US30 Short: The Bounce Was the Signal, Not the Recovery
SkyAnalyst AI journal entry: US30 Short on Jul 8, 2026 closed +3R on TP3. Full workspace view, decision log, and AI reasoning, unedited. SkyAnalyst AI journal.
Before the trade, meet the system.
SkyAnalyst is not one AI trader. It is four specialist agents, each with its own data pipeline, each maintaining state between evaluations. They don’t chat in prose. They write structured messages to a shared state object that each reads on every evaluation cycle. That is what makes the system auditable, and it is what this case study shows, step by step, on a setup whose trigger was written down before the bounce it needed even existed.
01. A tape leaning one way
The morning's sheet leaned harder than most. NYSE breadth printed -1101 advancers net against a 5-day average of -63, an order of magnitude below its own baseline. The VIX at 17.27 sat above its 5-day average of 16.64 and was climbing. Ten-year yields at 4.581% were above their average at 4.521%, and the dollar at 101.072 was firm against its own. The Macro Agent read US30 bearish at 72% confidence with moderate tradeability. The Trend Agent read bearish at 70% in a transitioning regime, and attached the qualifier that matters on a day with FOMC minutes due: reduce size.
The restraint was about location, not direction. Price was already pressing fresh lows with support shelves at 52442 and 52408 directly beneath, and shorting into those is paying the worst price on the chart for the most consensus idea in the room. So the plan published a condition instead: wait for a corrective bounce into the 52468 to 52480 zone, the EMA and retracement cluster overhead, and sell only if a 5-minute bar then closed back below 52458, proof that the reclaim attempt had been refused. Invalidation was a 5-minute close above 52567.
The market delivered the whole sequence inside the hour. The bounce came, tagged the zone, and the 14:09 UTC evaluation caught the failure bar as it closed: all seven confluence factors checked at the trigger, quality scored 8.7 out of 10, and the setup carried its C+ grade for the day's conditions rather than the pattern's shape. One evaluation, 63% confidence, and the short filled at 52453 a minute later with the stop at 52564, 111 points above.
The setup has a name among professional traders: a failed reclaim short. A market breaks below a level, bounces back up to retest it from underneath, and gets refused. The short triggers on the refusal, on the logic that the buyers were just handed their best chance to repair the damage and could not do it.
What the pattern is
When support breaks, the trapped longs above it do not disappear; they wait for the market to come back so they can exit at breakeven. The corrective bounce toward the broken level is their moment, and it is also the market's referendum on the breakdown. If the bounce pushes through and holds above the level, the breakdown was a false move. If the bounce stalls at the level and a bar closes back beneath it, the referendum failed: supply absorbed every buyer the recovery attracted, and the path lower reopens with fresh fuel from the longs who now know they are not getting out flat.
Why the bounce is the information
Selling fresh lows tells you nothing except that you were late. The bounce is where the tape reveals itself, which is why the plan treated it as the trigger rather than the threat. The refusal at 52458 carried real information: dip buyers showed up, pressed into a named zone, and were sold to. It is the mirror image of the breakout retest we bought on cable last week, the same referendum run in the opposite direction, and in both cases the entry is the market's answer, not the trader's guess.
Location did the risk work. Because the entry sat at the failure point rather than at the lows, the stop lived a defined 111 points above, just under the invalidation, and the targets stair-stepped into the space the failed recovery had just vacated: 52346, 52236, 52120. The market took all three inside 81 minutes.
Where it fails
The pattern fails when the reclaim succeeds on the second try. A refused retest does not forbid another attempt, and a 5-minute close back above the zone, here 52567, is the read's expiration date, not an inconvenience to be argued with. It also fails in the wrong regime: a failed reclaim inside a strong uptrend is usually a pause, not a turn, which is why the trigger only carried weight with breadth at -1101, volatility rising, and both macro and trend agents independently bearish. The pattern is a knife that only cuts when the tape is already leaning.
The desk doesn't favor this setup, or any single setup. It reads the tape first and fits the pattern to what is actually there, dynamically rather than by preference. Last week the same instrument earned pullback longs; Monday the Nasdaq earned a long at broken resistance; today the Dow earned a short at a resistance that refused to break back. The pattern rotates because the tape does. See SkyAnalyst run your markets with a 7-day free trial.
US30 is trading in a risk-off, breadth-led bearish environment for the NY AM session. NYAD/ADD is -1101 vs 5-day EMA -63, below yesterday’s low and near today’s breadth extreme, which keeps the default bias short. VIX is 17.27 vs 5-day EMA 16.64, above yesterday’s high and elevated enough to favor wider stops and fade/retest entries over aggressive chasing. The Macro Agent is bearish (72% confidence, moderate tradeability), citing price below the 5-day EMA and below yesterday’s low, with deteriorating breadth. Cross-asset confirmation is also bearish: 10Y yield 4.581 > 5-day EMA 4.521 and above yesterday’s high, while DXY 101.072 > 5-day EMA 100.977, a mild added headwind for Dow multinationals. Trend Agent is BEARISH, 70% confidence, transitioning, reduce size, with R=52567, S=52236, VWAP=52708. On 60m, price remains below falling EMAs and VWAP; 15m shows only a corrective bounce inside the larger downtrend. Longs are vetoed by the breadth-extreme / risk-off compound backdrop.
Directional Bias: Bearish
Volatility: High
Setup #1: US30 SHORT
- Directional Bias: Bearish
- Entry: 52452-52458
- Entry Trigger: Only sell a failed reclaim after a 5m bounce into the 52468-52480 retracement/EMA area that then closes back below 52458; do not chase at current price into 52442/52408 support.
- Stop Loss: 52564-52566
- Targets: TP1=52346, TP2=52236, TP3=52120
- R-Multiples: TP1=1.0R, TP2=2.0R, TP3=3.0R (using ~52456 entry / 52566 stop)
- Quality Score: 8.7/10
- Confidence: High - 7/7 confluence when triggered:
- NYAD strongly negative and deteriorating
- VIX above 5-day EMA / elevated
- Macro Agent bearish, confidence >60
- Trend Agent bearish, confidence >60
- 60m bearish structure below EMAs/VWAP
- Entry sits at 5m/15m Fib + intraday reaction zone
- No high-impact USD event within 30 minutes
Risks: 15m MACD histogram is still positive, so this is a sell-the-bounce setup inside a transitioning regime, not a clean momentum chase. FOMC-minutes day = reduced size.
- Invalidation: Any 5m close above 52567, or sustained acceptance back above the reclaim zone with breadth materially improving.
No additional high-probability setup right now.
- No longs: breadth-extreme veto active.
- No breakdown chase: a fresh short through 52408 is lower quality unless it first retests and rejects, because current support is too close relative to required ATR-based stop.
Risk note: Because this is a FOMC day and the Trend Agent says REDUCE_SIZE, keep risk conservative, typically below normal intraday size (for many traders, around 0.5%-0.75% equity risk, not full risk).
Decision log
My plan forbade the obvious trade. Price was pressing fresh session lows into the 52442 and 52408 support shelves, and shorting there means paying the worst location on the chart for the most crowded idea of the morning. The condition I published instead: a corrective lift into 52468 to 52480, the EMA and retracement cluster overhead, followed by a 5-minute close back below 52458 to confirm the reclaim had failed. At 14:09 UTC that close printed. Breadth stood at -1101 net against a -63 baseline, the VIX was above its 5-day average and rising, yields and the dollar both leaned risk-off, the macro read was bearish at 72% and my own read bearish at 70% in a transitioning regime that mandated reduced size. All seven confluences checked at the trigger and the quality score stood at 8.7 out of 10. Confidence 63%. Entering short at 52453, stop 52564, TP1 52346, TP2 52236, TP3 52120.
On a $100k account at 2.0% risk per trade.
Each trade risks +$2,000 (1R). The system's actual scale-out behavior may differ, see disclaimer.
| Scenario | R-multiple | Profit on $100k |
|---|---|---|
| Stop hit (invalidated) | -1R | −$2,000 |
| TP1 hitActual | +0.96R | +$1,920 |
| TP2 hit | +1.95R | +$3,900 |
| TP3 hit (max potential) | +3R | +$6,000 |
All six agents combined.
06. What this trade teaches
The lesson worth keeping is where the information lived. Not in the breakdown, which everyone saw, and not in the fresh lows, which were the crowded trade. The information was in the bounce: the market gave the buyers their retest, at a named level, on the clock, and they were refused. A system that waits for that answer enters later than the aggressive seller and knows something the aggressive seller does not. On this trade the difference was an entry five points below the 52458 confirmation level with a structural stop, instead of an entry at the lows with a stop in the noise.
The accounting deserves its one honest sentence. The ledger banks +0.96R (TP1), or +$1,920 (TP1) on the $100,000 simulated account, while the simulated panel shows the full run to TP3 was worth +$6,000 (TP3), and the metrics row records that the position never ticked against entry on its way there. That 2R gap between booked and traveled is the price of the conservative TP1 methodology, and we pay it every week because the weeks where the ladder does not complete are the ones the methodology exists for.
The subtler lesson is the grade. This setup scored 8.7 out of 10 on pattern quality and still wore a C+ grade with a reduce-size mandate, because the quality score priced the chart while the grade priced the day: FOMC minutes pending, a transitioning regime, a tape stretched an order of magnitude beyond its breadth baseline. Both assessments were correct at once. Conviction about the pattern and caution about the conditions are different axes, and the system is allowed to hold both.
07. From the desk
Case study #109 is the journal's deepest completed ladder to date, and the desk's first Dow short after a stretch where every US30 entry was a pullback long. We resist reading either fact as a headline. The flip is just the tape doing what tapes do, and the 333 points are what happens when a correct read meets a market in a hurry. The part we would teach is smaller and more durable: the trigger was written down before the bounce existed. The system did not react to the rally; it had already decided what the rally would have to prove.
Through July 8 the ledger reads 11 trades and +1.22R for the month at a 63.6% win rate, with the year at +6.73R across 50 trades at 60.0%. Quiet numbers, deliberately measured at TP1. Days like this one are why the full-potential column exists alongside them.
At a Glance
What this teaches about AI-driven trading
What is a failed reclaim short?
It is a short entered when a market that broke below support bounces back to retest the broken level from underneath and gets refused. The confirmation is a bar closing back below the retest zone, evidence that the recovery attempt attracted buyers and supply absorbed them. The entry sits at the failure point with the stop above the zone, so the trade risks a defined amount against the level whose reclaim would disprove it.
Why sell the bounce instead of the breakdown to new lows?
Because the lows are the worst location for a crowded idea. Shorting fresh lows pays the most extended price of the morning and parks the stop inside ordinary noise. Waiting for the corrective bounce does two jobs: it moves the entry to a structural level with a defined invalidation just above, and it extracts information, since a bounce that fails at a named zone proves sellers are still in control rather than assuming it.
How can a trade score 8.7 out of 10 and still be graded C+?
The two numbers price different things. The quality score measures the pattern itself: seven of seven confluences checked at the trigger, structure aligned across timeframes, a clean invalidation. The grade prices the conditions around it: FOMC minutes pending, a transitioning regime, and volatility rising, all of which argued for reduced size regardless of how clean the chart looked. A system can be confident in the read and cautious about the day at the same time.
When would this short have been invalidated?
A 5-minute close above 52567 would have ended the read, and the stop at 52564 sat just under that line, capping the loss at -1R on the 111-point risk. A successful reclaim on a second attempt would have meant the breakdown was a false move and the trapped longs escaped, removing the fuel the short depended on. Neither happened; the market never traded a point above the entry before the ladder completed.
Run your markets with SkyAnalyst
Seven-day free trial. No credit card. Full access to the Trend Agent, Macro Agent, and six-factor confluence scoring.
Trading involves substantial risk of loss. Past performance is not indicative of future results. The analysis shown was produced by an AI model operating on SkyAnalyst’s live trading infrastructure; it is shared for educational and research purposes only and is not financial advice. About reported results. Every AI Trader publishes three take-profit targets (TP1, TP2, TP3) per trade. The broker closes 100% of the position at TP1, so two distinct R-multiples appear in this article. The hero R-multiple is the full-potential R: where the market actually traveled (the highest take-profit hit, or the stop loss) before the setup was invalidated or exhausted. The realized R, shown on the TP1 row of the simulated returns panel, is TP1’s R (or -1R on a stop out). The realized R is what we log to our running track record. Both numbers are honest. Showing both is what lets readers see the full arc of the move and the conservative ledger entry it produced. Simulated returns in this article are calculated against a hypothetical $100,000 account at 2% risk per trade (1R = $2,000). These are educational reference figures and do not reflect any specific account or broker execution. Your actual result depends on your position size, your risk parameters, and live market conditions.
From the SkyAnalyst Journal
GBPUSD Short: A 15-Pip Stop That Only Works at the Overlap
Cable gave the desk a 15.3-pip stop and paid 38.2 pips against it. Case study #110 is about when the trade existed: the London to New York overlap, where the session itself is the risk tool.
Weekly Losses, Week of Jun 29: Four Stops, One Idea
Four losses for -4.00R inside a week that still closed green. Three of them arrived in fifty-two minutes on the same thesis, and that clustering, not any single stop, is what this report is about.
Weekly Recap: The Week the Pullback Paid, Then Charged Tuition
Eleven trades across five instruments, seven wins, four losses, +1.48R net. One entry shape ran through the whole week, paying through Wednesday's four-winner burst and presenting the bill on Thursday morning.