DC Market Intelligence Report –Sep 8, 2026

01 The Market Verdict

Yesterday’s strong-steamer set broadly did what the closing market said it should do. On 69 qualifiers, we saw 12 winners against 11.19 expected, producing a performance index of 107.2. That is a small but real beat versus the market’s own closing probability, and it matters because it shows the late contraction (10am to the off) carried information on the day rather than being pure noise.

However, the wider context remains the more important signal. Over the last 30 days, strong steamers have run at 87.2 on the performance index (296 winners vs 339.33 expected). One positive day does not repair a persistent underperformance versus the market’s final view. The implication is that while yesterday’s tape read cleanly, the underlying 30-day pattern warns against assuming that “strong steam” automatically equates to edge once prices have fully converged.

The pricing evidence also shows where the strain sits. The average Open→10am move was marginal (‑1.03pp), but the average 10am→final move was large (4.00pp), meaning the main information flow arrived late. Yet the theoretical level-stakes picture is split: a small loss at 10am (‑1.84 pts) but a much larger loss at SP (‑17.27 pts). That combination is consistent with a market that is good at finding the right horses, but unforgiving on the price once the late money has done its work.

02 Market Beaters vs Money That Misfired

03 Where the Money Went

04 The Practical Read

1. Use yesterday as a reminder that strong steam can be predictive on the day (PI 107.2), but do not overweight it against the 30-day baseline (PI 87.2). Treat a single good day as a data point, not a regime change.
2. Separate “big move” from “big statement”. Contractions like 101 → 51 can look dramatic in percentage terms, but the implied-probability shift may still be modest versus the type of squeeze seen in shorter-priced runners.
3. Respect the information in close losers. A 2nd like Crown The Future or a 3rd like Ghost Jury can still validate the direction of the money; it simply didn’t translate into the win outcome. That distinction matters when you’re judging whether a move was “wrong” or merely “not rewarded”.
4. Be forensic about price versus probability. Yesterday’s level-stakes profile (‑1.84 pts at 10am vs ‑17.27 pts at SP) is consistent with a market that compresses away most of the value by the off. The closer you operate to the final price, the more your outcome tends to resemble the market’s expectation rather than beat it.
5. Watch for confirmation across time windows. The average Open→10am move was small while 10am→final was large, so the day’s signal mostly arrived late. In live monitoring, that means late acceleration matters more than early firmness.
6. Sample size discipline: one-day clusters (e.g., a trainer with four supported runners) are worth logging, not concluding from. Only repeated recurrence over time can upgrade “where the money went” into a usable pattern.
7. No filter-supplied qualifiers were available at generation time, so today’s Live Market read should be built from real-time movement and context rather than pre-listed candidates.