DC Market Intelligence Report –Sep 18, 2026
What the market knew, what the results proved, and what the evidence means for today.
Yesterday’s Market Scorecard
Underperformed expectation
01 The Market Verdict
Yesterday’s STRONG STEAMERS profile was clear: the market was decisive late, but it was not accurate enough to justify that confidence. From 101 qualifiers (each contracting at least 15% from 10am to the off), the group produced 13 winners against 16.17 expected at final implied prices. That gap is material: a performance index of 80.4 says the late money under-delivered versus what those closing prices promised.
The price-action itself was not subtle. Average Open→10am movement was near-flat (-0.37pp), then the average 10am→final move was a hefty 4.68pp. In other words, most of the conviction arrived after 10am, and it arrived aggressively. When that late pressure repeatedly fails to convert at the rate the closing prices imply, it points to either (a) overconfident compression, (b) information that was broadly shared but not decisive enough to beat race variance, or (c) price-taking behaviour that forced horses shorter than their true chance.
Importantly, yesterday is not an outlier against the broader backdrop; it is consistent with it. Over the last 30 days, qualifiers ran at 306 wins from 2,276 with 359.68 expected: a performance index of 85.1. That is the same story as yesterday—closing prices have been systematically “too short” relative to outcomes. One day can be noisy; a month at mid-80s on the index is a pattern.
The returns framing underlines the timing problem. Theoretical level-stakes P/L was +9.34 pts at 10am prices but -27.91 pts at SP yesterday, and the 30-day split is even starker (+381.14 pts at 10am versus -437.66 pts at SP). That is not a claim of repeatable edge; it is evidence that much of the “value” is being competed away by late contraction.
02 Market Beaters vs Money That Misfired
Market beaters
- Nouveau Grey (10.00→5.00; SP 4.00): a clean example of sustained support that was confirmed again at SP and converted; this is what “right money, right direction” looks like.
- Santerno (2.88→1.91; SP 1.67): the market pushed it into a short-price profile and it delivered, helping to offset the broader underperformance of compressed favourites.
- Our Son Oscar (3.00→2.00; SP 1.67): another winner where the late steam aligned with an even firmer SP, indicating follow-through rather than a late squeeze that snapped back.
- Ariane Sky (4.33→2.75; SP 2.75): the final price held into the off and the result matched the market’s re-rating—useful evidence that not all strong late moves were “false”.
- Opal Storm (36.00→23.00; SP 23.00): a winning longshot steamer matters because it shows the support was not restricted to obvious contenders; however, these outcomes are inherently lower-frequency and should not be over-weighted.
Money that misfired
- It'sneverjustone (26.00→8.00; SP 7.00; 5th): the most violent contraction on the day did not translate into a win or even a place, which is the clearest “over-confident” signal in the sample.
- Sunny Orange (8.00→3.75; 5th): a major re-price into the front of the market that failed to hit the frame—exactly the kind of miss that drags the performance index down because expectation is high at these closes.
- Midnight Thunder (3.25→1.83; 8th): the market went very hard (23.88pp move) and was wrong; these are the losses that do the most damage to SP outcomes because they concentrate implied probability.
- Grindleton (9.00→5.00; SP 4.33; 2nd): a non-winner but not a “dead” read; the money identified a major contender and was close to being right, which is still useful information when assessing signal quality.
- Tommie's Angel (36.00→19.00; SP 29.00; 3rd): the place is meaningful, and the SP drift away from the final price suggests the late steam did not fully persist into the off—an example of why confirmation close to post time matters.
03 Where the Money Went
Repeated connections
- Trainer concentration was visible but not decisive on a single day: Tim Easterby, Linda Perratt and R. Mike Smith each had 3 strong steamers backed and none won. That reads as market interest without payoff yesterday, not proof of a stable-level pattern.
- Several trainers converted from smaller volume: Richard & Peter Fahey (2 backed/1 won), James Owen (2/1), Kevin Ryan (2/1) and Marco Botti (2/1). With counts this small, treat it as “money found winners” rather than any durable edge.
- Jockey support was similarly split: Shane Gray (3 backed/1 won) and Lewis Edmunds (3/1) landed a winner each, while Oisin Murphy, Saffie Osborne, Billy Loughnane, Cieren Fallon and Jason Hart each had 3 backed with 0 winners—again, notable but not statistically persuasive from one card-set.
- At the 2-backed level, W. J. Lee, Harry Vigors, David Nolan, Rossa Ryan and Silvestre De Sousa each returned 1 winner from 2. The market can cluster around riders on a given day; the key is whether that clustering continues over time, not whether it happened once.
The other side of the market
- Storm Force Ten (2.0→4.0; 5th): a major late drift that avoided winning; the drift did not “find the floor” but it also did not signal a collapse to the back—typical of a horse losing win-credibility while retaining some competitiveness.
- Nando Royale (2.75→4.5; 2nd): the most instructive drift on the list; the horse nearly won despite weakening significantly. This is a reminder that drifts can be about win confidence rather than overall ability to run well.
- River Nile (1.91→2.63; 4th) and Glitter Code (4.33→8.0; 4th): both ran credibly in defeat, reinforcing that the “drifter = no show” shortcut is unreliable.
- Darzah (2.2→3.5; 7th) and Wopbopaloomop (2.9→5.0; 8th): these are the drifts that did align with clear underperformance, illustrating the more straightforward case where the market is stepping away from a weak proposition.
04 The Practical Read
- Respect the direction, interrogate the price. With a one-day performance index of 80.4 and a 30-day 85.1, the recurring issue is not “no information” but “too-short closers”. Treat heavy late contraction as a prompt to measure whether the final number is now doing the backer’s work.
- Confirmation matters because some steam does not persist. Cases like Tommie's Angel (final 19.00, SP 29.00) show that a 10am→final move can partially unwind by the off; when it does, it changes the meaning of the earlier support.
- Use pp movement to separate gentle support from force. Yesterday’s average 10am→final move of 4.68pp says the cohort was being re-rated materially, but single-horse extremes (e.g., very large contractions) are the most expensive when wrong and the biggest driver of SP losses when the win doesn’t arrive.
- Do not treat “missed” as “no signal”. Placings and close defeats (e.g., Grindleton 2nd; World Star 3rd; Tommie's Angel 3rd) still indicate the market found live runners. The actionable distinction is between “live but not enough at the price” and “never involved”.
- Keep sample size discipline. One day’s trainer/jockey clustering is descriptive only; the more robust read is that both yesterday and the last 30 days show late-shortening prices have, on average, overstated true win chance.
- No filter-supplied qualifiers were available at generation time, so today’s Live Market work should start from price/volume behaviour and confirmation rather than pre-flagged names.
Data AppendixMethod and report totals
- Report generated
- 2026-09-18 10:01:57
- Analysis date
- 2026-09-17
- Strong steamer threshold
- 15%
- Races
- 39
- Runners
- 418
- Strong steamers
- 101
- Landed
- 13
- Missed
- 88
- False steamers listed
- 10
- Trainers tracked
- 12
- Jockeys tracked
- 12
- Qualifiers today
- 0
- Expected winners
- 16.17
- Performance index
- 80.4
- Placed
- 40
- 10am level-stakes P/L
- 9.34 pts
- 30-day qualifiers
- 2276
