The Dow Short That Refused at 83 and Entered at 76

The Dow short on September 8 was refused twice at 83 percent confidence and taken at 76. At 14:55 UTC the gate scored 48 and returned WAIT. A minute later it scored 83, and returned WAIT again. At 14:59 it scored 83 for a second time and still returned WAIT. At 15:00 it scored 76 and entered short at 52,895.4 with a stop at 53,008, a risk band of 112.6 points. Price went to 52,620. Two hundred and seventy-five points, all three targets, +2.45R. The sequence looks backwards for the same reason it did on the Cable short a week earlier: the high scores and the entry score were not measuring the same object. Two of them were describing a bearish tape. One was describing a trade that had become available.

Breadth led, and the trend was already extended

September 8 was a breadth-led bearish session, and the numbers were not close.

The NYSE advance-decline line sat at -500 against a 5-day EMA of -112.6, so participation was decisively below its own baseline. It was not at a five-day low, and that mattered: the prior day had printed -1132, so this was bearish without being the kind of washout extreme that invites a violent reversal.

VIX at 15.56 was above its 5-day EMA of 15.05 and above the prior day's high. Elevated volatility of that shape pushes the session away from clean breakout entries and toward mean reversion within the dominant trend, which is the specific instruction the analysis carried into the open.

Macro leaned the same way without being emphatic. The Macro Agent read lean_bear at 64% confidence with a tradeability score of 62 out of 100, citing stronger payrolls lifting hike odds. The 10-year at 4.792% sat above its 5-day EMA of 4.776, keeping rate pressure on equities. DXY at 98.833 was below its own 5-day EMA, which softened the case slightly without overriding negative breadth.

The Trend Agent was the most confident input on the board: BEARISH at 78% in a TRENDING regime, with resistance at 52,935.6, support at 52,688.8, VWAP at 53,032.1 and invalidation at 53,041.4. The New York open produced an opening-range breakdown, roughly 53,048.8 high to 52,909.6 low, and the break held.

Failed Reclaim of the Broken Opening Range. The pattern is a bounce sold at the level it just broke, and the entire reason this trade waited through two 83s is that the bounce is the pattern. Without it there is no setup, only a direction.

Why an oversold downtrend is dangerous to sell

The 60-minute RSI was at 29.6 with MACD below zero and a negative histogram. The 15-minute was below its EMAs and VWAP with strongly negative MACD. Both frames said the same thing: the downtrend is intact and it is stretched.

A stretched downtrend is the most expensive place to initiate a short. The direction is right, but price is furthest from the level a stop can sensibly sit against, so the risk band has to widen to survive an ordinary snapback. Every point of extra stop width divides directly into the R the trade can produce.

What the entry actually required

The analysis named the condition before the session: sell bounces and failed reclaims, do not chase sub-52,800. That is a conditional instruction, and until the condition prints there is nothing to score highly.

So the system waited for price to lift back into the 50 percent retrace and broken opening-range-low zone around 52,895 to 52,915, and for a 5-minute rejection to close back below 52,900. When that printed, the stop could sit at 53,008, comfortably under the Trend Agent's 53,041.4 invalidation, giving a 112.6 point band on a setup targeting 275.

What the 76 was actually saying

Professional traders: the score dropped from 83 to 76 because the entry carried acknowledged risks that the thesis did not. Both the 60-minute and 15-minute were oversold, DXY was soft rather than confirming, and an elevated VIX makes snapback rallies sharper. Those are real, and the number says so.

SkyAnalyst does not favor any single strategy or a fixed entry trigger. It reads the tape first, scores the entry in front of it rather than the opinion behind it, and accepts a lower number on a real trade over a higher one on a setup that does not exist yet.

What this trade teaches

For four minutes this desk was more confident about the Dow than it was at the moment it sold the Dow, and it made 275 points because of that gap rather than in spite of it.

The GBPUSD short a week earlier ran the identical shape: four refusals at 79, 81, 82 and 84, then an entry at 62. Two trades, two instruments, two very different setups, and the same underlying structure. In both cases the high scores were attached to a market view and the entry score was attached to a position with a defined invalidation.

The second lesson is about what an oversold reading does to a correct directional call. Both the 60-minute and 15-minute frames were stretched at the moment the thesis completed. That did not make the bearish read wrong, it made the immediate entry bad, and the difference between those two statements is worth roughly 275 points here.

From the desk

This is one of two winners the desk has taken this week, alongside a US500 short that closed at its third target for exactly 4R. September now stands at +5.69R across 10 trades, seven winners and three losers, on the conservative TP1 baseline the recaps use.

Three of those ten losses and wins are worth putting side by side, because they describe the same book behaving differently. US30 has now taken two trades in September, both shorts, both winners, for +1.78R combined on the TP1 baseline. That is a book doing exactly what it did through August, when it was the anchor of a losing month at +5.15R across eight trades while everything around it bled.

The EURUSD short that last week's recap flagged as still open has since resolved, and it resolved as a loss. We said at the time that it would be counted in full in the week it closed rather than marked at a favourable level while it ran, and it has been: it enters the September record at -1R, which is the same figure every other loss on this desk carries.

The year stands at +28.75R across 220 trades through the August close, and September's ten trades sit on top of that as an open month.