DC Market Intelligence Report –Sep 21, 2026

01 The Market Verdict

Yesterday’s strong-steamer cohort produced 6 winners from 45 qualifiers, a 13.3% strike-rate that looks superficially “normal” but sits below what the closing market said should happen. At final prices the group was priced to win 8.28 races; it delivered 6. That gap is the story: the performance index of 72.5 means the late money was materially less accurate than the market’s own expectation.

The place rate (35.6%) matters here because it shows the market wasn’t simply “blind” across the board: plenty of the steamers ran well without converting. That profile—competitive runs, fewer wins than implied—typically points to over-confidence in the win portion of the price rather than a complete misread of ability. The level-stakes figures underline the same point: a loss at 10am (-15.83 pts) worsened further by SP (-22.58 pts), consistent with a day where taking shorter late prices was punished rather than rewarded.

Crucially, the one-day miss is not an isolated blip. The 30-day picture shows the same direction of travel: 309 wins against 374.22 expected, a performance index of 82.6. That is a persistent underperformance versus closing odds expectation, not just variance. The other important 30-day clue is the split between 10am and SP outcomes: a very large positive theoretical P/L at 10am (367.11 pts) flips to a large negative at SP (-486.28 pts). In plain terms, the edge—where it exists—has been earlier in the day; chasing the move has been expensive.

02 Market Beaters vs Money That Misfired

03 Where the Money Went

04 The Practical Read

1. Separate competitiveness from win-value. Yesterday’s place rate (35.6%) says the market often found runners that belonged; the performance index (72.5) says it too often paid win prices that were too short. Treat a big contraction as evidence of intent/information, not proof that the remaining win price is fair.
2. Respect the time-of-day split. The 30-day record shows a stark divergence: theoretical returns are positive at 10am but deeply negative at SP. That is consistent with an environment where the crowd is good at identifying the right runners but bad at paying the right price once the move is obvious.
3. Use the size of the move as a prompt to demand more confirmation, not less. Yesterday’s average 10am→final move was 5.64 percentage points, with extreme examples like Chairmanforlife (17.00 → 5.50) still not winning. When the market compresses that hard, the risk is not “wrong horse” so much as “no margin left.”
4. Don’t build narratives from one day’s connection clusters. Trainers and jockeys with multiple backed runners mostly did not convert yesterday; that may be noise. Keep a rolling view before treating “stable money” as anything more than a note in your working file.
5. No filter-supplied qualifiers were available at generation time, so today’s Live Market work should lean even harder on process: price discipline, late-vs-early comparison, and insisting on a realistic win price rather than simply following the move.