SilentXx｜寂静猎手

RGTI Deep-Water Roll: Took In $84 to Push Assignment from Friday to January

Wed, 14 Oct 2026 00:00:00 GMT

Tonight’s trade was a mandatory question the tape pushed onto my desk.

RGTI has slid six sessions straight, from $15.17 down to $12.12 intraday. The 14.5P expiring 10/16 sits deep in the money — Δ −0.98, almost all intrinsic value, the time value thin enough to ignore. Do nothing, and after Friday’s close it becomes a 200-share assignment ticket.

At 10:03 ET, one BAG order settled it: bought back 14.5P×2 @ $2.34, and sold January-2027 14.0P×2 @ $2.76.

Net cash flow: +$84.

A two-day bomb swapped for a ninety-three-day bomb — and the toll was paid by the bomb itself.

1. The Ledger First

Ledger detail (click to expand)

Open (sold on 10/2)

Date (ET)	Contract	Action	Qty	Avg	Premium
10/2	RGTI OCT 16’26 14.5P	Sell-to-Open	−2	$0.394	+$78.80

Roll (10/14 BAG order, single fill)

Time (ET)	Contract	Action	Qty	Avg
10/14 10:03	RGTI OCT 16’26 14.5P	Buy-to-Close	+2	$2.34
10/14 10:03	RGTI JAN 15’27 14P	Sell-to-Open	−2	$2.76

Settlement (two bases, both real reads)

Basis	Amount	Meaning
Old-leg settlement	−$390.74	Broker realized P&L basis (the 14.5P’s full life)
BAG cash	+$80.98	Account cash $1,020.50 → $1,101.48
Coverage	141%	New premium $552 ÷ old-leg loss $390.74

Both numbers are real and mean different things: one answers “what did this leg lose in total,” the other answers “how much more cash does tonight’s roll leave in the account.” Quote them with their basis attached.

2. First, Cross Out “Take Assignment”: the Account Can’t Hold Those 200 Shares

The first of the three paths to go was “do nothing”:

Assignment would require $2,900 in cash (14.5 × 200), against an account holding only $1,020 — the balance flips negative (−$1,880) and rides on margin;
Post-assignment, the RGTI position becomes 400 shares ≈ 60% of NLV — twice the steady concentration line;
This wasn’t news, either: the 10/6 system check had already flagged “assignment $2,900 > available funds ⚠️”.

In our rulebook, assignment has exactly one identity: an accident. Using assignment to solve assignment upgrades an options problem into an account problem.

3. Of the Three Roads, I Took the One That Pays Cash

Straight buy-back: −$441 to clear the bomb and zero out the obligation — but cash drops to about $580, still below the floor;
Roll to 12P (Jan): pay $142–176 net, in exchange for a lower Δ (−0.39) and a 2.7% buffer — cash buys safety margin;
Roll to 14P (Jan): receive $84 net, downgrade the strike (14.5 → 14.0) and add 91 days — odds buy cash flow.

The cash difference between ③ and ④ is about $230. I chose the one that pays, and wrote its cost out in full: the new leg is still ~15% in the money with Δ −0.58 — assignment remains the base case (roughly a 60% probability), not a tail scenario. This is not defusing a bomb; it’s swapping one: a two-day bomb for a ninety-three-day bomb that came with a small cash rebate.

4. Execution Record: One BAG Order, Buy Before Sell

10:03:22 ET, a single BAG order filled: net +$0.42/contract ($2.76 − $2.34), 2 contracts = +$84;
Why the combo: both legs fill atomically — there is no naked window between “sold first” and “bought later,” and the price locks the moment the order is placed;
Post-fill checks: cash $1,020.50 → $1,101.48; the position-change detector logged it automatically (14.5P gone, 14P JAN on the books).

(As I write this, the new leg quotes $2.72/$2.89 — essentially flat against the $2.76 sale. This roll is currently neither up nor down; what it bought is time.)

5. The Cost, and What Comes Next: Three Months Is Not “Set and Forget”

The cost first, in three lines: assignment odds around 60%, assignment cost $2,800, and margin down only one notch (76.8% → 75.3%, still above the 60% hard gate). So the position gets three management trigger lines:

Leg price ≤ $1.40 (about 50% off the $2.76 sale) → evaluate buying it back and closing;
RGTI reclaims $14+ → let time value decay on its own, don’t fiddle;
Nodes: review before the 11/11 earnings; by 30 days before 1/15 expiry, a decision must be on the table — no improvising on expiry day.

And one line to my future self: if assigned, assignment day is reduction day — handled within two trading days, no averaging down, no adding, no waiting to break even.

6. On the Same Books: NOK’s 10-Lot Trade

One more action sits in the same week’s account: on 10/13 (ET afternoon) I bought NOK JUN 2027 10C ×10 @ $1.87. Its designation is a “trade lot” — a short-term catch that leaves on the bounce, booked separately from the 20-contract base position, with the exit timing in my own hands. It lowered the LEAP average from $2.563 to $2.334; and the process note goes on the ledger as usual: the add happened in a window where margin sat above the thaw line — writing the full record is the least “marketing account” thing about this system.

7. Hunter’s Review

Done right:

Didn’t solve assignment with assignment — cross out the most expensive default outcome first; only then are the remaining options worth comparing;
Clean execution — combo order, mid-price sale, books matching to the cent;
Positive cash-flow direction — between “pay to roll” and “get paid to roll”, the ~$230 stayed in the pocket.

To remember:

One day late cost about $114. The plan said hard deadline 10/13, limit $1.75–1.77 to ensure a fill — it went unexecuted. Over three days RGTI fell another ~5%, and the same buy-back went from ~$1.77 in the plan to $2.34 in the fill. Most of a roll’s bill is written by waiting.
A roll is not a recovery. 141% coverage means “this roll was net-positive,” not “nothing was lost” — the −$390.74 old-leg settlement is in the books and stays there.
Three months is three decisions. The trigger lines aren’t decoration; next time you read this, go check their current values first.

“Roll” sounds like giving it a little more time. On the ledger, it writes a different sentence: cash and time, spent to re-price risk.

This re-pricing came out at +$84 — small, but the direction is right: the account did not have to become something it doesn’t want to be, just because a contract reached its expiry.

Bomb disposal technicians shouldn’t fall in love with bombs. See you in three months.

Journal · part of the quantum-position management series with Two Legs on a Quantum Rally Day, Phased Stock Exit Day 1: RGTI’s 18C Rent Lease and QUBT/RGTI Twin Call Buy-Back.

Evidence: fills and settlement reads from TWS (get_fills: the 10/14 BAG legs at $2.34 / $2.76; reqPnL −$390.74; cash $1,020.50 → $1,101.48 cross-checked against before/after snapshots); account basis and ledger via the trading journal.

NOK Twin-Call Harvest: Down to 20–30% of Premium Left, and No Reason to Wait

Mon, 12 Oct 2026 00:00:00 GMT

Two actions tonight, less than two hours apart.

At 09:53 ET I bought back the NOK 10/23 12C×5 at $0.10; at 11:47, the 12.5C×10 at $0.05. Two limit orders, two clean fills.

These were not new trades — they were early check-outs on two old “leases”. They were rented out on 9/18 and 10/3 (sell-to-open); today the tenants handed back what was left: +$270.13 in total.

1. The Ledger First

Ledger detail (click to expand)

Open (two sales)

Date (ET)	Contract	Action	Qty	Avg
9/18	NOK OCT 23’26 12C	Sell-to-Open	−5	$0.44
10/3	NOK OCT 23’26 12.5C	Sell-to-Open	−10	$0.163

Close (two buy-backs today)

Time (ET)	Contract	Action	Qty	Avg	Cash out (incl. fees)
10/12 09:53	NOK OCT 23’26 12C	Buy-to-Close	+5	$0.10	−$53.71
10/12 11:47	NOK OCT 23’26 12.5C	Buy-to-Close	+10	$0.05	−$55.78

Settlement (broker realized P&L, as read)

Leg	P&L	Premium left	Profit locked in
12C ×5	+$162.82	$0.10 / $0.44 ≈ 23%	77%
12.5C ×10	+$107.31	$0.05 / $0.163 ≈ 31%	69%
Total	+$270.13	—	—

Cash reconciliation: account cash $3,007.43 → $2,953.72 → $2,897.94; total out $109.49, matching fills plus commissions exactly.

2. Why Today: the Tape Delivered the Price, the Ledger Delivered the Answer

NOK fell to $10.08 today (−2.7%), extending last week’s weakness (well below its 20-day average). For anyone holding short calls, that is good news: the 12 and 12.5 strikes drift further away, and the time value evaporates fast.

12C: $0.44 → $0.10 — 23% of the premium left;
12.5C: $0.163 → $0.05 — 31% left.

On the books, 70–80% of the profit is locked in. What remains is the temptation to “wait 11 more days to collect the last 20–30%”.

But those 11 days are not empty: 10/22 is NOK earnings, and 10/23 is expiry. Earnings is a binary event — if it lifts the stock, these two short calls turn from rent collectors into problems to manage; even if it doesn’t, tying up two short legs through earnings week for ~$100 combined buys no safety.

My exit plan was written for 10/15–21 (the pre-earnings window). Today, the conditions arrived early — once the residual value crosses the harvest line, there is no reason to wait. Three days early, the books are settled.

3. Execution: Two Limit Orders, Two Clean Fills

09:53:55 — buy back 12C×5, limit $0.10 — filled;
11:47:39 — buy back 12.5C×10, limit $0.05 — filled.

(System trail, checked while writing: the position-change detector automatically logged both legs’ disappearance right after the fills — every close is now fingerprinted by the system.)

After the buy-backs, NOK’s short legs: only the 11/20 13C×5 remains as the covered call; both 10/23 legs are cleared.

4. Hunter’s Review: Trading “Not Yet Happened” for “Already Banked”

Done right:

Residual-value discipline, again: the same algorithm as September’s NOK harvests — with 70%+ of the profit locked and only 20–30% of premium left, collect; don’t chase the last coin.
Pre-earnings exit, executed early: a window is a range, not a deadline; when the condition arrives first, use it. A binary event avoided at zero cost.
Clean execution: limit orders, two fills, books agreeing to the cent.

Worth noting:

If NOK keeps falling before expiry, an even better price was possible — early harvesting is never the optimum; it is the “exchange uncertainty for certainty” solution. Selling options is a game of probability, not of winning every hand.
With both short legs cleared, the margin reading will move accordingly — figures to be reconciled at close, and reviewed in the next weekly journal.

Waiting 11 days and an earnings event for the last 20–30% of premium — that math needs no hesitation. The selloff offers the cheap price; discipline offers the reason to bank it: what is left on the table matters more than what happened at the table.

Journal · part of the NOK structure-management series with NOK PMCC add, NOK three-tier put ladder, One Buy-Back, One New Lease and A 24-Hour NOK Day Trade.

Evidence: fills and realized P&L read from the broker app (10/12 buy-backs, +$162.82 / +$107.31); open record from TWS fills import (9/18) and user confirmation (10/3); cash basis and ledger via the trading journal.

A 24-Hour NOK Day Trade: Bought the Panic, Sold the Rebound, Handed the Structure Back to Discipline

Fri, 09 Oct 2026 00:00:00 GMT

October 8, US session. AI and tech sold off across the board, and NOK was dragged down with everything else — 4.5% on the day. I went through the headlines; not one of them was about Nokia.

The sector wind lands on every name. At 14:01 I made four clicks: bought NOK JUN 17’27 $10 calls ×10 at an average $2.050.

This was not a position add — it was a day trade. Those 10 contracts had one job: absorb the mispricing, not stay around as base stock. Twenty-four hours later the job was done — just in a way that felt more like a discipline lesson than a victory lap.

1. The Ledger First

Ledger detail (click to expand)

Buy (open the day-trade lot)

Time (ET)	Contract	Action	Qty	Avg	Amount
10/8 14:01	NOK JUN 17’27 10C	Buy-to-Open	+10 (4 fills)	$2.050	−$2,057.45 (all-in)

Sell (reset the structure, next day)

Time (ET)	Contract	Action	Qty	Avg	Amount
10/9 15:16	NOK JUN 17’27 10C	Sell-to-Close	−10 (2 fills)	$2.240	+$2,232.50 (net)

Two bases, both true

Basis	Amount	Meaning
Cash basis	+$175.05	Cash-flow difference of the pair ($190 gross − $14.95 fees)
Settlement basis	−$330.72	Sale at $2.24 vs position average of $2.563 (TWS reqPnL read)

One number answers “what did this 24 hours earn in cash”; the other answers “how does this close settle in the books”. Journalists, cite the basis.

2. Why Buy: A Sector Mispricing, and a Gap Only Panic Opens

Before pressing the keys, I checked three things:

The selloff was sector-wide, not company-specific. AI/tech dragged everything down; no fresh negative on Nokia itself. The AI-infrastructure narrative was intact, and the October 22 earnings were still two weeks out — no binary event in between that had to be avoided.
The price was genuinely cheap. $2.050 against a 20-contract LEAP average of $2.816 — 27% below it. Ten more contracts pulled the average straight from 2.816 to 2.563.
It had a return ticket from day one. The mandate was clean: a day-trade lot. Take the rebound, or take the stop. It had no license to become a habit at 70% concentration.

3. The Risk Plan: The Real Question Wasn’t NOK — It Was the Account

Right after the fill, the account readouts put a different question on the table — not about Nokia, but about the whole book:

Metric	Reading after the add	Line
Cash	$774.93	≥ $2,000 floor
Maintenance margin	69.9%	≤ 50% thaw line
LEAP concentration	≈70% of NAV	≤ 30%

Thirty LEAPs = seventy percent of net assets, with the October 22 earnings under the hood. That is not a structure you “hold and see.”

One more thing for the record: this add went in while the system was refusing to clear it — good timing does not make a good process. The reckoning is in section 5.

4. Execution: 15:16, Handing the Structure Back

On October 9 the rebound arrived: NOK +3.0% to $10.45, and the new lot was up 12% in a day (+$250). The cash audit closed to the cent: $2,832.38 → $774.93, the $2,057.45 gap exactly matching fills plus fees.

Then, at 15:16, I pressed sell — −10 contracts @ $2.240. Two reasons:

The exit condition was met. Twelve percent in a day: the mispricing repair was paid. Whatever comes next belongs to the base — the twenty contracts that stayed.
The risk rules outrank the trade. All three lines were crossed: the cash buffer was too thin for any surprise, margin was closing on forced-liquidation territory (a modeled NOK −20% would do it), concentration sat at 70%. Holding on would not have been a bet on NOK’s direction — it would have been a bet on the account’s life.

After the sell-back, the structure was back in line:

Metric	After the add	After the sell-back
NOK LEAPs	30 (≈70%)	20 (≈47%)
Cash	$774.93	$3,007.43
Maintenance margin	69.9%	55.9%
Leverage	1.04x	0.80x

5. Hunter’s Review: Entries Are Given, Exits Are Demanded

What went right: the read (sector mispricing, no company event), the execution (books tied to the cent), the exit (condition met, no hesitation habit).

But the line worth remembering is the one from section 3: this lot went in while the system was refusing to clear it. What the system was blocking was the structure — thirty LEAPs, seventy percent of NAV, earnings on the other side — not the $2.05 price. Selling back was the remedy; the correct defence line sits before the buy. Next day trade: through the gate first, then press the key.

In 24 hours this trade made $175 in cash and settled $330 of loss in the books — both numbers are true, the way a day trade always is: it gives you an answer and a bill.

The real result isn’t in the numbers though. The 30-contract gambling table was carried out, and the cash buffer, the margin, the concentration all came back in line. A hunter does not live by every bullet hitting the bullseye — a hunter lives by the fact that when the rifle comes down, the structure is still intact.

Trade journal · part of the NOK structure-management series with the PMCC expansion, the three-tier put sale, and the big-up day close-and-open.

Evidence: fills and settlement come from TWS reads (four buys on 10/8, two sells on 10/9, reqPnL −$330.72); account basis and the full ledger live on the Trading Journal.

One Close, One Open: NOK 10P Banked +$107.94 as 13C Completes the 20/20 Lease

Tue, 06 Oct 2026 00:00:00 GMT

October 6, US Eastern — a big-up day for Nokia. Two clicks, 32 minutes apart.

11:17: NOK Oct 9 $10 puts ×5 bought back at $0.04. 11:49: NOK Nov 20 $13 calls ×5 sold at $0.32.

One leg came from the second tier of the September 22 three-tier put sale — a residual down to 15%, and a $5,000 assignment obligation priced too cheap to keep carrying. The other is a brand-new lease: strike 13, expiry 11/20, straddling the October 22 earnings — the last piece of the lease board on 20 LEAPs.

From today, all 20 LEAPs are working. Every base unit is producing cash flow.

1. The Ledger First

Ledger detail (click to expand)

Trade one: the second tier closes (buyback)

Time (ET)	Contract	Action	Qty	Sold avg	Buyback	Gross	Net banked
11:17	NOK OCT 9 ‘26 10P	Buyback	5	$0.263	$0.04	+$111.50	+$107.94

Basis: net banked per TWS settlement (all-in), same source as the October 1 buybacks. Residual at 15.2% ($0.04 ÷ $0.263); 82.1% of the $131.50 sold premium realized.

Trade two: a new lease at the 13 strike (sell-to-open)

Time (ET)	Contract	Action	Qty	Sold	Net collected	Expiry	Parameters
11:49	NOK NOV 20 ‘26 13C	Sell-to-open	5	$0.32	+$156.20	11/20	Δ 0.245 · IV 65.6% · OTM +21%

Both trades hit during the NYSE session: Nokia up over 5% on the day ($10.76 against a $10.20 prior close), the rally continuing to price the company’s AI-infrastructure narrative.

2. Trade One: The Comparison at a 15% Residual

Three lease signboards went up on September 22 — the first (OCT 2) redeemed last Thursday, the second (OCT 9) taken down today, the third (9.5P) still on the wall waiting for earnings (handling window 10/16–21).

Why now? Three reasons, all pointing the same way:

The residual comparison came due again. 0.263 → 0.04: the entire remaining potential is $20 (500 shares × $0.04). A seller realizes profit not by “waiting it out,” but through one comparison: is the residual still worth the risk budget it occupies? Two weeks ago this was a $131 business; today it’s $20 of loose change.
The rally pushed the answer even further away. NOK up over 5% puts the $10 strike 7% below spot — the leg is now safe enough to be boring, which means its story is finished: the full theta rent has been collected; the last three days hold only pocket change.
The account is in reduce-only mode. $5,000 of assignment obligation (500 shares × $10) came back off the table, and the maintenance-margin reading eased from 67.9% pre-market to 52.7% midday. Under the “unfreeze only below 50%” rule, every leg retired is a step toward thaw.

3. Trade Two: Why Sell Calls Into a Rally

Selling calls on a big-up day is a rare but comfortable posture for a seller:

IV up = rent up. The 13C went out at roughly 66% implied vol — a big-up day’s premium contains the market’s full mood premium. That is what the seller is paid for.
The strike can sit higher. The new leg lifts the short-call strike ladder from 12 / 12.5 to 13 — the first rung above the breakeven line (LEAP strike $10 + net cost $2.82 ≈ $12.82). NOK must gain 21% to touch it, with the October 22 earnings sitting in between.
The final five complete the board. Of 20 LEAPs, 15 were already leased (12C×5 + 12.5C×10). This trade put the last 5 units of capacity on the wall — 20 short calls against 20 LEAPs, fully leased. From today, new short calls are frozen: beyond full coverage, anything added is naked. That line is not a slogan; it is the discipline.

45 days for $156.20, roughly 20% annualized — unspectacular, but it is incremental rent on base units that were already paid for.

4. The Board: Fully Leased

The full Nokia structure now reads:

Leg	Expiry	Qty	Role
LEAP 10C JUN ‘27	2027-06	+20	Base (cost $2.816)
12C OCT 23	10/23	-5	Collecting rent
12.5C OCT 23	10/23	-10	Collecting rent
13C NOV 20	11/20	-5	This entry
9.5P OCT 23	10/23	-5	Last put tier, handle before earnings

The calendar items that require attendance:

10/16–21: pre-earnings handling window — review the 9.5P together with the fifteen 10/23 short calls;
10/22: Q3 earnings (pre-market);
11/20: 13C expiry — if the tape hasn’t taken it away, another $156 settlement day.

Account level: cash $2,832, leverage 0.85x, maintenance margin 52.7% — still above the 50% thaw line. Right direction; the discipline continues.

A seller’s log rarely has dramatic entries. More often it looks like this: late night, one old lease down to a 15% residual, redeemed; the same day, a new lease goes up — the price was handed over by the rally, the strike was chosen by discipline.

Three signboards went up on September 22: the second comes down today, the first already came down last week, the third waits in the wind for its earnings. And on the other side of the book, all 20 units are now on shift — as of today, not one of the 20 LEAPs is idle.

Trade journal · part of the NOK short-leg management series with the three-tier put sale, the 11C buyback, and the NFP-night double buyback.

Two Legs Pulled Before NFP Night: NOK 10P and RGTI 18C Bought Back, +$129.78 Banked

Thu, 01 Oct 2026 00:00:00 GMT

October 1, US Eastern. Two buyback clicks.

10:39 AM ET: NOK Oct 2 $10 puts ×5, filled at $0.05. 11:34: RGTI Oct 9 $18 calls ×2, filled at $0.10. Combined net banked: +$129.78.

One leg came from the first tier of the September 22 three-tier put sale (expiry 10/2). The other was the covered call sitting on the RGTI shares (expiry 10/9 — and its own “T-7 review line” also lands on 10/2). The board in front of NFP night got cleared in one pass.

1. The Ledger First

Ledger detail (click to expand)

Time (ET)	Contract	Action	Qty	Sold avg	Buyback	All-in cost	Realized
10:39	NOK OCT 2 ‘26 10P	Buyback	5	$0.18	$0.05	-$27.36	+$58.64
11:34	RGTI OCT 9 ‘26 18C	Buyback	2	$0.47	$0.10	-$21.49	+$71.14
					Total	-$48.85	+$129.78

Basis: both per TWS settlement (all-in, both sides); combined commissions & fees of $9.22 derived from the settlement figures — down to the penny. The RGTI leg was opened 9/17 via two single-contract fills at $0.47 (PHLX).

The full life of each leg:

Item	NOK 10P Oct 2	RGTI 18C Oct 9
Sold / bought back	$0.18 → $0.05	$0.47 → $0.10
Gross profit	+$65.00	+$74.00
Net banked	+$58.64 (65.2% of $90 nominal)	+$71.14 (75.7% of $94 nominal)
Buyback / sale price	27.8%	21.3%

2. Why Now: Taking Tomorrow’s Answers Off the Table

Both residuals had thinned to 27.8% and 21.3%. The entire remaining potential is $25 and $20 respectively. A seller realizes profit not by “waiting it out,” but through one comparison: is the residual still worth the risk budget it occupies? Both legs answered “no.”
The calendar stacked both review points onto 10/2. NOK expires on NFP night itself (8:30 AM ET / 20:30 Beijing); the RGTI call’s T-7 review line lands on the same date. And the tape already ran one rehearsal: hours before the buyback, NOK dipped to $9.95 — one layer of sentiment off the $10 strike — before rallying back above $10.3. Rather than babysit two answers tomorrow, take both off the table tonight.
Each buyback handed something back. The NOK leg released about $1,378 of maintenance margin (TWS whatIf); the RGTI covered call held no extra option margin (stated for honesty) — what it released was different: discretion over the shares. With the call gone, the 200 RGTI shares are a free hand again. After both, the account’s maintenance margin reading sits at $6,095, roughly 63% of net liq — right direction, still above the 50% safety line; reduce-only discipline continues (that honesty belongs to the account itself).

3. The $48.85 Price of Leaving Early

Combined all-in buyback cost: $48.85 ($27.36 + $21.49) — that is the distance from “theoretical max” to “banked today”: NOK gave up $27.36 ($25 residual + $2.36 fees); RGTI gave up $21.49 ($20 residual + $1.49 fees).

$48.85 buys three things: $1,378 of margin freed a day early, one data release that no longer needs attending, and full discretion over 200 shares. The bill is mine to pay — because the more expensive scripts (NOK opening south of $10, or a one-night quantum squeeze) were never mine to write.

4. The Board: One Leg Closed, One Stock Unshackled

Three lease signboards went up on September 22. The first one has officially been taken down:

Tier	Contract	Status
First	10P OCT 2 ×5	✅ Sold $0.18 → bought back $0.05, banked +$58.64
Second	10P OCT 9 ×5	Collecting rent — next decision point 10/9
Third	9.5P OCT 23 ×5	Straddles 10/22 earnings → handle before earnings

The full Nokia board now reads:

Leg	Expiry	Status
10P OCT 9 ×5	10/9	Collecting rent, next decision point
9.5P OCT 23 ×5	10/23	Straddles 10/22 earnings → handle before it
12C OCT 23 ×5	10/23	Earnings-straddling call, management line 10/15
LEAP 10C JUN’27 ×20	2027-06	Base position, untouched

RGTI, meanwhile, is back to a pure stock position (200 shares): no obligations overhead — and the next questions, whether to re-sell calls for the next cycle and when the trim condition triggers, go back to price doing the talking.

By the morning of October 2, there is not a single leg in the account waiting on an answer. When the NFP number lands, neither Nokia nor RGTI is on my to-do list; the next card sits on the October 9 page of the calendar.

A seller’s log rarely has dramatic entries. More often it looks like this: late night, prices settled, residuals thinned out, two clicks — and two half-written leases redeemed ahead of term.

$48.85 paid: $1,378 of margin freed, one data release that no longer needs attending, and a stock handed back its full discretion. The board is clean — next hand, on the calendar.

Trade journal · part of the September NOK short-leg series with the three-leg put sale, the 11C buyback, and the crash-day double buyback.

The Three Switches of Option Selling: Can an Options Cash-Flow System Be Stable Income?

Thu, 01 Oct 2026 00:00:00 GMT

The Problem

A few days ago I came across a post on Xiaohongshu titled “Selling Options for Stable Income?”

The post was short: friends of the author reportedly earned regular, stable income from options — some of them no longer needed their salaries and quit their jobs. “Is it really that good? Which stocks do you all trade? Any tax issues?”

The post had 100 comments.

I am in no position to laugh at it. Because for the past few months, I have been asking myself the same question — just in different words. I built an options cash-flow system, ran it for most of a year, watched the ledger grow thicker. And still, some late nights, a whisper: can this system really earn “stable income”?

So this time I stopped guessing and stopped listening to slogans. I gathered every piece of evidence I could find and put it all on the same trial table:

Global evidence: academic papers, forty years of CBOE index records, third-party backtests, the full ledger of real products — and the most famous “autopsy files” of the past three decades;
Human samples: the 100 comments under that post, top ones read line by line;
Mechanism analysis: why it looks stable has to be taken apart at the mechanical level;
My own ledger: 143 live options trades and 48 stock trades, audited one by one.

This article is the record of that trial. It may be the most important thing I have ever written — because it answers, not “how to make money with options,” but “are we actually built to live off it.”

The short version: Yes — but what is stable is not the cash flow; it is the system and the discipline. It behaves like an insurance business: regular income, irregular losses. The three switches — collateral, premium, management — must all be on. Miss one and everything earned earlier gets given back.

Layer One: Global Evidence — Do Sellers Really Have an Edge?

Start with the most fundamental question: as a group, are options sellers profitable over the long run?

The answer: Yes — with a long list of conditions attached.

First, academia. In 2003, Bakshi and Kapadia showed in the Review of Financial Studies that delta-hedged option portfolios systematically underperform the risk-free rate. In plain words: options, on average, are priced richer than they “should” be. The excess is called the volatility risk premium.

CBOE’s own numbers are even more intuitive: from 1990 to 2018, the average implied volatility on the S&P 500 was 19.3%, while realized volatility was only 15.1%. That four-point gap is what buyers pay extra for “insurance” — and it flows steadily into sellers’ pockets.

This is not mysticism. It is the mathematics of the insurance business. An insurer cannot profit on every policy, but premiums exceed payouts over time.

The Forty-Year Exam: The PUT Index

If you allow me only one piece of evidence, I choose this: the CBOE S&P 500 PutWrite Index (PUT). Since 1986 it has sold one at-the-money put every month, fully collateralized, mechanically executed, with zero human judgment.

Through August 2023, its report card:

	PUT Index (put selling)	S&P 500
Annualized return	9.40%	9.91%
Annualized volatility	10.3%	15.4%
Max drawdown	-32.7%	-51.0%
Drawdown recovery	30 months	66 months

Same returns, about two-thirds of the risk. In years when the S&P fell, it outperformed 95% of the time.

There is a starker comparison: running the exact same index options monthly, the put-selling index compounded +1835% while the put-buying index compounded +708% (1986-2018). One market, mirrored operations — and the seller made more than twice as much as the buyer.

The price is written into the distribution: skewness -2.72, kurtosis 13 (a normal distribution is 0 and 3). In human terms: small wins most of the time, a big loss now and then. In 1987, 2008 and 2020 it took heavy beatings — but it never died. Because it is fully collateralized, nobody can ever force it to liquidate at the worst possible moment.

The Strategy Level and the Product Level

Forty years is too long for details. We checked two more layers.

Strategy level — a wheel backtest (2018-2025, covering the COVID crash and the 2022 bear market): 11.2% annualized, slightly behind buy-and-hold’s 12.8%. But max drawdown -18.4% versus -33.7%; in 2022, -2.1% versus -19.4%. In a bull market it structurally loses by 3-5 points a year; in choppy markets it leads by roughly 8%; bear markets are its home turf.

Product level — a real product’s complete ledger: The most famous covered call ETF, “QYLD,” pays a 12% distribution rate. Sounds gorgeous. But its 12.5-year record: 8.7% annualized total return, against 19.9% for the Nasdaq 100 it writes options on. Its share price drifted from $25 to $18 — and in one month, 99% of its “distribution” was classified as return of capital.

This is not a scam. It is the textbook case of distribution rate ≠ total return: the monthly cash is real, but it was bought with long-term upside.

When You Sell Matters More Than What You Sell

A 20-year study by tastytrade found the key condition: when IV Rank is above 30, selling options has positive expectancy at roughly a 3:1 reward-to-risk ratio; when IV Rank is below 20, expectancy turns negative. Selling insurance in calm markets is a losing business.

Backtests from institutional desks in China reach the same conclusion: the identical selling strategy returns -13% annualized when IV is at 10%, and +185% when IV is at 30%. There is exactly one engine: the gap between implied and realized volatility.

The Autopsy Files: Nobody Died from “Selling Options”

Now the scary part — isn’t selling options how accounts blow up? We opened decades of case files:

OptionSellers.com (2018): Cordier, a man who wrote a book teaching option selling, had twenty “stable” years. Then natural gas spiked 18% in a day. Clients lost $150 million — and owed money on top. Cause of death: naked selling + no protection.
The LJM “Preservation” Fund (2018): A fund with “preservation” in its name sold naked S&P puts with leverage. The VIX doubled from 17 to 37 in a day; two days later the fund was down 80%; $650 million evaporated. Cause of death: leverage.
Karen “the Supertrader” (2015-16): A social-media star who publicly claimed 25-30% annualized — in reality she sold naked strangles with heavy leverage, rolled over $50 million-plus of losses for 18 months, and ended up with an SEC fraud complaint. Cause of death: leverage + hiding losses.

Three case files. Not one died from the act of selling options. All three died from the same thing: losing the power to choose at the exact moment it mattered most.

Layer Two: The 100 Comments — Real Humans

Back to the post. I read the top comments one by one. Three camps:

The doomsayers (holding the top-voted spots):

“I’ve sold puts for years. Tax-wise it doesn’t pay, and the tail risk is real. I’m rethinking my approach.” — The most experienced voice in the thread was the most cautious.
“Sell options long enough and within two or three years you’ll find it doesn’t really make money at all — you’d be better off just buying stocks.”
“Clearly someone who’s never lived through a crash. You make small money for years, then one market shock gives it all back and worse.”

The hands-on traders (a minority, seriously posting):

“My goal is $1,000 a week. I don’t have cash for shares, but I have tens of thousands in buying power, so taking on $100k-200k assignments should be safe enough.” — The author replied: “A thousand a week is already amazing.”

The bystanders: “How much income?” “Could you share more details?”

One comment was the sharpest, and it deserves an answer — “Can you beat the market?” Of 100 comments, nobody answered it directly.

And the most striking number of all: 100 comments, and zero about risk rules. No position sizing, no collateral structure, no drawdown limits. Everyone asked what to buy, what to sell, and how much they’d make.

That gap is the real distance between retail traders and the market.

Layer Three: The Mechanism — Why Does It Look Stable?

The illusion of stability deserves to be taken apart mechanically:

Premium is certain; profit is not. Every time you open a position, cash flow +1 immediately. But the real result is only revealed at close, expiry, or assignment. High-frequency small credits against low-frequency large debits — and the account curve looks “regular,” even stable.

The correct scorecard is:

Long-term net result = premiums + closing P&L − commissions − slippage − financing cost − taxes

Then look at max drawdown, recovery time, worst months — not at “how much premium I collected this month.”

One number to remember the difference: QYLD distributes at a 12% rate, and its real 12.5-year total return is 8.7%. The monthly joy fades the moment you look at total wealth.

Layer Four: My Own Ledger — 143 Trades on Trial

Evidence, comments, mechanics — all of it belongs to others. Time to open my own books.

In the first nine months of this year, my options cash-flow engine ran 143 trades:

Options: +$2,048 (averaging about +$14 per trade; six of eight tickers positive);
Directional stock trades: -$4,064 (48 trades; seven of eight tickers negative).

Same tickers: the options made money, the stocks lost it. The rent collected was eaten by the “house purchases.”

This is not a strategy failing — it is the proof that the strategy works. The engine produces about +$228 a month, a machine with positive expectancy. What actually ate the profits were operations outside the engine: directional bets, leverage, and holding losers.

On the strictest basis (net liquidation value minus cumulative deposits): at the latest snapshot, five years of deposits total $10,989 against an account value of $9,891 (the September 30 snapshot) — $1,098 from being back in the green across all history. At the engine’s pace, that is four or five more months.

So when people ask “is the system even profitable?” — my own data answers: the system is profitable; what lost money was not the system.

The Verdict: Three Switches

Put the four layers together and the conclusion surfaces. It is not “yes” or “no.” It is three switches — all three on, or everything earned earlier gets given back:

Switch 1 · Collateral. Fully collateralized, no naked selling, no leverage to inflate your “salary.” Every blowup story died here; every surviving system passed here. My red lines: leverage ≤ 1.5x, hard stop 2.0x.

Switch 2 · Premium. Sell insurance only when volatility is expensive. When IV Rank is high and the volatility premium is thick, expectancy is positive; force-selling in calm markets is donating money. My gates: IV Rank ≥ 60, or VRP ≥ 1.15.

Switch 3 · Management. Management beats stock selection. What doubled results in the backtests was not which stock was chosen, but disciplines like “never sell a call below cost basis on a losing position.” Manage actively at 21 DTE, take profits at 60-70%, and treat assignment as an incident (exit within two days). Details decide survival.

And three footnotes:

“Can it beat the market?” — Depends on the scorecard: it loses to the bull (structurally, 3-5 points a year) and wins at drawdowns (outperforming in 95% of down years). If your benchmark is risk-adjusted return at equal drawdown, it wins. If you only compare who rises more, it loses by design.

“My friend quit his job.” — That is not evidence. Evidence is “my friend quit, then lived through a full bull-bear cycle and is still here.” The bar for quitting is not “this month’s premium ≈ my salary.” It is: capital large enough (a 30-50% drawdown does not change your life), a track record long enough (3-5 years, full cycle), and no bleeding outside the system.

“What does ‘stable’ even mean?” — What is stable is not the cash flow; it is the system and the discipline. Income is regular; losses are not. Small wins are the norm; big losses are mandatory questions on the exam. Every genuinely stable system ends up the same way: an edge in probability, plus a margin of survival.

Closing

As I write this, my account is $1,098 away from being green across all five years. Five years, one full loop of “tuition → transformation → verification” — enough, perhaps, to say a few true things.

The post’s question — “is it really that good?” — is best answered this way: No, it is not that good. But it can be solid. It will not make you rich overnight, and it will not ruin you overnight either (as long as you don’t sabotage yourself). It is closer to a small business you run with care: buy inventory (take risk), collect premiums (sell insurance), pay claims (accept big losses) — and outlive your competitors through discipline.

I have pinned the three switches from this study into my system: into my knowledge base, into my skill rules, and onto the screen I open every day.

A note to myself five years from now: You do not make money by predicting the market. You make money by living long enough.

Addendum (2026-10-13): Figures in this article have been re-verified against broker trade-level settlement records. Options: 121 trades / +$2,162 corrected to 143 trades / +$2,048; directional stock trades: 41 / -$2,875 corrected to 48 / -$4,064 (assignment and expiry records now included); five-year deposits $10,203 corrected to $10,989 (a September wire was booked late); the distance from all-time breakeven corrected from $312 to $1,098. The conclusion stands: the system is profitable; what lost money was not the system.

Sources and References

Sources and references (feel free to verify):

Bakshi & Kapadia, Delta-Hedged Gains and the Negative Market Volatility Risk Premium, Review of Financial Studies, 2003 — paper
Cboe, Generating Income and Managing Risk: Cash-Secured Put Writing (PUT Index 1986–2023) — cboe.com
Cboe (Bondarenko), Historical Performance of Put-Writing Strategies (+1835% vs +708%) — cboe.com
EnnisKnupp, Evaluating the Performance Characteristics of the Cboe S&P 500 PutWrite Index — PDF; Neuberger Berman, Simply PutWriting — PDF
OptionsPilot, Wheel Strategy Backtesting Results — link; OptionsBrainiac, Wheel Backtest Research — link
Are Covered-Call ETFs Like JEPI Worth It? (QYLD total return and return-of-capital) — link; The Options Bench, QYLD 12-year covered-call data — link
tastylive, Selling Options Yields Consistent Success / The Expectancy of Trading Strangles (IV Rank evidence) — link
Shenwan Hongyuan / Bohai Securities research on option-selling strategies (IV sensitivity and position control)
CNBC / BBC / Forbes (2018, OptionSellers.com / Cordier); SteadyOptions (LJM); TheStreet and SEC filings (Karen Bruton, 2015)
de Silva, Smith & So, Losing is Optional (retail option-buyer losses) — PDF
Original Xiaohongshu post “Selling Options for Stable Income?” comment section, Sep 2026 (read line by line)
This article’s personal ledger: IBKR trade-level settlement records, 2026-01-01 to 09-30, re-verified before publication

The discussion thread for this article is on Tiger Community — Can an options cash-flow system really earn stable income?

Trading Insights #010 · To be continued

NOK 11C Buyback: $0.03 Never Filled, $0.04 Did — +$114.80 Banked

Mon, 28 Sep 2026 00:00:00 GMT

On September 28, at 11:15 US Eastern — 23:15 Beijing — I bought back the last 10 contracts of my NOK Oct 2 $11 call: filled at $0.04, $43.20 all-in including fees.

No dramatic tape, just a quiet closing: this old PMCC top leg (Group B) ended up leaving +$114.80 in the account — and removed a $11,000 notional obligation from the books of a RED-status account.

1. The Ledger First

Ledger detail (click to expand)

Time (Beijing)	Action	Contract	Qty	Price	Cash Flow
22:03	Limit buyback (DAY)	NOK OCT 2 ‘26 11C	10	$0.03 (no fill)	—
23:15	Buyback filled	NOK OCT 2 ‘26 11C	10	$0.04	-$40.00
	Commission & fees				-$3.20
				All-in total	-$43.20

And the full life of this leg, sell to buy:

Item	Value
Sold (opened)	avg $0.158 × 10 = +$158.00
Bought back (closed)	all-in -$43.20
Net banked	+$114.80 (kept 72.7% of the maximum gain)

Basis: both sides per TWS settlement figures; every number recomputed locally, no mental math.

2. What Happened That Night: $0.03 to $0.04

At 22:03 Beijing I placed the order: BUY NOK OCT 2 ‘26 11C ×10, limit $0.03, DAY — a one-cent bargain under the then-mark of $0.04.

The night passed; no seller showed up. At 23:15, the fill came at $0.04. With $3.20 in commission and exchange fees, the buyback cost $43.20 all-in.

Hold out a whole night over one cent? Not worth it. A bargain price is a bonus; getting filled is discipline — this order had exactly one KPI: take the obligation off the books and cross the task off the calendar. Look at the trajectory: the pre-market plan estimated ~$100 to clean this leg up; the actual cost came to $43.20. The market had already handed over plenty of discount — the last cent, let it go.

3. Why “Now”: The Right Move in a RED Account

This buyback hit three lines at once:

The account is RED — only two kinds of moves are allowed: reduce obligations, reduce risk. This was the former: an $11,000-notional potential delivery obligation, cleared for $43.20. Afterward the maintenance margin reading eased from 79.2% to 76.28% — the direction is right, but the account is still above the 50% line, still RED (honesty stays with the account itself).
Only a quarter of the value was left on the table. Sold at avg $0.158, bought back at $0.04 — 25% residual. Hang on four more trading days plus an NFP countdown for that last $40 of maximum remaining profit? No. A seller’s profit never comes from “waiting it out”; it comes from one comparison: is the residual worth the risk that remains?
Decluttering the calendar. October 2 was its original expiry — also the September NFP release night (20:30 Beijing). Retire this old lease early, and the Oct 2 board holds only 10P ×5 — reducing the decision surface is itself a return.

4. Coverage Check (never skipped)

Before: LEAP 10C JUN’27 20 contracts ≥ short calls 15 → fully covered ✅
After: LEAP 20 ≥ short calls 5 → +15 headroom, no naked ✅

The buyback didn’t weaken the PMCC skeleton — it just peeled off the oldest layer of lease on top. Same skeleton, lighter shoulders.

5. The Board: Nokia’s Next Node

Leg	Expiry	Status
10P OCT 2 ×5	10/2	~+4% buffer, watch $10 (NFP night)
10P OCT 9 ×5	10/9	Collecting rent
9.5P OCT 23 ×5	10/23	Straddles 10/22 earnings → handle before it
12C OCT 23 ×5	10/23	Earnings-straddling call, management line 10/15
LEAP 10C JUN’27 ×20	2027-06	Base position, untouched

The next decision point is 10/2: NFP plus the 10P ×5 expiry. And when that night comes, the board will no longer carry the name 11C.

The most valuable thing a seller does rarely looks dramatic: not the thrill of opening a position, not the pride of squeezing out the last cent — but quietly returning to the market a lease no longer worth managing.

The rent is small; the discipline is large. With the obligation cleared, the next hand starts on a cleaner table.

Trade journal · part of the September NOK short-leg series with the 10C roll and the three-leg put sale.

NOK Three-Leg Put Sale: $393.50 in Rent, and a Ledger That Must Be Managed

Tue, 22 Sep 2026 00:00:00 GMT

Mid-session on September 22 (US Eastern) — just past midnight, September 23, Beijing time (00:03–00:26).

Late at night, I hit the sell button three times on NOK — three tranches of puts, 5 contracts each, 15 in total.

NOK is the name that shows up most in my account: a batch of LEAPs, two sets of covered calls, and several rounds of puts sold and bought back. Tonight’s action was putting its volatility up for rent.

1. The Ledger First

Ledger detail (click to expand)

Time (ET)	Contract	Action	Qty	Fill	Cash Flow
12:03	NOK OCT 9 ‘26 10P	Sell	5	$0.27	+$135
12:13	NOK OCT 2 ‘26 10P	Sell	5	$0.18	+$90
12:26	NOK OCT 23 ‘26 9.5P	Sell	5	$0.36	+$180
				Net (fees included)	+$393.50

Note: six fills grossed $405, net $393.50 (fees of $11.50 deducted — TWS settlement basis).

Here is the board as it looked when I sold:

Where the stock stood: NOK had just completed a full wild swing — closing at $11.13 on 9/11, crashing -13.3% in a single day on 9/14 (close $9.65), then grinding back to $10.94 by 9/21. Midday on 9/22, the tape was around $10.5–10.6;
Implied volatility: 56%–66% across the three strikes — that is the “rent price” a high-volatility window puts on the table;
Buffer: the two 10P strikes sat about 5% below the tape, the 9.5P about 10%;
Effective assignment cost (if exercised): 9.82 / 9.73 / 9.14 — cheaper than any price I saw in September.

2. Why These Three Tranches

Selling a put is writing the market a promise: “I will buy here,” and collecting rent for it. Two questions decide whether to write it: Am I willing to take the shares, and is the rent worth writing?

Willing to take delivery: NOK is a core name I track long term — the thesis behind the LEAP position (AI optical networking, the carrier and defense base) has not changed. If it really trades down to 9.5–10, that is not an accident for me — it is within discussion range for restocking.
The rent is worth it: ±10% daily swings pushed IV to 56%–66%. The market is charging a separate premium for NOK’s fear — that is the window a seller collects in.
Three staggered expiries (10/2, 10/9, 10/23): spread the possible assignment dates out, instead of stacking all 15 contracts on a single expiry.

Boundary condition: these three reasons only hold for volatility. If NOK’s decline turns out to be fundamentally driven news (guidance cut, orders collapsing), the promise is void — buying back or rolling then is not “restocking at a discount,” it is incident management. That line gets written on day one.

3. Do the “What If I Get Assigned” Math First

The part of a put sale people skip is never the premium — it is the assignment column:

All 15 contracts exercised = buying 1,500 shares of NOK, requiring $14,750;
Cash in the account: $3,762 (25.5% coverage).

In other words: the “what if” on this table is not something the account can casually absorb.

A fair note here: the three tranches expire at different dates — $14,750 is the extreme, cumulative case where all three get assigned. The closer-to-reality read: 500 shares per tranche, three expiries, three independent decision points — and after any single assignment there is still room to act (hold for rent, sell, or roll). This is not a bet on swallowing 1,500 shares in one gulp; it is a promise list managed by expiry date.

That fact defines this position, from the moment it was opened, as one that must be actively managed — not left alone, not bet on “it probably won’t drop that far.”

4. Rent Isn’t Free: The Margin Bill

At the same moment the $393.50 landed, the account’s maintenance margin went from 30.4% to 63.4% ($2,997 → $6,258) — short puts occupy margin by rule; that is the bill, not an accident.

In plain terms: the price of collecting $393.50 in rent is that the account’s risk budget is now half-reserved by these 15 puts. That is why management priority for this position sits above its income contribution — written down here as a memo to myself.

5. The Management Calendar (Written on Day One)

Date	Action
10/2	First tranche expires. Above $10 → rent earned, evaluate a re-sell; near $10 → buy back or roll early
10/9	Second tranche expires — same logic
10/22	NOK Q3 earnings. The 9.5P expires on 10/23, after the report — by my rules, a short leg spanning earnings either gets handled before the print, or accepts gap risk explicitly. Default: handle it before
10/23	Third tranche expires
Anytime	Fat profit (60–70%+) → buy back into strength (the 9/17 muscle memory); near the strike → decide actively, never gamble on expiry day

In one line: the rent from these three tranches is a promise written to the market — and every expiry node of that promise needs someone in the room.

That someone is me.

The rent is small; the discipline is big. The final score of this trade will not be decided by whether NOK goes up or down next — but by whether that calendar above gets executed.

Whatever comes next — buyback, roll, or assignment — will get its own entry.

TRON Final Exit: All 2,300 Shares Sold — The Inventory Era Ends, Capital Moves Fully into Options

Mon, 21 Sep 2026 00:00:00 GMT

Midday, September 21, US time. 00:14, September 22, Beijing time.

One sell order fills: TRON, 2,300 shares, $1.542 — the entire position, sold.

Fourteen months, counting from the first buy in July 2025. As of this moment, the number is zero.

The “TRON daily monitor” line in my pre-market scans can finally be deleted. In one sentence: TRON and I are officially even — goodbye, for good.

But this is not a farewell letter. It’s a closing report — because what this trade means reaches far beyond 2,300 shares.

I. The Ledger First

Ledger detail (click to expand)

Three fills tonight, two battlefields:

Time (Beijing)	Instrument	Action	Size	Price	Reference	P&L
Evening	NOK 10P OCT2	Buy back	2 contracts	$0.15	$0.463 sell avg	+$61.13
00:06	NOK 10.5P SEP25	Buy back	4 contracts	$0.12	$0.218 sell avg	+$34.34
00:14	TRON stock	Sell	2,300 shares	$1.542	$2.0839 cost	-$1,246.43

Basis notes: the two NOK entries are TWS settled values (fees included); TRON is a gross loss against average cost (commission excluded). Combined TWS realized for the US session (9/21): -$1,225.20.

The two NOK puts pocketed +$95.47 combined — the premium-selling machine has been quietly paying out on NOK all along.

On the TRON side, a gross loss of -$1,246.43 (-26.0%).

All three combined, tonight’s net is about -$1,150. That is the final invoice of a fourteen-month inventory war.

Now the part that matters more than any single trade — the account-level shift:

Metric	Before	After
Cash	-$80.78	+$3,368.45
Leverage	1.16x	0.82x
Margin usage	61.8% of NLV (9/17)	29.0%
TRON concentration	36.6%	0%
Excess liquidity	—	$7,125.56
Buying power	—	$44,990

In one line: the account went from “every red line flashing” back to “breathing normally”. Leverage under 1x for the first time.

II. What Happened in These Fourteen Months

A quick recap for readers who joined midway (veterans, you’ve been here the whole way):

July 2025. I opened TRON for reasons that sound very “retail” in hindsight: a blockchain narrative, big volatility, good for selling covered calls. I kept adding, brought the average cost to $2.08, held 4,300 shares at the peak, with 23 CCs sitting on top (6× 7.5C + 17× 5C).

Late 2025 onward, the plot flipped: the stock slid from the $4s toward $1.50. The CC rent was a band-aid against a falling price — it stops the bleeding, it doesn’t cure the disease. Worse was the seal: 17 of those JAN’28 5Cs welded 1,700 shares shut. Couldn’t sell them; and whenever IV spiked, buying the calls back got absurdly expensive.

August 19. Into a rally, I trimmed 2,000 shares (4,300 → 2,300), moving the heaviest block off the table.

August 31 → September 18. The two batches of short legs were bought back one by one (+$129.53, +$109.16). The seal was fully lifted.

September 21. The last 2,300 shares, sold in one order. Endgame.

The full fourteen-month ledger (system books):

Options side: +$584
Stock side: about -$2,097 (including tonight’s -$1,246)
Lending side: +$693.86 (Stock Yield Enhancement Program)
Net across the whole journey: about -$819

One more entry has to go on the books — and this +$694 wasn’t traded, it was rent. Throughout these fourteen months the position stayed enrolled in IBKR’s Stock Yield Enhancement Program (SYEP): shares lent out to short sellers, the fee accruing daily. Total: $693.86; best single day $7.36; when TRON was squeezed hardest, the market borrow rate topped 77% annualized (I keep half). So the dead inventory wasn’t just bleeding — it was quietly collecting rent all along.

With that added, the full-journey net narrows from -$1,513 to about -$819 — the tuition cut almost in half.

I’m not going to dress this up as a victory. Roughly $800 of tuition bought an upgrade worth far more than $800 — and here is what the upgrade actually contains.

III. Why This Is an Era-Level Trade

Because what just changed isn’t a number — it’s the structure of the account:

As of now, aside from the last 200 RGTI shares (and their covered short leg), all capital goes into options trading.

Three shifts underneath that sentence:

One: from inventory manager back to premium seller. For over a year I was nominally an options trader, but most of my energy went to servicing inventory: handling assignments, averaging down, tracking cost bases, lifting seals. The books say it plainly — 8 tickers of directional stock trades, 7 of them losses, -$4,064 cumulative; the options-selling engine, +$2,048 cumulative. The data wrote the answer long ago: stocks were where I bled; options are where I earn. My only job was to execute it.

Two: capital efficiency. The same dollar sitting in a stock is dead — it can only wait for a rebound that may never come. The same dollar inside an options engine is alive — CSPs collect rent, CCs collect rent, PMCCs lever the same exposure with far less capital. Time value works for me every day, instead of me standing guard for a stock every day.

Three: attention. What inventory really consumes isn’t money — it’s attention: checking the price, recalculating the cost, agonizing over whether to sell. With the inventory at zero, that attention finally goes back to the system itself: scans, signals, discipline.

IV. The Two NOK Puts Came Off the Same Night

Both NOK put positions were closed the same night: 10P @ $0.15 (+$61.13) and 10.5P @ $0.12 (+$34.34).

Puts are the biggest consumers of margin; with both off, the account’s room came right back:

Cash: -$81 → +$3,368 (the “cash ≥ $2,000” red line ✅ cleared)
Margin usage: 61.8% → 29.0%
Leverage: 0.82x — below 1x for the first time

One line still owed: RGTI concentration remains around 33% (over the 30% cap). The plan is unchanged — keep its covered structure, handle it on a rally, no new moves. But tonight, two of the red lines cleared first.

Updated to-do list: hold RGTI covered, keep the NOK PMCC structure, resume normal premium-selling pace once the account turns green (waiting on scans for candidates).

V. Three Lessons

As usual, three lines to keep:

Dropping the wrong position matters more than finding the right one. This -$1,246 stop-loss reset the whole account out of RED state. Had I understood this earlier, the tuition would have been half price.
The cost of a liquidity trap is never the price — it’s the time. Money can be re-earned; those welded-shut months can’t. From now on, the first question about any CC candidate is: “can I get out anytime I want?”
A system upgrade usually starts with admitting failure. What I cleared tonight wasn’t just a stock — it was a playbook: the playbook of making money by being directionally right. From here, probability and time do the work: sell premium, harvest time value, roll the wheel.

Thank you, TRON, for the three lessons. See you in the next life.

Trade Journal · Phased Stock Exit Series · The TRON Finale

Addendum (2026-09-25): +$693.86 of stock-lending fees has been added to the full-journey ledger — the fourteen-month net is revised from -$1,513 to about -$819. A pre-exit tail of roughly $2-4 is pending the October statement.

Addendum (2026-10-13): The ledger figures cited above re-verified against broker trade-level settlement records: directional stock trades corrected from 7 tickers / 7 losses / -$2,875 to 8 tickers / 7 losses / -$4,064; the options-selling engine’s cumulative from +$2,162 to +$2,048. The fourteen-month TRON-specific books and trade records are unaffected.

The Leverage Blowup Files: From Nobel Laureates to the AI Stock God, They All Died in the Same Place

Sun, 20 Sep 2026 00:00:00 GMT

The Problem

In article #001, I wrote about why I cap my leverage at 1.5x.

At the time, that number came from instinct and one personal scare — my account had drifted to 2.25x leverage with 77% margin usage, and one night I woke up with a single thought: “Another 10% drawdown and I get a margin call.”

But instinct needs evidence. So this month I did something simple: I pulled out the most famous blowups of the last three decades, one file at a time.

I only wanted to check whether my discipline held up. After reading them all, I turned on every light in the house.

Six Files

File 01 · LTCM (1998) — The smartest team on Earth, 25x leverage

Long-Term Capital Management had two Nobel laureates in economics on its board (Scholes and Merton) and the former bond chief of Salomon Brothers at the helm. Their models were nearly mathematically perfect — “the market makes small mistakes, but it doesn’t stay wrong for long.”

In 1995 and 1996, they returned over 40% annually. Then, from January to September 1998, they lost more than 90% of their capital. Balance-sheet leverage: 25-to-1. Notional exposure: over a trillion dollars. It took the Federal Reserve convening 14 banks to inject $3.6 billion to stop LTCM from dragging Wall Street down with it.

File 02 · XIV (2018) — The “safe income” trade that went to zero in a day

XIV was an exchange-traded note that shorted the VIX. The logic was seductive in its simplicity: volatility trends down over time, so stand on the selling side and collect rent every year.

The logic was right for five straight years. On February 5, 2018, the VIX went from 18.44 to 37.32 in a single day — more than double. XIV fell 84% after hours, then closed down 93% the next day as Credit Suisse announced liquidation. The bank lost over $500 million on the product.

One individual investor watched three years of savings — $4 million — evaporate overnight.

File 03 · Crude Oil Treasure (2020) — Ordinary people can blow through their accounts too

On April 20, 2020, the May WTI contract settled at −$37.63 a barrel — the first negative oil price in human history.

Bank of China’s “Crude Oil Treasure” product failed to roll its positions in time. The result: investors didn’t just lose their principal — they owed the bank money. This wasn’t a game for professionals. Participants were ordinary office workers buying what was advertised as a “100% margin, no leverage” product. It still blew through their accounts.

File 04 · Archegos (2021) — The leverage nobody could see

Bill Hwang, a legendary trader from the Tiger cubs lineage, built positions across multiple investment banks using total return swaps. Every bank thought he was a small client. Nobody saw the whole picture.

In March 2021, two trading days erased his $20 billion fortune. Credit Suisse lost $5.5 billion. Nomura lost $2.9 billion. No matter how well you hide leverage, a blowup is always public.

File 05 · The AI Stock God (2026) — Four times leverage, 36 hours

This one is last month’s news. Leopold Aschenbrenner, 24, former OpenAI researcher, ran a hedge fund with eight employees and +439% returns in the first half of the year. Then he applied four-times leverage to his AI thesis.

In July 2026, he lost 67% in a month. Over the following 36 hours, he sold roughly $16 billion in equities to Citadel in a single block trade — not because he chose to sell, but because he had to. “We let you down this month,” he wrote to investors. The Bank for International Settlements cited the case in a systemic risk warning.

File 06 · A contemporary specimen (2026)

I recently researched a Chinese-language options education platform that publishes its real account data on its website. I actually respect the transparency.

But I ran the numbers on their leverage: portfolio value ÷ net equity = 3.93x. They had borrowed $461,000 from their broker and paid $89,000 in margin interest over four years — 48% of their total deposits. Their worst month in 2026 was −71%. The account fell from $1.3 million to $157,428 — −87.9%.

And he’s still selling courses on how to earn stable income.

What I Took Away

Put all six files side by side, and the pattern is counterintuitive:

Common thread	What it means
All were brilliant	Nobel laureates, Wall Street veterans, AI prodigies. Intelligence does not prevent blowups
All had won first	None were beginners. Every one made real money, then added leverage
None got the direction wrong	LTCM’s model was right. The VIX mean-reversion thesis was right
All were forced out	Not one exited by choice. Every one was liquidated at the worst possible price
All went to zero	Not drawdowns — permanent capital loss

The first four aren’t fatal. The fifth is.

And of the five, I think the most dangerous is the second: they all won first.

Because success is the best leverage salesman.

You win once, confidence rises. You win twice, you start to think your capital is too small. You win three times, and you think: “If this is so stable, why not use someone else’s money and earn more?”

LTCM increased leverage after two consecutive years of 40%+ returns. The AI stock god used 4x leverage after +439%. The platform I studied earned 410% in 2025 and lost 79% the next year.

Adding leverage after winning is the most natural-feeling, most lethal move in investing.

One more thing deserves its own paragraph:

They didn’t lose on judgment. They ran out of time.

LTCM’s mean-reversion model wasn’t wrong — the market just stayed irrational longer than their margin lasted. XIV’s “volatility trends down” thesis wasn’t wrong — February 5, 2018 just didn’t belong to “the long run.”

The real mechanism of high leverage isn’t that it makes you wrong. It’s that it strips away your options at the worst possible moment.

At 1x leverage, you can wait out a crash. At 4x, when the crash comes, you can only sell — and you sell when everyone else is selling, at prices set by panic.

That’s what “forced to make the right decision at the wrong time” actually looks like.

The Insight

Leverage cannot buy you returns. It can only buy you speed. And the price of speed is your margin for error.

Investing is a survival game, not a speed game. The ones who last aren’t the fastest — they’re the ones who never get knocked out.

Here’s how I’ve turned this into non-negotiable rules:

Leverage cap at 1.5x, hard stop at 2.0x. The math is simple: at 1.5x I can survive a 20% drawdown; at 3.9x, a 25% decline wipes me out.
Never use margin loans for directional trades. Interest is a certainty; trading returns are not. Using certain costs to chase uncertain gains means starting 6% in the hole every year.
In a losing streak, cut leverage — never average down. The platform in File 06 kept high leverage through six consecutive losing months. That’s not trading. That’s acceleration.
Before every trade, ask: if the underlying drops 20% tomorrow, do I get liquidated?

Closing

I filed these six cases in my knowledge base under a single tag: cautionary tales.

Not because they scare me. Because I know — I haven’t won enough yet, which means my most dangerous moment hasn’t arrived.

The truly dangerous moment isn’t now. It’s the future day when my options income engine has run smoothly for a dozen months, when the equity curve is so clean that risk feels theoretical. On that day, someone — possibly me — will say: “If it’s this stable, why not add just a little leverage?”

These six files are my answer for that day.

Trading Insights #009 · To be continued

125 Years of Nobel Prizes — Why the Money Keeps Growing

Sun, 20 Sep 2026 00:00:00 GMT

A counter-intuitive headline

On 18 September 2026, one piece of news got buried in the feed: the Nobel Foundation announced it would raise this year’s prize money by 9%, to 12 million Swedish kronor per prize. The occasion — 2026 marks the 125th anniversary of the awards.

I paused for a second and asked myself a dumb question:

A “person” who lives off the income from a principal, paying out bonuses for 125 years — why has the money not run out, but instead grown bigger and bigger?

After digging into the numbers, I realized the Nobel Foundation’s books might be the most perfect “cash-flow system” sample on this planet. And the math it proves happens to be exactly what I do in my options account.

The books first

In 1900, the Nobel Foundation was established with roughly 31 million Swedish kronor — almost the entire fortune Alfred Nobel left in his will.

By the end of 2025, the Foundation’s invested capital had a market value of SEK 6.744 billion.

Mind you, it has paid out prizes every single year of those 125 years — in 2025 alone, prizes plus operating expenses totaled SEK 148.1 million, about 2.2% of assets. Yet the principal still grew from 31 million to 6.7 billion.

Even more interesting is the return sequence. From the Foundation’s 2025 annual report, the five-year scorecard:

Year	Return on investment
2021	+18.3%
2022	−1.9%
2023	+10.6%
2024	+11.6%
2025	+0.9%

Five-year average: +7.7% per year. And the goal written in its annual report is merely “a total inflation-adjusted return of at least 3%”.

It does not chase explosive returns (just 0.9% in 2025), and it does not guarantee a positive year (2022 was negative). Yet over a century-plus, the return rate has stayed reliably above the spending rate — that is the entire secret.

The math is one formula

Return rate − Spending rate = Net principal growth

The Nobel Foundation’s actual numbers: 7.7% (five-year average return) − 2.2% (annual spending) ≈ 5.5% net principal growth per year.

Unimpressive on its own. But put that 5.5% into the machine of time, and it is where the 6.7 billion came from.

Flip it around: if return is 5% and spending is 6%, the books slowly shrink until prizes can no longer be paid. The life-or-death line of a perpetual system is not how high the return is — it is the relative size of return versus spending.

But its first 80 years were anything but glamorous

This part comforted me most while reading the history — because it looks so much like an ordinary person’s account:

For the first decades, the will restricted investments to “safe securities” (bonds, real estate, municipal loans). Taxes ate into it — the fund actually shrank;
By 1925 it was so poor that it simply skipped a year of prizes as a protest — which won it a tax cut from 40% to 3%;
Real estate was only allowed from 1939; equities up to 60% only from the 1970s; up to 70% in the 1990s;
Real growth only began in the 1980s — after which assets quadrupled over 20 years;
Only in 1991 did the fund get back to its 1901 size (in real terms);
Then came the dot-com crash: assets fell from almost 5 billion to just over 3 billion kronor — down nearly 30%; the 2008 financial crisis hit hard again.

See that? The first 80 years of this “perpetual system” were mostly foundation-laying. It lived through two crash-level drawdowns and even skipped paying prizes once.

And it survived for exactly one reason: income covered spending, so it never had to sell assets in a panic. Crashes could be waited out, rules could be adjusted slowly — as long as cash flow never dries up and you are never forced to sell, time is on your side.

Versus myself: a miniature Nobel experiment

My options cash-flow system is essentially the same structure at a tiny scale:

Nobel’s “principal” = my account’s net liquidation value;
Nobel’s “prize money” = the premiums I collect selling options;
Nobel’s “spending rate” = what I take out of the account (currently zero — everything compounds).

Here is my real ledger (as of end-September, all figures from broker trade-level settlement records):

Options cash flow (settlement basis): +$2,048, across 143 trades, about +$228 per month;
That translates to a monthly cash-flow rate of roughly 2.3% (realized option P&L ÷ current NAV ÷ 9 months);
But over the same period, directional stock positions lost −$4,064. Net: −$2,016.

Honestly, my “foundation” is still in its 1900–1950 phase: the engine is built, the structure is not finished — legacy stock positions still weigh on the account, and principal integrity is only 90.0% (NAV ÷ cumulative deposits).

So I will not claim “I make 2.3% a month, 28% annualized — 3.6 times faster than the Nobel Foundation.” That would be self-deception. Nobel’s 7.7% is 7.7% built from 55% diversified equities, hedge funds and property. My 2.3% is 2.3% built from small-cap options with unresolved concentration and unverified tail risk. Speed and scale cannot be compared — only the structure can:

Return must stay above spending — my “spending rate” is currently zero, but what really matters is “return minus drawdown”;
Never be forced out — leverage ≤1.5x, assignment handled within two trading days, only verified seller structures; all of this buys one thing: the right to never sell at the bottom;
Rules must evolve — it took the Nobel Foundation 80 years to get its asset allocation right. My Q4 validation period is my own “1980s”.

Three things to take away

If you also want a “perpetual system” — an options account, rental income, a business, a side hustle — the Nobel books offer three immediately actionable lessons:

Calculate your spending rate first: how much do you withdraw from the system each year, as a percentage of assets?
Make the return rate stand firmly above the spending rate before talking about how high returns can go. A stable 1% beats a volatile 10%.
Treat time as a multiplier: a 40% loss needs a 67% gain to recover; but a system compounding 5.5% net per year is 5x in 30 years. The first principle of compounding is not the return rate — it is “never interrupt it”.

The Nobel Foundation’s former CEO Lars Heikensten once said their investment philosophy could be summed up in four words:

“We’re in this forever.”

Not “how much this year” — but “here forever”. That sentence deserves to be posted above every options position I hold.

Trading Insights #008 · to be continued

Addendum (2026-10-13): Figures re-verified against broker trade-level settlement records (through Sep 30): options corrected from +$2,101 / 121 trades / +$246 per month to +$2,048 / 143 trades / about +$228 per month; monthly cash-flow rate corrected from about 2.6% to about 2.3%; stock direction corrected from -$2,875 to -$4,064, net account balance from -$773 to about -$2,016; principal integrity from 94.3% to 90.0%. Structure and conclusions unchanged.

The 10C Roll, Filed at the Deadline: -231 to Close the Old Leg, +220 to Open the New — and an Earnings Calendar We Almost Missed

Sat, 19 Sep 2026 02:51:00 GMT

The 10C Roll, Filed at the Deadline: Two Fills, Three Lessons

September 18, 2:51 pm ET. Nine hours left before the close. Nine hours left before my five 10C SEP18 contracts would have turned into an accidental short of 500 NOK shares.

One combo order, two fills at the same instant:

Bought back 10C SEP18 ×5 @ $0.66 (paid $330)
Sold 12C OCT23 ×5 @ $0.44 (collected $220)

Old leg closed, new leg open. That’s what a roll is.

The Ledger: What This 10C Leg Really Cost

Ledger detail (click to expand)

The 10C SEP18 was sold in early September at $0.198, collecting $99. Tonight it was bought back at $0.66, paying $330.

Realized loss on this leg: -$231.

The new 12C OCT23 brought in $220 of premium. Net of the two: roughly -$11 — a wash.

Compare with the alternative: buy back and walk away, and -$231 lands for good, with zero coming back. The roll converted a certain -$231 into “probably -$11, possibly worse” — as long as NOK stays below $12 through October 23, this cycle closes at -$11.

That’s the seller’s exchange: use a new obligation to subsidize an old wound.

Lesson One: A Qualifying Roll Must Cover the Buyback with New Premium

In the post-mortem, one question had to be asked: was this roll qualified?

The standard should be: new premium ≥ buyback cost. Covering the loss is the floor; collecting more than the buyback — making money on the roll — is what makes it good.

By that standard: paid $0.66 to close, collected $0.44 on the new leg — $11 short of covering. This roll was only “nearly qualified.”

The fully qualified candidate was sitting right there on the chain: NOV20 12C at $0.63, which would have covered the $0.66 buyback. I had ranked “shorter lock-up” above “loss coverage” and demoted it to a backup.

Wrong order. The first question of any roll is “can the loss be recovered?” The second is “how long must the obligation run?” Cover first, profit second; coverage before tenure.

Lesson Two: The New Expiry Must Avoid the Earnings Calendar — We Almost Missed This

More worth recording than the $11 is the calendar problem.

I chose the 12C OCT23. Expiry: October 23. NOK’s earnings: October 22.

The new short leg expires one day after earnings — it is itself an earnings-spanning short call.

Before rolling I checked the option chain and never checked the earnings calendar. What does an earnings-spanning 12C mean? If NOK gaps on earnings morning, a $12 buffer can evaporate in the first tick. Sellers don’t fear slow declines — they fear gaps.

So the management line is clear:

Before October 15: let the 12C burn theta. Do nothing.
If NOK approaches $12, or when the October 15 earnings blackout window arrives: roll earlier, to a date further past earnings — first choice 13C DEC18 (strike up to 13, +22% buffer from spot, and net credit positive under the new rule), second choice 13C NOV20 (near-free upgrade to the 13 strike).

Rolling early beats rolling the night before earnings — cheaper, and a lot calmer.

Lesson Three: Two Checklists Before Every Roll

Tonight compresses into two iron rules, in this order:

Money: is the new premium ≥ the buyback cost? If not, keep searching further out in expiry and higher in strike; if nothing on the whole chain qualifies, a plain buyback is the only option — no fake rolls.
Calendar: does the new expiry fall inside an earnings or event window? For any event-spanning leg, either move the expiry further out, or move the strike up to where even a gap can’t reach it.

Order first, calendar never.

The Full NOK Short-Leg Map (after the Sep 18 close)

Leg	Qty	Expiry	Status
10.5P	×4	SEP25	$0.07 from the strike, assignment risk , decide by Sep 23
11C	×10	OCT2	Δ0.41, roll up before Oct 2
12C (new)	×5	OCT23	earnings-spanning, pre-roll by Oct 15
10P	×2	OCT2	holding for theta, take-profit at 60-70%
LEAPS 10C	+15	JUN’27	the covered base for all short calls

Fifteen short calls, exactly covered by fifteen LEAPS — the short-call side is at full capacity; no new NOK seller legs until one of these batches clears.

A roll is never skipping class — it’s moving the homework from tonight to next week. But only if you actually flip next week’s calendar first.

TRON Seal Lifted: All 17× 5C Bought Back, +$109 — the Final Piece of a Nine-Month Liquidity Trap

Fri, 18 Sep 2026 22:26:00 GMT

September 18, 2026. 22:24 and 22:26. Two buy orders.

7 contracts @ $0.50, 10 contracts @ $0.48 — every one of the 17 TRON JAN’28 5C short legs, bought back.

The seal that went up nine months ago has been lifted.

I. The Ledger First

Ledger detail (click to expand)

Time	Contract	Side	Qty	Sell avg	Buyback	Commission
22:24	TRON JAN’28 5C	BOT	7	$0.548	$0.50	$3.43
22:26	TRON JAN’28 5C	BOT	10	$0.548	$0.48	$3.70

Buyback average $0.488 against a $0.548 sell average — $101.66 gross, +$109.16 on the settlement basis (+11.7%).

A short leg that hung around for nine months and spent part of that time underwater closes with a small win. But the ledger is the least important page of this story.

II. What the Seal Locked Up

The 17× 5C was the last piece of the TRON wreck. Early in 2026, the TRON CC complex held 23 short legs (6× 7.5C + 17× 5C). The 7.5C group closed on August 31 (+$129.53); these 17× 5C were everything that remained.

What they locked: 1,700 of my 2,300 shares. To sell the stock I had to buy the Calls back first — and they expired JAN’28, sat on an options chain with almost no volume, and quoted bid-ask spreads so wide they looked broken. For nine months, 1,700 of my shares had no freedom.

And then there was the double-loss structure: the stock fell from above $4 to $1.47, so the 5C should have gotten cheaper — instead, small-cap IV expansion pushed its mark to $0.875 at one point. The stock was bleeding, and the Calls were bleeding too. That’s the true horror of a liquidity trap: not one side losing — both sides welded shut, hurting whichever one you touch.

III. Why the Unlock Happened Today

Unlocking costs money. Today’s two buys cost about $830 — money I couldn’t afford to spend for months.

Not because the account was empty — because the account state didn’t allow it. When cash was negative and leverage over the line, spending $830 to buy back an “underwater leg” meant paying old debts while still bleeding. Completely wrong sequence. So the answer was always: fix the account first, break the seal second.

And today’s price was the right time to pay: at the $0.875 mark earlier this year the buyback would have cost $1,490. Today $0.488 got it done — waiting for IV to cool saved about $660. Buying back a painful leg is painful; buying it back at its most expensive would have been pain squared.

IV. After the Unlock

All 2,300 TRON shares are free — for the first time in nine months.

Everything left follows the constitution: GTC ladder orders — 1,200 shares @ $1.50, 1,100 shares @ $1.80. If they fill, I’m out; if not, I wait. No watching, no betting on highs. Once the stock is gone, the TRON story is over, and the account enters options-only for real.

Hunter’s Summary

The real cost of a liquidity trap isn’t the loss. It’s the time. Money can be earned back; nine locked months are the position’s true opportunity cost.

Three takeaways:

Unlocking costs money, and a RED-state account doesn’t have the right to spend it. Fix the account first (cash positive, leverage in line), then handle the wreck. Reverse the order and you’re still waiting for IV to cool.
Time buybacks on IV, not on mood. $0.875 versus $0.488 is a $660 difference — same legs, same account, the only difference being patience.
Wrong CC parameters equal time lock-up. 16-month DTE plus near-zero liquidity — see The Hidden Cost of Covered Calls.

The hardest seal is broken. Selling the stock — the last step — goes to the GTC ladder, one rung at a time.

Trade Journal · Phased Stock Exit Series · TRON Chapter

Phased Stock Exit Day 1: QUBT Sold Outright, RGTI Covered with 18C Rent (+$92 Premium)

Thu, 17 Sep 2026 11:10:00 GMT

September 17, 2026. Quantum names rallied — RGTI +7.75% intraday, QUBT +5.56%.

On green days, retail thinks about adding. I thought about exiting.

That’s not bravado. Nine months of account data already delivered the verdict: my option-selling engine made +$2,048 across 143 trades, while 48 stock-direction trades lost -$4,064 — eight tickers, seven in the red. Stocks are where I bleed, and a rally day is exactly the right day to execute an exit plan. This journal covers Day 1 of the phased stock exit.

I. Today’s Trade Ledger

Ledger detail (click to expand)

Time	Action	Qty	Price	Realized
Pre-market	NOK 10P SEP18 bought back	2	$0.04 (sold avg $0.243)	+$40.60
Intraday	NOK 10P SEP25 bought back	2	$0.10 (sold avg $0.353)	+$50.60
15:10	QUBT liquidated (50/100/150)	300 shares	$8.422 / $8.425 / $8.4223	-$371.16
15:11	RGTI 18C OCT9 sold	2	$0.47	+$94 premium (unrealized)

Daily realized on the broker’s settlement basis: -$374.26 (the three lines above plus commissions). The $371.16 QUBT figure uses the $9.66 tracked cost; the ~$91 gap to the settlement figure (≈$0.30/share) comes from the broker’s cost-basis treatment and historical fee adjustments — settlement basis wins.

The two NOK put buybacks were housekeeping: 10P SEP18 was up 92% with expiry tomorrow, 10P SEP25 up 72% — both hit the seller’s profit-taking discipline. The real work was the two stock exits below.

II. Premium Thickness Decides the Exit Method

QUBT and RGTI both needed to go — but by different routes. One criterion ruled: is the CC premium worth locking the stock for 30 days?

Live chain data that day (29 DTE, “strike ≥ cost” exit-style CCs):

Ticker	Spot	Exit strike	Premium
RGTI	$15.94	18C (needs +13% to exceed $17.86 cost)	$0.63-0.67
QUBT	$8.45	10C (needs +18% to exceed $9.66 cost)	$0.22-0.23

Same “assignment = profitable exit” structure, but RGTI’s rent is three times QUBT’s. $22 to lock shares for 29 days while carrying an 18% upside gap? Not worth it. $65 on RGTI? Now renting your way out makes sense.

So: QUBT sold outright, RGTI exited via Calls.

QUBT filled at an $8.423 average — above the $8.25 top rung of my old GTC plan; the rally day handed me a better exit than I had planned. RGTI got 2× 18C OCT9 sold at $0.47, $94 of premium in (about $92 after fees). That Call isn’t rent collection — it’s writing the sell decision into the price in advance: if RGTI runs past 18, I sell at 18, above my $17.86 cost, and exit in profit; if it doesn’t, I keep the $92 and run the same play in 29 days.

III. The Phased Stock Exit Framework

Three steps, each with its own selection rule:

Thin premium? Sell outright. Rent that doesn’t justify a lock-up means no lock-up. Timing belongs to GTC ladder orders — no staring, no betting on highs.
Thick premium? Exit via CC. Strike must be ≥ cost — assignment isn’t an accident, it’s your sell order filling. The premium is your pay for waiting.
Liquidity trap? Handle it last. TRON’s 2,300 shares plus 17 short Calls sit at the end of the queue: sell the 600 free shares first, buy back the Calls to unlock, then liquidate. Order matters.

Two iron rules run through all of it: sell first, repair second, deploy last — no new positions the night you liquidate, recovered capital first restores positive cash and sub-1.5x leverage; and exit-vehicle exceptions during RED state must be reviewed and logged — this RGTI Call was approved and recorded by the AI review process, not casually “rent collected”.

IV. The Account After Execution

One night of liquidation plus rent turned the account from RED toward repair: cash -$1,758 → +$860 positive, leverage 1.55x → 1.29x, and RGTI concentration starts shrinking toward control (assignment at 18C zeroes it out).

One hard bone remains in the stock book: TRON. Schedule unchanged — step three above.

Hunter’s Summary

A rally day isn’t a greed day. It’s a plan-execution day. Split the sell into three rungs — thin premium sold outright, thick premium sold as Calls, trap positions dismantled slowly — and exiting stops being an emotional decision and becomes an assembly line.

Day 1 done: cash positive, leverage back in the safe zone. The phased exit showed its first advantage: you don’t need to nail every price in one bet — you just need every step to stay on the plan.

Trade Journal · Phased Stock Exit Series · Day 1

Addendum (2026-10-13): The ledger figures cited here re-verified against broker trade-level settlement records: options corrected from 121 trades / +$1,825 to 143 trades / +$2,048; stock direction from 41 trades / -$2,874 to 48 trades / -$4,064 (7 of 8 tickers in the red). The day-one trade ledger and account figures are unaffected.

On the Day NOK Jumped 7%, I Bought Back Two Groups of Puts I Had Sold: +$91 Surfing With the Tide

Thu, 17 Sep 2026 09:54:00 GMT

September 17, 2026. Nokia announced an expanded Microsoft partnership: Nokia Data Suite joins Microsoft Fabric, bringing agentic AI to telecom network automation. NOK opened +5% pre-market and touched +7% intraday.

Three days earlier, on September 14, NOK crashed 13% — and the first thing I did was buy back the Calls I had sold. Today, the direction flipped completely: on the rally day, I bought back two groups of Puts I had sold.

The two moves look opposite. They’re the mirror image of the same action. This journal is about why high tide and low tide days see me doing the same thing.

I. The Ledger First

Ledger detail (click to expand)

Time (UTC)	Contract	Side	Qty	Sold avg	Bought back	Realized
13:37	NOK SEP 18 ‘26 10P	BOT	2	$0.243	$0.04	+$40.60
13:54	NOK SEP 25 ‘26 10P	BOT	2	$0.353	$0.10	+$50.60
					Total	+$91.20

Combined gross profit +$91.20 (76.5% of premium captured); roughly +$91 net of commissions on the settlement basis. Both Put groups were CSP short legs sold around 9/14. After NOK’s three-day run, their residual value evaporated fast — closing them here locks in most of the profit.

II. Why Buy Back Puts on a Rally Day Instead of Waiting for Expiry

Selling Puts gets paid for “not falling”. The harder the underlying rallies, the faster “not falling” gets settled — which means the profit has already arrived early.

Look at what was left on the table:

10P SEP18: 83.5% of profit banked, expiry tomorrow, $0.04 of residual left (≈$8 on 2 contracts). Holding $8 of residual through one more overnight of event risk? Not worth it.
10P SEP25: 71.7% banked, 8 days and $0.10 of residual left (≈$20). That hit my profit-taking line for short legs (60-70% and up). The remaining $20 isn’t profit — it’s a risk wage for holding 8 more days of directional exposure. NOK had just run three straight days, the news was out, and it faded from the high — sentiment is resetting and the next wave’s direction is unknown. A $20 risk wage isn’t worth boarding the ship for.

Realizing Put profits shouldn’t depend only on “waiting for expiry”. When the profit line is high enough and the residual is thin enough, buy back proactively — take the money you’ve already earned out of the next wave’s gamble.

III. Two Sides of the Same Ledger: Crash Day vs. Rally Day

Put the two days side by side:

	9/14 Crash day	9/17 Rally day
Market action	NOK -13.30%, Call residuals evaporated	NOK +7%, Put residuals evaporated
My action	Bought back 12 sold Calls, +$166 banked	Bought back 4 sold Puts, +$91 banked
Logic	The down-wave already paid my Calls	The up-wave already paid my Puts

Same muscle memory both times: whichever leg the wave is pushing toward fast profit, reel that one in.

Trading really is like surfing. When a wave comes, you either ride it or get slammed. But smart investors aren’t surfers — a surfer bets on the direction of the next wave. A helmsman bets on nothing: he only makes sure every wave that passes the boat leaves a little momentum behind. Premium is the momentum the boat collects. Big wave or small, rising or falling — the boat keeps moving forward.

Three straight up days, news out, faded from the high — whether NOK continues up or pulls back tomorrow is a question I don’t need answered. Because when this wave passed my boat, I had already stowed everything it owed me.

IV. The Ride-the-Wave Profit-Taking Discipline

Act when the profit line hits. Short option legs: start closing at 60-70% profit (stock-direction positions use 20-30% — different standards, don’t mix them).
Do the residual math. What’s left on the table, divided by days remaining, weighed against the event risk ahead. Not worth it? Close it.
Never bet on the next wave. Replace “I think it’ll keep rising/falling” with “the profit is already settled”. Wave prediction is the surfer’s job; the helmsman just collects momentum.

Trade Journal · Mirror image of On the Day NOK Crashed 10%, I Bought Back Two Calls I Had Sold

The Hidden Cost of Covered Calls: Sell the Stock or Sell the Call After a Big Rally? — The RGTI Decision and the TRON Trap

Thu, 17 Sep 2026 00:00:00 GMT

On a +6% day, the most uncomfortable person isn’t the one holding cash. It’s the one holding underwater stock.

Sell now? What if there’s another wave tomorrow. Hold? What if this rally is the escape hatch. That’s when the Covered Call looks like the perfect compromise: keep the shares, collect a premium first, and “wait gracefully.”

But the CC math is nowhere near that clean. This guide uses my real RGTI decision from September 2026, plus the TRON trap that has been hanging around my neck for nine months, to lay out the hidden costs of using CCs as an exit strategy.

I. Two Half-Truths About Covered Calls

There are three popular claims about CCs. Two are true. One is a trap.

“A CC is hedged by stock, so it’s not naked” — true. With 200 shares you can sell at most 2 Calls. The iron law stands.

“A CC doesn’t use margin” — also true. The shares themselves are the collateral; a covered call adds no margin requirement.

“Therefore a CC is free” — that’s the trap. “No margin” is exactly what makes CCs dangerous: zero-cost entries are easy to overdo, and one day you discover your stock is locked up by your own Calls.

The real cost of a CC is never margin. It’s liquidity lock-up:

Upside locked: the stock runs past your strike, your gains freeze, and selling the shares now means buying back an expensive Call first;
Downside fully yours: the stock drops, your Call expires worthless and the premium is nice — but the shares bleed all the same. The rent is just a cushion;
Exit lane welded shut: selling a CC is a promise to “hold to expiry or buy back”, for up to several weeks — during which “reducing the position” is no longer a free action.

In one sentence: when you sell a CC, you’re not selling premium — you’re selling the freedom to sell your stock for the next few weeks. That bill is far bigger than margin.

II. The RGTI Decision: Sell or Sell a Call After a +6% Day?

On 2026-09-17, RGTI jumped 6% in a day. I held 200 shares at a $17.86 cost basis, down 12%. Three options on the table: sell the shares outright, sell part, or sell a CC.

First, answer the real question — “is there a second wave after a big up day?”

I ran the numbers on two years of daily data (108 samples of single-day moves ≥ +5%):

Next-day median move: +0.09%, with a 51% win rate — statistically a coin flip;
Day 3 median: +0.53%, 54% win rate — a coin slightly biased positive;
Probability of falling back below the rally’s starting point within 5 sessions: 34% — two-thirds of the time, the gain isn’t fully given back.

“The second wave” doesn’t hold up in the data. It can exist, but it’s not a trend — it’s a coin with a tiny positive bias. Betting on it is betting on coins.

Now the CC math (OCT16 expiry, 29 days out, live premiums from the decision day):

Strike	Premium/contract	If assigned (per share)	If not assigned
16C	$1.13	Sold at 16: net -$0.73	Keep shares + $1.13
18C	$0.55	Sold at 18: net +$0.69 (above cost)	Keep shares + $0.55
20C	$0.27	Sold at 20: net +$2.41	Keep shares + $0.27

Look closely at the difference between the 16C and the 18C: the 16C pays the fattest premium ($1.13), but it’s a promise to sell at 16 — locking in a loss. The 18C pays half the premium, but assignment means exiting at a profit.

That’s the first law of CC exits: the strike is the price you’re willing to sell at. Assignment isn’t an accident — it’s your sell order getting filled. Choose the 18C and you’re happy when assigned; choose the 16C and you’ll wince. And the “losing on both ends” script almost always starts with a strike that makes you wince.

The worst path is rally-then-crash: RGTI runs to 17, the Call goes ITM, and closing the position means buying back an expensive Call; then it dumps back to 14 (August was exactly this: +11% followed by -8.6% one session later). You buy the Call back — and the stock loss didn’t miss a single bite.

III. The TRON Trap: Wrong Parameters, Locked Up for Nine Months

The RGTI decision was a 29-day problem. TRON shows what happens when the parameters are wrong: a lock-up that has lasted nine months, with no way out on either end.

My TRON position: 2,300 shares at a $2.08 cost basis, with 17 short $5 Calls expiring JAN’28 sitting on top. The logic at entry was standard: hedged by stock, no margin used, collect a “safe” premium.

Looking back today:

17 Calls lock up 1,700 shares — of the 2,300 shares, only 600 are free;
Selling the stock first requires buying back 17 Calls for roughly $1,500, and those Calls have thin liquidity and wide bid-ask spreads;
The stock side is deeply underwater, and the Call side is underwater too because small-cap IV exploded — losing on both ends, and neither end can move.

This wasn’t a “wrong direction” problem. It was a structural problem:

Wrong DTE. A 16-month lock-up handed away the right to sell my stock a year and a half in advance. On high-volatility small caps, keep CCs short — 30 days is my cap.
Wrong liquidity check. Small-cap options have low open interest and wide spreads; the moment you want out, slippage and buy-back costs blow past expectations. Iron law: illiquid underlyings don’t deserve CCs.
Wrong account rhythm. When the stock went underwater, I kept it and tried to “rent my way back to breakeven” — and the rent never covered the house. The ledger for those small caps says it plainly: options +$482, stock -$712.

The TRON trap wasn’t the fault of “selling a CC”. It was wrong parameters plus refusing to cut when cutting was due. Tuition for those nine words: $4,000+.

️ IV. The Right Way to Use CCs as an Exit Tool

When to do it (three steps):

Fix the account before opening the CC. When cash is negative, leverage is over the line, or concentration is breached, the first move is always selling stock — not selling a Call. A CC is an exit tool for an already-healthy account, not a tourniquet for a bleeding one.
Strike ≥ your cost, or ≥ the price you genuinely want to sell at. Before opening, ask one question: “If this Call gets assigned, am I happy or am I wincing?” Happy, open it. Wincing, pick a different strike.
Keep DTE short. Cap it at 30 days on high-volatility names. Your selling plan must be able to tolerate a 30-day lock.

When not to do it (checklist):

Cash negative, leverage over the line → don’t. Sell stock first;
A definite selling plan within the next few weeks (earnings, lockup expiry, GTC orders) → don’t. The CC will block your sell order;
Small cap + long DTE + high IV (the TRON script) → don’t;
Picking a below-cost strike just to “collect more rent” → don’t. That’s a promise to sell cheap;
“I feel like it can go higher” → don’t. That’s not a reason for a CC. That’s betting on coins.

Hunter’s Summary

The CC is one of the smoothest tools in a seller’s toolbox — but its real cost isn’t margin. It’s freedom: the freedom to sell your stock for the next few weeks.

Use the CC as an exit tool (strike = your selling price), and there’s no losing on both ends — assignment is exactly the outcome you wanted. Use it as a rent machine bolted onto underwater stock, and it becomes the next TRON.

On the big rally day: fix the account first, then think about the Call.

Related reading: The Wheel Strategy Explained, PMCC: The Poor Man’s Covered Call

The Rent Never Covered the House: Nine Months of Hard Data — Options Cash Flow vs. Stock Direction

Thu, 17 Sep 2026 00:00:00 GMT

The Problem

Two people live inside my account.

One sells insurance: sells Puts, sells Calls, collects premium, and gets paid for time. The other bets on direction: buys stock, waits for it to rise, and holds on when it falls.

Nine months later, the insurer made $2,048. The gambler lost $4,064.

That’s not a metaphor. Those are the real numbers from my IBKR settlement reports, January through September 2026. Today I’m publishing the verdict — and answering the question that keeps circling back: for the stock still stuck underwater, do I keep holding, or sell it and turn the capital into a cash-flow machine?

My Thinking

First, the scorecard (settlement basis, through 2026-09-30):

Engine	Trades	P&L	Avg per trade	Record
Option selling (CSP/CC/PMCC)	143	+$2,048.22	+$14.32	6 of 8 tickers positive
Stock direction	48	-$4,063.80	-$84.66	7 of 8 tickers negative

The monthly track makes it even clearer: May through July, before the rules were enforced, the account bled -$1,807. August and September, after “seller-only” became the constitution, the account turned +$27. Same hands, same account — the only variable that changed was discipline.

And the most painful part is the same-ticker ledger:

Ticker	Options P&L	Stock P&L	Net
TRON	+$584	-$851	-$267
RGTI	+$482	-$712	-$230
QUBT	+$241	-$382	-$141
NVTS	+$202	-$492	-$290

Four tickers, options profitable on every one, stock losing on every one. The rent never covered the house.

The Insight

On the same ticker, selling its volatility pays better than betting on its direction — not because selling options is smarter, but because “waiting to break even” is a wound that keeps bleeding.

TRON is the most expensive footnote to that sentence: the stock once showed a +$2,000 unrealized gain I didn’t take. It ended as a -$2,838 unrealized loss — a $4,800 round trip. Nine months of lock-up, and not just the money: the opportunity cost. That same capital in the seller engine has run at roughly 2.4% a month.

The math of “holding on” was never “once it recovers I win”. It’s: by the day it recovers, you’ve already lost every dollar of cash flow that capital could have produced — plus nine months of sleepless nights.

Practical Advice

If the stock must go, it should still leave with discipline. Three rules:

Sell by plan, not by emotion. No panic-selling on crash days, no greed on rally days. The correct way to “find a good price” is to place an order — not to stare at the tape.
Use GTC ladder orders. Split the exit into limit orders at a couple of price rungs. If they fill, you’re out; if not, you wait. No watching, no betting on highs — let the system do the waiting.
Sell first, repair second, deploy last. No new positions the night you liquidate. Recovered capital first turns cash positive and brings leverage back under 1.5x. Only after those gates pass do new positions enter the review pipeline.

Where does the money go afterwards? Back where it belongs: cash-secured Puts, and CCs/PMCCs backed by shares or LEAPS. Selling “nominal CSPs” while holding underwater stock was really renting with borrowed money; after the stock is gone, the same strategy becomes genuinely cash-secured — a better risk structure, not a worse one. Core tickers still follow the old standard: under fifty dollars, long-term uptrend, thick liquidity, fat premium.

Trading Insights #007 · To be continued

Addendum (2026-10-13): Ledger figures re-verified against broker trade-level settlement records (through Sep 30): options corrected from 121 trades / +$1,825.77 to 143 trades / +$2,048.22; stock direction corrected from 41 trades / -$2,874.83 to 48 trades / -$4,063.80 (assignments and expiries now fully included). Monthly track re-aggregated: May–July revised from -$1,875 to -$1,807; August–September (full months) revised from the old snapshot +$923 (Aug 15–Sep 15) to +$27 (the full months include the late-September TRON liquidation of -$864; the Aug 1–Sep 17 window alone was +$903). The conclusion stands: the rent never covered the house.

When You Get Called Away, Your LEAPS Won't Automatically Deliver — The Most Costly PMCC Misconception

Wed, 16 Sep 2026 00:00:00 GMT

“I bought 5 LEAPS and sold 5 Calls, built a PMCC — now I’m bulletproof.”

If that sounds like you, this post is for you. Because the name “PMCC” has fooled almost every beginner — including me, once.

I. Misconception #1: A PMCC Is a Strategy, Not a Position

In your head, 5 LEAPS + 5 short Calls form one packaged “combo” — one rises, the other falls, and at expiry the system “settles it as a whole.”

Wake up.

To the broker and the OCC (Options Clearing Corporation), the word “PMCC” simply doesn’t exist. They only see two independent contracts:

Leg	Contract	Expiry	Role
Long	Buy $10 Call	9 months out	Your long
Short	Sell $10 Call	This Friday	Your short

“PMCC” is just a name you made up in your head — like naming your cat “Lucky.” The bank account doesn’t pay interest on it.

⚙️ II. Misconception #2: When You Get Called Away, the System “Backs You Up”

This is the deadly one.

Many people assume: when the short leg is assigned, the system will automatically exercise my LEAPS, “deliver the shares” for me, the two legs cancel out, and everyone’s happy.

No. None of that happens.

Here’s the real sequence (you sold 5 × $10 Calls and hold no shares):

On expiry, the stock closes above $10 → your short Call is automatically assigned (this part is automatic; you can’t stop it)
Your account instantly becomes: short 500 shares + $5,000 cash
Your 5 LEAPS? Sitting there, untouched, waiting for you to act

Notice: the system only did one thing automatically — “assign.” It will never touch your LEAPS. Because your LEAPS are still 9 months from expiry, and the auto-exercise mechanism simply doesn’t apply to them yet.

So the next trading day, you open your account and find: short 500 shares + 5 LEAPS + $5,000 cash. The short position is a knife over your head — the higher the stock goes, the more you lose — and only you can close it.

Silent’s Note: In the options world, only two things are automatic: auto-exercise at expiry, and auto-assignment for sellers. Everything else — rolling, closing, unwinding — lives in your own hands.

️ III. Profit or Loss? Depends on Which Path You Manually Choose

Say the stock hits $10.5 at expiry and you get called away. Earlier you:

Bought 5 LEAPS @ $2.2 = $1,100
Sold 5 Calls @ $0.198 = $99

Same assignment, different manual actions, opposite outcomes:

What you manually do	Result
✅ Sell the LEAPS + buy back the short (don’t touch the shares)	≈ +$59
❌ Exercise the LEAPS to deliver (“my gut says close it”)	≈ -$1,001

Where’s the difference? In the “time value” inside your LEAPS.

The LEAPS you bought at $2.2 is worth about $2.6 when the stock hits $10.5. Break it down:

Intrinsic value $0.5 — what you can cash in by exercising
Time value $2.1 — the “waiting value” of 9 remaining months

If you “exercise the LEAPS to deliver,” you only get the $0.5 intrinsic and throw away $2.1 × 500 ≈ $1,050 of time value. The $99 premium you collected can’t fill that hole.

The right move is to sell the LEAPS (not the shares): the options market pays for time value, so your $2.2 turns back into about $2.6 — nothing wasted.

IV. Three Paths — All of Them Require You to Place an Order

① Don’t want assignment, want to keep collecting rent: before expiry, buy back the short Call and roll to a later, higher strike.

② Want to take profit and exit the whole combo: sell the LEAPS + buy back the short Call together (use a spread order in TWS, avoid a naked single leg).

③ Do nothing and get assigned into a short: the next trading day, buy back 500 shares to close the short. Never touch the LEAPS’ exercise button.

The timeline — memorize it:

Before Friday close  → you can still close/roll manually (the only calm window)
After Friday close   → OCC processes, short Call gets assigned
Monday open          → the short appears, now it's firefighting

Hunter’s Summary

The PMCC teaches us to “collect rent with less capital” — but it never promised “the system will clean up after you.”

What you thought was a “combo” is just two independent contracts that happen to live in the same account. Know that they go their separate ways, and you won’t be tricked by the “auto-backstop” fantasy into going naked short.

Related: PMCC: The Poor Man’s Covered Call — 1/3 the Capital, Same Rent

0DTE Options and the Survivorship Bias Factory

Tue, 15 Sep 2026 00:00:00 GMT

The Problem

Every few days, my feed shows the same screenshot: a 0DTE contract bought at $0.20 and sold at $8.00, captioned “turned $200 into $8,000 — zero-day options are how ordinary people get rich.”

The comment section is full of “teach me.” And every so often someone DMs me the same question: can zero-day options work as a long-term, consistently profitable strategy?

My answer is one line: Don’t ask whether it works. Ask what happened to the people who never posted.

My Thinking

I went through every hard dataset I could find. The conclusion is brutal but clear.

First dataset: academic research. After Cboe introduced daily expirations in May 2022, a widely cited paper (SSRN, “Retail Traders Love 0DTE Options… But Should They?”) tracked retail P&L: retail traders lose $350,000 a day, on average, trading 0DTE. The breakdown is the interesting part — buyers lose $8.05 per contract on average, sellers make $4.55 per contract after fees. The problem: retail overwhelmingly prefers to buy. The sellers’ thin positive edge gets drowned by the buyers’ large negative one.

Second dataset: exchange statistics. Cboe’s own report shows 0DTE volume grew five-fold in three years, averaging nearly 2 million contracts a day, with retail making up 50-60%. The report praises retail traders as “sophisticated and disciplined.” Discount that sentence heavily — the exchange earns fees on every trade. Of course it wants this game running forever.

Third dataset: backtests. SPX 0DTE backtests from 2023 to 2026, thousands of trades: buying ATM calls for direction wins about 28% of the time and loses 0.18R per trade on average; selling 10-delta spreads wins about 78% of the time but earns only 0.06R per trade.

Put the three together and the full picture emerges:

Buying is a lottery ticket: 28% win rate, theta evaporating by the hour, the bid-ask spread eating double digits of your premium. A lottery ticket doesn’t become a salary just because you can buy one every day.
Selling is a thin blade: winning 78% of days sounds beautiful. But the profit/loss ratio is inverted — the average loss is about 2.5x the average win. One fat-tail day gives back eight wins, plus interest. And that thin 0.06R edge? Commissions eat half of it.

The Insight

0DTE riches are survivorship bias at its purest. The only honest use of 0DTE options is as insurance, not as an investment.

The bias has three layers, each more subtle than the last:

Layer one: posting bias. Losses are 100% silent — what is there to post about going to zero? Gains are 5000% loud — you must post them. Tens of thousands of contracts quietly expire worthless every day; one lucky soul turning $200 into $8,000 is enough to flood your feed. You are not seeing the distribution of a strategy. You are seeing its loudest survivor.

Layer two: dropout bias. The blogger “earning $100 a day” selling 0DTE blows up on a volatility spike and stops posting. The algorithm quietly promotes the next blogger still inside their winning window. The sample purges its failures in real time; you only ever see the living.

Layer three: ecosystem incentives. While researching, I found a righteous-looking article titled “Retail loses $350K a day — Wall Street is taking your money.” Halfway through, I realized it was an ad for an e-book. Course sellers, subscription vendors, data-tool vendors, and the exchanges themselves — everyone in this industry makes money when you keep trading, and no one makes money when you make money.

One counterintuitive note: 0DTE sellers and my PMCC/Wheel system harvest the same thing — the variance risk premium. The difference is I play the 30-45 day “slow table” while 0DTE is a same-day “fast table.” Changing tables doesn’t change the game; it only changes how much faster the odds turn against you.

Practical Advice

Ask three questions before believing any screenshot: How much capital? How many total trades? What’s the total P&L? Screenshots only answer “how much did this one win” — which happens to be the least important number.
If you must trade 0DTE: sell only, defined-risk spreads only, position size under 1% of net worth, close same-day, never overnight — and run the commission math before entering. Many strategies cancel themselves once you do.
The right way to use it: on FOMC, CPI, or earnings days, buy a cheap 0DTE protective put as one-day insurance for your holdings. Pay only for the time you need — the one use case Cboe and every serious study endorse.
Keep “win rate” and “expectancy” separate: a 78% win rate is not an edge. +0.06R is, and it barely survives commissions. Next time you’re tempted, ask yourself — am I here to collect premiums, or to buy lottery tickets?

Trading Insights #006 · To be continued

On the Day NOK Plunged 13%, I Bought Back Two Short Calls: $166 Banked Early

Mon, 14 Sep 2026 00:00:00 GMT

Monday, September 14. Three headwinds stacked up, and NOK crashed 13.30% on the day to close at $9.65:

An AI-infrastructure repricing: over the weekend, Anthropic CEO Dario Amodei called for slowing frontier AI development (endorsed by Altman and Musk), triggering a sector-wide de-rating of AI data-center names. NOK has been re-rated by the market as an “AI-infrastructure stock” (optical networks, IP routing, data-center switching) and was already down about 40% from its $17.45 June high — high-multiple momentum stocks are the most sensitive to exactly this kind of headline;
The China exit confirmed: Nokia confirmed it would close nearly all of its mainland China operations by year-end — 1,600 layoffs at the Hangzhou R&D center, with Beijing, Chengdu, Qingdao and Shanghai sites also on the closure list. China revenue shrank from a €2.2B peak to €913M (after losing the 2020 5G tenders, the market was gradually taken over by Huawei and ZTE);
The Modulate Space deal collapsed: a potential business combination with satellite-communications company Modulate Space fell apart, removing another piece of the growth story.

On a crash day, the loudest topic online is “how much pain the put sellers are in.” My portfolio has put legs hurting too. But the first thing I did that day was buy back two short calls — while others stared at the wound, I banked whatever cash was ready to bank.

1. The Ledger First

Ledger detail (click to expand)

Time (ET)	Contract	Side	Qty	Avg sell	Buyback	Cash out	Realized
10:08	NOK SEP 25 ‘26 12C	Buy to close	10	$0.179	$0.06	-$70.30	+$108.70
10:10	RGTI SEP 25 ‘26 18C	Buy to close	2	$0.453	$0.155	-$33.02	+$57.58
					Total	-$103.32	+$166.41

Realized total is the broker’s settlement figure, +$166.41; the per-leg sum is $166.28, a $0.13 difference from rounding of sell-side commissions.

The NOK 12C was built in two batches (the last 4 on 9/11 at $0.18), averaging $0.179; the RGTI 18C went on during the 9/8 quantum rally, averaging $0.453. Both were covered calls — the NOK 12C fully covered by 10 LEAPs (a diagonal spread), the RGTI 18C covered by 200 shares. No naked calls.

2. Why Buy Back Instead of Waiting for Expiry

The math is simple. NOK 12C: sold at an average of $0.179, bought back at $0.06 — 66% of max profit is already in hand, with only 33% of the premium left on the table (10 contracts ≈ $60). Same for the RGTI 18C: sold at $0.453, bought back at $0.155 — 66% banked, 34% left (2 contracts ≈ $31).

One question settles it: for that $91 sitting on the table, what am I willing to carry over the next 11 days?

The FOMC meeting opens tomorrow. The market is pricing a 60-82% chance of a rate hike (the first since 2023), and the decision lands 9 days before the 9/25 expiry. Any directional shock brings OTM calls back to life.
A -13% single-day move in NOK just happened. The 12C has a 24% buffer to the close, the RGTI 18C 17% — it looks thick, but in this volatility regime a thick buffer is one snap-back rally away.
The buyback also frees margin — but only for the NOK 12C. The 12C is covered by 10 LEAPs (a diagonal), yet IBKR still charges maintenance margin on it at spread terms (about $200 per contract × 10 = $2,000), which the buyback releases. The RGTI 18C, by contrast, is a stock-covered call that ties up no extra option margin — buying it back leaves the shares and their margin untouched. The real margin hogs are the put legs, so this reason only deserves third place.

Profit-taking on short calls should not come from “waiting for expiry” — it comes from one comparison: is the premium left on the table worth the risk you keep carrying? With a third of the premium left and an FOMC ahead, there was nothing to think about.

3. RGTI: The Stock Is Losing, the Call Already Won

This one deserves its own section, because it shows the difference between single-leg thinking and portfolio thinking.

Look at the stock alone: 200 shares at a $17.86 cost against a $15.43 spot — an unrealized loss of about -$486. Ugly.

Now add the option leg: the 2x 18C sold on 9/8 collected roughly $92 in premium; today I bought them back at $0.155, banking +$58. While the stock was taking hits, the call leg was collecting. “You get paid even when the stock doesn’t rise” is the whole point of a Covered Call — and this fill was a textbook cash-in.

The same decline looks like tragedy leg by leg. Put the legs back into the portfolio and you see the hedge has been alive all along.

4. The Other Side of the Same Crash

This review is not going to show only the winning half.

The same day, my NOK 10.5P flipped from OTM to ITM — the put-selling legs were bleeding; the account’s total PnL for the day was -$720. The $166 I banked doesn’t mean “I won today.” It means it was the only money I proactively took off the table on a crash day — it can’t fill the hole the puts dug, but it proves something else: the two-sided nature of selling options is itself a hedge. The call legs realized their profit faster than usual precisely because they fed on this very decline.

Don’t play dead when losing, don’t brag when winning — that’s the only way the ledger stays trustworthy.

5. Three Takeaways

Profit-taking comes from comparison, not from waiting. “Just let it expire for free” sounds free, but it is really re-exposing profit you already have to the market. The right questions come in threes: How much premium is left? What events are ahead? Do I need the margin? The answers decide between buyback and waiting.
The first reaction on a crash day is not panic — it’s opening the ledger. Every leg reacts to a decline differently: stock down, puts hurting, calls accelerating toward max profit. Scan them once and you’ll know which action is with the wind. That day, buying back calls was the only one.
Portfolio thinking is the only cure for single-leg fear. RGTI’s -$486 stock loss and +$58 call gain belong on the same row; so do NOK’s -$1,071 LEAP loss and +$109 12C gain. Leg by leg it’s all blood; as a portfolio it’s net equity.

A hunter doesn’t fear the storm. He fears not knowing, in the middle of it, which hand is collecting cash and which hand is bleeding. This time, I had a clear look at both.

Current state: NOK 12C×10 and RGTI 18C×2 fully bought back; remaining short calls are TRON 5C×17 (covered by 2,300 shares); remaining short puts are NOK 10.5P×4 + 10P×6 + QUBT 8P×2. Account leverage ≈ 1.56x, cash -$1,619.

Open items:

NOK 10.5P (9/25) ×4 now ITM — buy back or manage to expiry; written decision by 9/23 at the latest
NOK 10P (9/18) ×2 — the FOMC decision lands 2 days before expiry; buy back first
QUBT 8P (9/18) ×2 — back OTM, close for about $48 to cut the assignment tail