Options desk

The Desk

Tape Fri 2:09 ET

$35,952

19 names · 51 contracts

Tape Fri 2:09 ET

August 2026

The edge paid. The activity didn’t have to.

+$11.5k on a 70% book with winners 1.76× the losers. That’s the style. 118 closes across 40 tickers is not. TFSA and zero commissions took the brake off. The written cut and target have to be the brake now.

Realized
$11,505

YTD $62,587

Win rate
70%

62 / 26 · 15 BE

Closes
118

5.9 / day · 40 names

Payoff
1.76×

+$244 / -$138

Keep doing

This is the book.

Two lots

Sell one. Ride one.

HOOD +$820, LEU, DT, MRK harvest +$391 then the $10.75 add, AAPL holding both through 315. That’s the book working. One lot is how NVDA and SPCX became moods.

Wolfe shorts

The only short-dated.

CRWD Aug, LLY Aug, LEU Sep, MRK Oct $135, GILD Oct. He’s the machine. If it isn’t Wolfe, it isn’t a week-to-month call.

Payoff

70% and 1.76×.

Average winner +$244, average loser −$138. That’s a real edge. August didn’t need 118 closes to print +$11.5k. The edge already pays. Activity is the leak, not the engine.

BE closes

Fifteen of them.

SONY +$54, QLYS $200s +$45, MNST scratch. Killing a name that never earned a seat is process. Selling SPCX on the pin because the tape felt loud is not.

Analysis

What the month actually said.

Concentration

Nine lines made the month.

Top nine winners = +$10,635. That’s 70% of the +$15,097 gross. The other 53 winners averaged ~$84. August still prints +$11k if you never touch the long tail. The 118 closes live in that tail — AMPL 13, PFE 6, GIS 5, KTOS 4, BABA 4, eleven day trades. The edge is the nine. The fatigue is the other fifty-three.

The leak

GOOGL and the day-trade pile.

Two GOOGL strikes −$2,078 is 58% of all loser dollars. Add the listed day trades (−$385, −$200) and you’re at −$2,663 before AMKR/MNST/RELL/AAPL $305. Losers that had a written cut (MNST, RELL, AAPL 305) are the cost of doing business. GOOGL twice and the unlisted same-days are optional.

Payoff is not the problem

1.76× already works.

62 winners at +$244, 26 losers at −$138, 15 BE. You don’t need more trades to lift +$11.5k. You need fewer names so the two-lot structure can actually run. NVDA and SPCX felt like moods because they were one lot with no second to sell. MRK today is the correction.

This week

Process vs reaction, same Friday.

MRK: harvested +$391, bought the 147 bounce at $10.75, two lots, $145 cut. AAPL: held both through 315 instead of selling 320 mid-breakout. QLYS $200s scratched +$45 (thin chain, fade), $190s back on a hold — that’s a rewrite, not a chase. SPCX: cut was $134.20, you sold the 140 pin, close 141.45. Same day. The first three are the book. The last one is why this tab exists.

TFSA

No fees, no tax, no brake.

That’s a gift on PLTR and GLD. It’s a trap on GIS fill five. The market will not charge you for a sixth PFE. The desk has to. If the cut and target aren’t written before the order, it doesn’t go in. Short-dated stays Wolfe. Everyone else is dated, two lots, and the next action after a fill is the target — not a new ticker.

September

Same edge, fewer closes.

Watch closes per week, not P&L. A week that only prints targets and cuts that actually hit is the month you already know how to run. Jan wind-down stays on. No restocking Sep/Oct unless Wolfe. Re-entry needs a new weekly close, not a 52-week high the day after you harvested the ladder.

Kill

This is August leaking.

Day trades

Eleven. Dead.

TSLA, BABA, GDX, MITK, DRAM, QUAD, MUU, plus the Aug 11 unlisted −$385 and Aug 21 −$200. Zero-fee TFSA made this free to attempt. It is not free in attention. If it isn’t on the desk with a cut before the fill, it doesn’t get a same-day round trip.

Pin sells

SPCX today.

Cut was $134.20. Target was 147. You sold $16.23 into a 140 pin. Close 141.45. The first lot would have been green too. Volatility between the two written numbers is not a signal. If the cut didn’t print, the next action is the target.

GOOGL

Two strikes, −$2,078.

$350 Dec and $400 Dec. That’s 18% of August’s gross winners eaten by one name, two expressions. No third line on a name. If the first is wrong, the second doesn’t fix it.

AMPL 13 closes

A ladder is one trade.

Thirteen closes is a hobby. You harvested the Jan ladder +$345 the right way, then almost bought Jan ’28s at the 52-week high the same week. One setup, two lots, done. Re-entry needs a new weekly close, not a new mood.

Churn names

PFE 6 · GIS 5 · KTOS 4 · BABA 4.

GIS wasn’t even a position. PFE doesn’t belong in a 118-close month. If it isn’t Wolfe and it isn’t on the desk, it doesn’t get a second fill.

Focus

September is fewer closes, same edge.

Write it first

Cut and target before the fill.

The desk exists so the fill doesn’t invent the plan. If you can’t say the cut and the one-lot target in one line, don’t send the order.

Two lots max

Unless Wolfe says size.

Third UBER was the exception that proves it — cheaper fill on a 200-day hold, still one strike. GOOGL two strikes is the anti-pattern. NVDA is stuck as one lot and that’s why it feels like a mood.

Short-dated = Wolfe

Everyone else is dated.

Jan and later, or don’t bother. You already started the Jan wind-down. Keep it. GH Sep and SRPT Sep were the last of that pile. Don’t restock it.

Closes per week

Count them.

118 in 20 days is 5.9 closes a day. September target: if a week has more than the targets and cuts that actually printed, you’re back in August. The number to watch isn’t P&L. It’s closes.

What paid

What ate it

GOOGL alone −$2,078. That’s one name, two strikes. The day-trade pile is next.

Observed, not scolded: the two-lot days (HOOD, MRK, LEU, AAPL through 315) are you. The pin-sell into 141.45 is the other guy. Short-dated stays Wolfe. Everything else is dated, two lots, cut written, and the next action after a fill is the target — not a new ticker.