The structure that cost us most, and why stopping it was not enough
For the first 3 weeks it published anything, tickerseer reached for one options structure on nearly half its calls. That structure went on to produce more of our losses than every other structure combined. This study is about those 3 weeks: what we were doing, what it cost, and whether stopping helped. The whole record is 142 settled calls, 105 right and 37 wrong, every one published before the earnings print it was about. Nothing here is a backtest.
What we did in May
A bull call spread buys an option and needs the stock to move far enough, fast enough, to cover what it paid. Its mirror image sells an option and needs the stock to finish on one side of a line, which it can do by going nowhere. Over these weeks the second bet was usually the right one: the stock stayed on the safe side of the short strike often enough for those structures to win 88.0% and 86.7% of the time.
We knew that. We wrote the other one anyway, 21 times in 3 weeks, because each call was argued on its own merits and nothing in the pipeline was watching what the structures added up to.
What it cost
Bull Call Spreads are 21.1% of everything we have published and 51.4% of everything that lost money. They went 11 and 19. Take them out and the rest of the record is 94 and 18.
| Structure | Wins | Losses | Win rate | Share of all losses |
|---|---|---|---|---|
| Bull Call Spread | 11 | 19 | 36.7% | 51.4% |
| Bull Put Spread | 72 | 11 | 86.7% | 29.7% |
| Bear Call Spread | 22 | 3 | 88.0% | 8.1% |
| Bear Put Spread | 0 | 2 | 0.0% | 5.4% |
| Long Call | 0 | 2 | 0.0% | 5.4% |
The structures where we sell an option won 88.0% and 86.7% of the time over the same period. The one where we buy an option won 36.7%.
What changed, and whether it worked
Use fell away. The structure appears in 9 of the 19 weeks on the ledger, and 21 of its 30 calls sit in those opening 3 weeks.
It helped, and not as much as we would like. Across every week since, the structure was used 9 times and went 1 and 8. That is still 36.4% of the losses on 9.3% of the calls, so the problem was never only that we used it too much.
Two rules came out of this record rather than out of a plan. The generator now refuses a credit spread paying less than a fifth of its width, and a debit structure is no longer allowed to lead a week's document on valuation alone. Neither is scored yet: both arrived after most of the calls above had already settled.
What this study cannot tell you
19 weeks and 142 calls. Enough to see a structure-sized difference, not enough to pin a rate, and the thinnest row in the table has 2 behind it.
The loss share counts losing calls, not money lost, and those are different rankings. A credit spread wins often and gives back more than it made when it loses; a debit spread loses more often for less each time. Counting rows flatters the structure that fails rarely and badly, which is the same direction as the finding. Read 51.4% as half the calls that went wrong, never as half the damage.
Structure is chosen, not assigned. These were written when the week's thesis was directional, so the option mechanics above are one explanation and not a tested one. The structure may be standing in for the weaker half of the selection process rather than causing the losses itself.
Outcomes are graded at settlement against published strikes, with no transaction costs. The ledger's dollar column models each spread's premium rather than recording a fill, and is generous to credit spreads, so this study quotes outcomes and no profit total. 2 of the 142 rows carry strikes read off a real chain, 131 carry strikes worked out from the note's stated offset, and 9 show none, because the reconstruction failed a plausibility check and those rows show no dollar either.
These weeks were mostly a rising market, which is kinder to a bullish book than a falling one. Nothing here is advice.
Frequently asked
Which structure caused most of tickerseer's losses?
Bull Call Spreads. They are 21.1% of the 142 calls on the public ledger and 51.4% of the losses, going 11 and 19. Most of that use falls in the first 3 weeks of the record.
What is tickerseer's track record?
142 options calls published from 2026-05-11 had all settled by 2026-09-18. 105 won and 37 lost. Every call is listed individually on the public ledger with its strikes and outcome, and a published call is never edited or removed.
Did tickerseer stop using it?
Mostly. It appears in 9 of 19 weeks and 21 of its 30 calls are in the opening 3 weeks. The 9 since went 1 and 8.
Are these real trades?
No. These are published calls scored at settlement against the strikes they were published with. No position was opened, and the ledger's dollar column models each spread's premium rather than recording a fill.
How current are these figures?
They are recomputed every week from the same ledger the public track record renders, so this page and that table cannot disagree. The reading above is as of 2026-09-24.