Doubled the Size, Cut It in a Day

Week of May 4–May 10, 2026

This is the week the account got bigger for exactly one day. Tuesday, May 5: close the old strangle for a clean win, then reach for size, first trying to sell five contracts of a new /ES strangle, settling on a fill for four at a 4.25 credit. Every /ES trade before this one had been two contracts. This was double. By Wednesday afternoon it was gone, bought back at 4.90 for a $163.92 loss, and the very next order that evening was a new strangle at the old size: two contracts, wider call strike. The experiment opened and closed inside 24 hours, and the account never traded four lots again.

What actually happened

Three moves, all packed into Tuesday and Wednesday:

Date	Action	Details
Tue May 5	Closed	2x /ESM6 Jun 5 4900P/8000C strangle, 2.00 debit. +$283.04
Tue May 5	Opened	4x /ESU6 mid-June 5000P/8100C strangle, 4.25 credit
Wed May 6	Closed	Same 4-lot, 4.90 debit. -$163.92
Wed May 6	Opened	2x /ESU6 Jun 26 5200P/8400C strangle, 5.40 credit

Two details from the raw order log are worth pulling out, because they say more than the fills do. First: before the 4-lot filled, there was a canceled order for five contracts at a lower credit. The intent Tuesday was to go even bigger than what filled. Second: Wednesday’s exit wasn’t one order. It was a ladder of five repriced buyback attempts inside about an hour, 4.45, 4.55, 4.75, 4.80, 4.85, before finally paying 4.90 to be out. Whatever the reason for leaving, the log shows someone chasing the exit up, willing to pay more every few minutes rather than stay in the position.

Also this week: the roll from the June /ES contract to the September contract (/ESM6 to /ESU6), which is routine futures housekeeping rather than a strategy change, but it’s in the log so it’s noted.

What moved the market this week?

The S&P 500 was in the middle of a melt-up: up 2.3% on the week, a sixth straight winning week, with new all-time closing highs on Friday, May 8, the same day April payrolls came in at 115,000 jobs against expectations of 55,000.

Two honest caveats. The exit happened Wednesday, two days before the jobs report, so payrolls explain nothing about this trade. And while a market grinding to record highs does mechanically lean on the call side of a short strangle (the 8100 calls were the near side of the risk in a rally), the log records what was done, not why. The mechanical pressure is a fact; the motivation is not in the data.

Anything new tried?

Size. Four contracts instead of two, with an attempt at five. Handled by reversing it within a day at a contained loss and going straight back to the standard size, with the replacement strangle’s call strike moved further out (8400 vs 8100).

Takeaway

This one the data states almost by itself: the size experiment cost $163.92, less than the single win booked the day before, and the account went back to two contracts and stayed at or below that size for every subsequent /ES trade in the record. A loss that undoes half a win and permanently ends an experiment is cheap tuition, if the lesson actually sticks. On this record, it stuck.

The trader’s own verdict, given after the fact, is simpler than any theory: four contracts was too much size for this account. Full stop. Realistic sizing for an account this size works out closer to three contracts, roughly $35K in buying-power terms, and four was past the line. The lesson wasn’t about the market at all; the market barely moved. It was about position size relative to account size, recognized once the position was on and the number was staring back. The full cycle gets its own journal entry.

Frequently asked questions

What was the biggest position size ever taken in this account?

Four contracts, a /ES short strangle opened May 5, 2026 after first trying to fill five. It was bought back at a loss the next day, and no /ES trade after it was larger than two contracts for the rest of the documented record.

How big was the loss on the 4-contract strangle?

$163.92 net of fees: sold at 4.25, bought back at 4.90 within about 24 hours. For comparison, the trade closed the day before it made $283.04, so the loss gave back a bit more than half of that win.

Did the May 8 jobs report cause the loss?

No. The position was closed on May 6, two days before April payrolls came out. The broader backdrop that week was a market grinding to record highs, which mechanically pressures the call side of a short strangle, but the log doesn't record why the exit happened when it did.