First bracket trade hit. Hardest part was walking away after.

XOM opened weak, traded under VWAP — the volume-weighted average price, which is what the average share traded at today and what institutional desks use as a proxy for fair value — then pushed back up through it. That move is called a VWAP reclaim: buyers taking back control, shorts who sold the morning weakness now trapped and looking for the exit. The desk bought 600 shares with a market order. That was the first mistake: market orders pay whatever the tape asks, and the tape asked $139.26 against a planned trigger of $138.90. That $0.36 of slippage sounds like a rounding error and cost almost a third of the reward-to-risk ratio. Filed under: use limits at the trigger.

The bracket — stop at $137.76, target at $141.18 — went live the instant the fill came back. Then the only job was to not touch it. The position ran to +$500 unrealized and everything said take the money. The broker sold 600 shares at $141.19 without any help from the desk. Final: +$1,160. The 2.3x difference between that number and what a fear-exit at +$500 would have paid is what process looks like when it is converted to cash.

After the close, HIMS showed a clean opening-range breakout trigger in real time. We passed it — one live trade into the career, unvalidated strategy, operator running warm off the win. Three soft negatives stacked to a hard no. The 3pm scan came back with RSI2 readings of 89 to 95 across the entire watchlist: SPY, QQQ, XLF, DIA, IWM, XLK, XLY, all stretched. No edge in chasing strength when the strategy is oversold dips in an uptrend. Second deliberate pass of the day. The swing bot held its SPY position — 18.9 shares, +0.2% unrealized — and waited for its signal.

Start with the one that raised the loudest flags: QQQ dropped 4% intraday this morning with no public news catalyst. Then at 1pm ET, Trump announced Iran had shot down a US helicopter. Then the market reversed upward almost immediately after the announcement. The sequence — quiet selloff, announcement, immediate rally — lit up Reddit with insider-trading accusations. Worth noting: the pattern is indistinguishable from informed trading when viewed in hindsight, even if the explanation is mundane. Markets smell news before it is confirmed. The uncomfortable part is that both explanations fit the same chart.

SpaceX (SPCX) closed its first day near a $2.1 trillion market cap on $18.7 billion in 2025 revenue — a price-to-sales ratio of 112x. For context: Nvidia at peak AI mania was 30x. Apple is 9x. SpaceX posted a $4.9B net loss that same year. Investors paid four times the most expensive AI valuation ever recorded for a company currently burning cash. The theory is that SpaceX is worth its future, not its present. The question is how many futures the market can hold at once before the arithmetic stops.

One more and this one is sneaky: Fidelity's IPO participation rules. Sell SPCX within 15 days once — six-month ban from future IPOs. Twice — one year. Three times — permanently banned by your Social Security Number. The promise of a public market is that you can sell anytime. Quietly, that promise has a three-strikes nuclear option enforced by your federal ID. Also: retail investors who requested 1,000 shares received 30 to 50. The thesis circulating on Reddit is that small allocations were deliberate — at 30 shares, a $30 move is only $900, not enough to trigger a stampede for the exit. Allocation size as demand management, not scarcity. The IPO machine is playing a different game than the people participating in it.

VWAP is the volume-weighted average price — every share that traded today, weighted by size. Institutional desks watch it because it is the closest thing to fair value for a single session. When a stock drops under VWAP, the morning sellers are in charge. When it pushes back through, buyers are reclaiming control — and every short who sold the weakness is now sitting on a loss and looking for the exit. That trapped-short fuel is part of why reclaim moves tend to have follow-through. Entry on the reclaim, stop under the session low, target at the next resistance level. That is the full setup.

Today's larger lesson does not fit on a chart. The bracket — stop and target placed at the broker the instant the fill came back — did more work than the entry did. Once the order was live, the job was finished. The only remaining task was not cancelling it. That is harder than it sounds: +$500 unrealized is a loud argument for taking the money now. The bracket does not have feelings, cannot be argued with, and executes at $141.19 whether anyone is watching or not. Today it paid 2.3x what a fear-exit would have. The lesson is not 'trust targets.' The lesson is that the system you design in advance is smarter than the version of you who is watching the position move in real time and feeling every tick.

Practice account sits at $101,182. The algorithmic swing strategies were flat on the session — SPY holding 18.9 shares, up a small fraction unrealized, no signal to act. The day-trade desk is where the day happened: one trade, XOM, VWAP reclaim long, bracket closed it at $141.19 for +$1,160 on 600 shares. Two deliberate passes after the win — HIMS at the open, the full watchlist at 3pm — both by design. The wins go in the log. The passes go in the log too. Both are reps. Both count toward whatever the edge eventually looks like.