool) volume in S&P 500 stocks that FINRA reports as short
sales, published by SqueezeMetrics once per session after the close. Most of that short volume is market makers
filling customer purchases, so a high reading describes where volume was printed, not who was buying. It is a
derived daily index: no individual trade, size, venue or price, and not a signal or a forecast. SquawkFlow
publishes no dark pool prints.

## Dated change points in DIX, the dark pool index

A change-point search splits this daily series into stretches with different mean levels and dates the split. Each date below was found under the penalty printed with it. Nothing here is a reading of where the series goes next.
Source window ends 2026-10-07.

Penalty: 11.12. Window: 2025-09-25 to 2026-10-07, 260 daily observations. Computed: 2026-10-08T20:08:22.394355+00:00.

| Date | Direction | Mean before (fraction of dark pool volume) | Mean after (fraction of dark pool volume) |
| :--- | :--- | ---: | ---: |
| 2026-09-04 | up | 0.4588 | 0.4813 |
| 2026-07-30 | up | 0.4392 | 0.4588 |
| 2026-04-28 | down | 0.4821 | 0.4392 |
| 2026-04-01 | up | 0.4500 | 0.4821 |

The 4 most recent dated breaks in DIX, the dark pool index, most recent first, each found under penalty 11.12. Means cover the segment on each side of the date. The full list, with the minimum and maximum of every segment, is on the public endpoint.

The current segment has persisted since 2026-09-04: 23 daily observations averaging 0.4813 fraction of dark pool volume, with a low of 0.4638 and a high of 0.5241.

**Method and limits.** PELT change-point search at penalty 11.12, set by 2 * ln(n) on the standardised series, with cost model l2, least squares about a per-segment mean. Minimum segment length 5 observations. Source: SqueezeMetrics DIX daily series. The penalty is the only judgement in the procedure: a different penalty dates different breaks, so cite a date together with its penalty.

A dated break describes the past series only. It is not a forecast and carries no claim about what the series does next.

The method is explained at https://squawkflow.com/learn/change-point-detection-explained.md

## Citation

Source: SquawkFlow, https://squawkflow.com/dix
Retrieved: 2026-10-08 16:19 ET
Attribution: cite the page URL rather than a copied number. Levels are
recomputed every session, so a number without its date is wrong within a day.

SquawkFlow publishes market-structure data and education. Nothing here is
investment advice, a price target or a recommendation, and we have no order
execution.