Aug 24, 2026: +$1,093 on a 20% weekend gap, and the same mistake for the third time

My best session since June 10, and once again almost all of it came from a position I bought on Friday afternoon and did not look at over the weekend. The part I traded actively lost money. That is now four sessions running.

NSSC: a 20% weekend gap

Bought 150 NSSC (Napco Security Technologies) at $37.95 at 13:49 on Friday. Sold all 150 at $45.6354 at 07:46 this morning, pre-market. +$1,152.81.

That is a 20.3% move across a weekend on a $5.7bn security-hardware company. Napco reported earnings, and — as with Buckle on Friday — I did not check the calendar before buying. Twice in three sessions I have been paid for a catalyst I didn't know was coming.

Being right by accident spends the same as being right on purpose. It does not repeat like it, though.

I want to sit on that rather than skate past it, because it is the single most important caveat in this entry. If I keep buying into unchecked earnings dates, the distribution has a fat left tail I haven't met yet. A 20% gap down on 150 shares is minus $1,140, and nothing about my process would have prevented it.

Twelve orders, +$2,021, one loser. Over the same eight sessions I placed roughly 120 intraday orders and they are net negative.

XPON: the same mistake, a third time

This is the part I would rather not write, which is exactly why it goes in.

XPON (Expion360) opened weak and I traded it three times in three minutes:

• 09:38:09 in at $7.85 → 09:38:22 out at $7.9195 — +$14 in thirteen seconds
• 09:39:28 in at $7.60 → out at $7.72 and $7.57 — +$2
• 09:41:14 in at $7.50 → 09:41:31 out at $7.12 — -$76

Net -$59.70. Every entry lower than the last. The final one lasted seventeen seconds and lost more than the first two made combined.

I did the identical thing to SGLY and Flowers Foods on Thursday — buy, take a small win, then re-enter repeatedly into a falling price until one of them is large enough to erase the rest. Three sessions, three names, same shape. At this point it isn't a bad day; it's a documented habit with a receipt.

The tell is obvious in hindsight and invisible at 09:41: if my second entry is below my first exit, I am no longer reading the tape. I'm arguing with it.

The honest split

Two orders made a thousand dollars. Eight orders lost sixty. That is the 890-trade finding reproduced inside a single Monday, and it is the fourth consecutive session where the split runs the same way.

Still open

At 12:50 I bought 30 DKS (Dick's Sporting Goods) at $177.30 — a $5,319 position, the largest single line I have taken. It is marked +$96.30 tonight.

Worth flagging the sizing honestly: after two good gaps I have just committed more capital to one overnight position than the account's entire realised profit for July. That is precisely the behaviour that produced the -$1,108 day in July, and noticing it here is not the same as having controlled it.

Day: +$1,093. Running total across published sessions: +$4,441. Net liquidating value $6,416 — the most cash the account has ever held, though the realised running total is still short of the +$5,635 peak from June 11.