PORTFOLIO RESET · INCOME + CONVEXITY + INDEX + CASH
Built 2026-11-04 00:10 SGT on the Friday 30 Oct 2026 close · NLV $2,534,530 · income from the INCOME_BOOK pin (fortress_v28 --live, Sat 11:00 SGT), HBMax from the 2 Nov RTH run and after-hours runs of 3 Nov, and your 90-day Coverage INSIGHTS · crypto from CBOE close quotes · backtest 28 Aug to 30 Oct on Yahoo closes · read-only, zero IBKR, no orders
In plain words
This was a factor crash, not a market crash. The S&P is only 3% off its 6 Oct high, but MARA fell 50%, IREN 51%, SNDK 39%, HOOD 32% and DELL 28% from their October highs. Your NLV went from $3.56M on 28 Aug to $2.53M on 30 Oct, -27.7% after the one withdrawal, while SPY was -1.5%. The book was a 3.7x leveraged bet on high beta and crypto, and the Risk Cockpit, which moves everything by its S&P beta, understates that kind of fall.
The plan fits. $2.53M of NLV less about $125k to close today's book and open the new one, plus about $566k brought in (the BTC swap's cash after the bets, and spare cash), pays for the $940k income account (Max Loss $866k held as cash), $1.20M of index ($929k in shares (738 SPY, 538 QQQ) + $266k in Dec '28 LEAPS (12x SPY 735C, 8x QQQ 675C), bought in one go, weekly calls on every lot and LEAPS, written your way) and $340k of crypto (500 MARA + 250 BMNR, 250 more at BMNR $30: Dec '28 / Jan '29 calls, weekly calls on a quarter of them), with $500k kept in cash.
Income: HBMax pays $22,943/mo at the INCOME_BOOK basis on $866k of Max Loss, about $11,248/mo at your 90-day rate. Replaying its own calls on the real paths, it made $35,628/mo over the last 90 days and averaged $30,648/mo over two years (both at a normal week's volatility, earnings weeks excluded). Your $30-40k comes from those calls plus the SPY/QQQ calls, interest and the reserve (Income tab).
Crash levels improve, but the drawdown is still deep. At S&P −20% the new portfolio is worth about $1.84M (-38%) against today's book at $1.03M (-60%), same model. Nothing in it can be force-sold to −95%. The income book still loses most of its Max Loss in a −20% to −30% market, because that is what Max Loss means.
Sep and Oct, replayed: the new portfolio built on 28 Aug ends -5.1% (income -$4k, crypto -$94k, index -$55k), against your real -27.7%. The crypto sleeve is the drag, by design: its calls can only lose what they cost.
My recommendation: reset, in stages (Hold vs Reset tab). Swap crypto first, then close the high-beta fortresses on up days, buy the crypto sleeve and the index in one go right after the reset, then build the income book over two to three weeks. Total book income of the plan: $49,087/mo (Total Income tab).
The reset portfolio
NLV at the 30 Oct close $2,534,530 (Main $1.19M, Joint $578k, Neville $458k, RetireInc $307k). Friction = closing today's 156 option legs (4,109 contracts) and opening the new ones a quarter of the spread from the mid, how the Desk's NUDGE fills ($247k at the full natural), plus about $0.70 a contract.
Income book
$940k
HBMax NLV · Max Loss $866k held as cash 13 fortresses, $22,943/mo (book basis), $11,248/mo (your rate)
Convexity
$340k
500 MARA + 250 BMNR, 250 more at BMNR $30, Dec '28 / Jan '29 cost = Max Loss; weekly calls on 25%: ~$9k/mo gross
Index
$1.20M
$929k in shares (738 SPY, 538 QQQ) + $266k in Dec '28 LEAPS (12x SPY 735C, 8x QQQ 675C) 60% SPY / 40% QQQ, bought in one go
Cash reserve
$500k
$1,408/mo at 3.38% (IBKR bench −0.5%)
Friction
$125k
quarter spread; $247k at the natural
Brought in
$566k
$643k freed by the BTC swap − $318k of convexity bets, + $241k of your spare cash funds the index and reserve above NLV (Convexity tab: 300x IBIT Dec '28 90C)
Monthly income of the whole reset:$49,087 on the book basis, everything in (Total Income tab: HBMax $22,548, SPY/QQQ calls $15,616, crypto calls $2,483, IBIT calls $2,453, bet calls $3,347, interest $2,640); plan on about $32k to $44k. The income book alone: $25,583 on the book basis (calls $22,943 + idle cash in the income account $1,232 + reserve $1,408), or $13,888 at your 90-day hit rate. Today's book earned $64,732/mo of calls over the last 90 days and paid about $12,587/mo of cash hedge rent, so about $52,146/mo net, on $3.17M of fortress Max Loss. The reset keeps about the same income only if you write at the book basis; at your current hit rate it earns less than half, on less than a third of the Max Loss. The crypto weekly calls are about $9k/mo gross but roughly break-even over time on the real path (Convexity tab).
Crash levels: what each S&P drop leaves
The Risk Cockpit's own model (module_risk_cockpit, each name moves by its cockpit beta times the drop, volatility held still as IBKR's What-if does, fortresses stop at Max Loss, hedges pay). Today's book from Repair's What-if legs of 30 Oct 15:50 ET. The reset row includes the $566k brought in from outside IBKR (today's book row does not hold the BTC), with the whole $1.2M index in. Names the cockpit has no measured beta for (XOM, LMT, LLY, JPM, CAT, XBI) move at its default 1.5, which is harsh: this year XOM, LMT and LLY rose on the S&P's worst days.
Account value
S&P now
S&P -10%
S&P -20%
S&P -30%
S&P -40%
Today's book
$2.56M
$1.65M
-35%
$1.03M
-60%
$811k
-68%
$851k
-67%
Today's book, Risk Cockpit (IBKR-matched, 29 Oct close)
$2.68M
$1.70M
-37%
$989k
-63%
$741k
-72%
$768k
-71%
Reset portfolio
$2.98M
$2.34M
-21%
$1.84M
-38%
$1.52M
-49%
$1.32M
-55%
income book
$940k
$578k
$302k
$144k
$79k
crypto calls
$340k
$238k
$171k
$149k
$149k
index
$1.20M
$1.02M
$870k
$728k
$596k
cash reserve
$500k
$500k
$500k
$500k
$500k
Reset portfolio, measured betas (250d, floor 0.5)
$2.98M
$2.41M
-19%
$1.97M
-34%
$1.68M
-44%
$1.47M
-51%
The reset is not a low-volatility portfolio. At S&P −10% it still loses 21% of its value (today's book 35%), and a +10% bounce adds about $723k (today's book $818k, both before calls cap the gain). What changes is the deep end and the path: the loss at −20% falls from 60% to 38% of NLV, nothing can be force-sold, and a crypto or high-beta crash like this one costs a capped amount. At −20% the loss comes from the income book (-$638k), the crypto calls (-$168k) and the index (-$325k). The two knobs that move it are the size of the income book and how much of the index is in when the fall comes.
AppendixSep and Oct 2026, replayed
Sep and Oct 2026, replayed
The reset built on the 28 Aug close (same moneyness and tenors, same contracts, volatility held still), revalued on every daily close to 30 Oct. Income book at your 90-day hit rate; crypto calls at the same moneyness and dollars; index 60/40 SPY/QQQ, shares at the real path and LEAPS at the same moneyness and dollars; reserve at 3.38%. Real book = NLV ledger, net of the 18 Sep withdrawal.
Piece
Start (28 Aug)
P&L to 30 Oct
Return
Income book (HBMax)
$940,000
-$4,434
-0.5%
Crypto calls (500 MARA + 250 BMNR, 250 more at BMNR $30)
$339,500
-$94,215
-27.8%
Index
$1,195,654
-$54,886
-4.6%
Cash reserve
$500,000
$2,917
+0.6%
Reset portfolio
$2,975,154
-$150,619
-5.1%
Real book
$3,555,262
-$985,828
-27.7%
Read the replay with care. It is hindsight on one regime: oil and defence rallied while high beta and crypto broke. That is close to a bad case for the tech and crypto half of this income book and for the crypto sleeve; its XOM, JPM, AXP, MCD, PEP, LLY, ABT and CI half did well in it. Volatility rose in October and the replay holds it still, which flatters the short puts a little. It shows the shape of the risk, not a forecast.
Monday plan, in three phases.Phase 1, right after the reset: close every position (the Desk's RESET page tickets; about $108k of spread and commissions at a quarter of the spread), move the cash, then at once buy the crypto sleeve in Neville and the index in Joint (in one go, as far as the cash goes), so a bounce in MARA, BMNR or the market reaches you. Phase 2, the following weeks: the income book in Main, name by name. Phase 3, when the BTC sale lands: the IBIT calls and the three bets in Neville (IREN is separate: on REBOOK now, from Neville's own cash), and the rest of the index shares in Joint the same day. Each account is closed out before it buys, so nothing is bought next to old positions.
Phase 1: cash moves
Cash after the reset = the 30 Oct NLV less that account's close cost. + money in, − money out.
Account
Cash after reset
Move
Holds then
Main
income book, phase 2
$1.13M
-$173k
$953k
Joint
index + $500k reserve
$559k
+$812k
$1.37M
Neville
crypto sleeve now
$442k
-$98k
$344k
RetireInc
closed
$299k
-$299k
$0
Main → Joint: $173k. Main keeps $953k for the income book.
Neville → Joint: $98k. Neville keeps $344k for the crypto sleeve.
RetireInc → Joint: $299k, all of it; then close the account.
Your spare cash → Joint: $241k.
Check that IBKR allows each internal transfer: between accounts with different owners (Joint, or any account not in your sole name) it may have to go out and back in as a withdrawal and a deposit.
Phase 1: buy right after the reset
Neville, the crypto sleeve ($203k): 250x BMNR 40C Jan '29 and 500x MARA 25C Dec '28, once Neville's own positions are closed; the other $136k stays there as cash for 250 more BMNR calls if BMNR reaches $30, strike your call that day, on its own ticket. MARA reports on 4 Nov: a Dec '28 call carries little earnings premium, so buying before is a bet on the report rather than a cost; half before and half after splits it.
Joint, the index in one go ($871k): all the LEAPS (12x SPY 735C, 8x QQQ 675C) and 65% of the shares (480 SPY, 350 QQQ): everything Joint can pay for with the $500k reserve untouched. A long-term hold with no margin loan, so a further fall cannot force a sale; no waiting for pullbacks. The last $325k of shares waits only for the BTC money (phase 3).
Phase 2: the following weeks
Main, the income book (HBMax, $940k): one name at a time over two to three weeks, every call written the first Friday. First: COIN, XOM, SPCX, LLY, MU, ABT, JPM, AXP, PEP. Then: MCD after its 4 Nov report, CI after its 5 Nov report, NVDA after its 17 Nov report, DELL after its 24 Nov report.
Every order is yours: the Desk or TWS. Nothing on this page places one.
Phase 3: when the BTC sale lands ($767k)
Neville: deposit $442k; buy the IBIT calls (300x 90C Dec '28, $125k) and the three bets ($318k). IREN is separate: on REBOOK now, from Neville's own cash.
Joint: deposit $325k and buy the rest of the shares ($325k: 258 SPY, 188 QQQ) the same day.
Appendixthe 31 Oct partial derisk (Plan D) and its account checks
The plan: close 18 positions on Monday, keep the rest for the later reset. Every crypto holding that writes no calls (BMNR, MARA, IBIT, MSTR and the small ones), GLD, MU in Main and MU and GOOG in Neville, and buy the crypto sleeve (500 MARA + 250 BMNR, 250 more at BMNR $30). That releases $1.43M of forward Max Loss for about $27k of friction, so the $500k of SPY/QQQ (shares + LEAPS) and the $500k reserve are fully outside the book's worst case, with $226k to spare. It keeps $64k/mo of the calls your book earned over the last 90 days (of $82k). S&P −20% leaves about $1.55M instead of $1.03M.
What "secured" has to mean
Your accounts already hold $1.36M of cash (Main $844k, Joint $443k on margin, Neville $155k, RetireInc $0), but all of it is behind the book: its forward Max Loss, the further loss from today's prices if every name kept falling with the hedges paying, is about $2.39M ($2.54M on Repair's own count). So the $1M is only secure when the book left behind fits inside what is left after setting it aside: NLV minus friction minus $1M has to cover the remaining forward Max Loss. Raising cash alone does not do it; shrinking the Max Loss does.
Four ways to do it
Each plan also buys the $500k of SPY/QQQ (its crash loss included). Crash numbers: the Risk Cockpit's S&P −20% and "October again" (each name's own 6-30 Oct move, from today). Calls = the last 90 days' call income of the positions kept, before hedge rent.
Plan
Forward Max Loss released
Friction
$1M secured? (room to spare)
October again
NLV at S&P −20%
Calls / mo kept
Positions / contracts
Today
–
–
$839k short
-$755k
$1.03M
$82k
–
A. Crypto holdings only, no sleeve yet
$829k
$13k
-$24k
-$529k
$1.41M
$79k
14 / 2,111
B. A + GLD + MU Main
$1.24M
$16k
+$388k
-$429k
$1.69M
$71k
16 / 2,171
C. B, and buy the crypto sleeve
$1.24M
$22k
+$43k
-$576k
$1.40M
$71k
16 / 2,171
D. C + MU and GOOG in Neville plan
$1.43M
$27k
+$226k
-$500k
$1.55M
$64k
18 / 2,259
Plan D: the positions
Account
Name
Forward Max Loss
Cash raised
October again
S&P −20%
Calls / mo given up
Contracts
Main
MU
$309k
-$24k
-$70k
-$258k
$7,442
20
Main
BMNR
$217k
$293k
-$58k
-$143k
–
375
Joint
IBIT
$189k
$248k
-$24k
-$61k
–
58
Joint
BMNR
$139k
$362k
-$58k
-$119k
$1,741
532
Main
MARA
$122k
$123k
-$64k
-$92k
–
400
Neville
MU
$111k
-$30k
-$28k
-$97k
$3,462
8
Main
GLD
$106k
-$14k
-$30k
-$26k
$1,188
40
Neville
GOOG
$78k
$2k
-$48k
-$58k
$3,695
80
RetireInc
MSTR
$40k
-$12k
-$13k
-$33k
$387
16
Main
MSTR
$38k
$39k
-$8k
-$22k
–
shares
Joint
MSTR
$36k
$279k
-$17k
-$34k
–
30
Main
SBET
$20k
$20k
-$6k
-$9k
–
shares
RetireInc
MARA
$16k
$16k
-$12k
-$15k
–
500
Main
BTCS
$4k
$4k
-$1k
-$2k
–
shares
Main
MSTX
$3k
$3k
-$1k
-$3k
–
shares
Joint
CAN
$2k
$2k
-$759
-$1k
–
shares
Joint
MSTX
$1k
$1k
-$507
-$1k
–
shares
Joint
BTBT
$129
$20k
-$129
-$129
–
200
Account by account
Accounts do not share margin. The sleeve sits in Neville (your convexity account); the last column assumes the SPY/QQQ is bought in Joint, as in your two-account plan.
Account
NLV
Cash
Forward Max Loss
Room left (NLV − Max Loss − index)
Calls / mo given up
Main
$1.20M
$830k → $1.27M
$1.25M → $425k
+$772k
$8,631
Joint
$590k
-$381k → $530k
$529k → $161k
-$766k
$1,741
Neville
$460k
$469k → $102k
$405k → $556k
-$97k
$7,157
RetireInc
$308k
$446k → $450k
$215k → $159k
+$149k
$387
Joint cannot take the index on Monday. Its other seven positions (HOOD, AMZN, COPX, GOOG, IREN, TSLA, QUBT) still carry forward Max Loss, and after the index there is not enough left behind them. Buy the SPY/QQQ in Main first (it has the room), and move it to Joint once Joint's other positions are closed in the second stage.
Monday sequence
US clocks change on 1 Nov: the open is 22:30 SGT from Monday.
At the open: sell the crypto holdings (shares first, then the LEAPS as combos with NUDGE), GLD and MU in Main. MU does not report until 23 Dec.
Same session: buy the sleeve's BMNR leg (250x 40C Jan '29). BMNR reports 20 Nov.
Thursday 5 Nov: buy the MARA leg (500x 25C Dec '28) after MARA reports on 4 Nov. Until then you have no MARA: that is the price of not paying pre-earnings volatility.
SPY/QQQ $500k: $929k in shares (738 SPY, 538 QQQ) + $266k in Dec '28 LEAPS (12x SPY 735C, 8x QQQ 675C), in Main, in thirds if you like; move to Joint later.
Tuesday or Wednesday: MU and GOOG in Neville.
Next two to three weeks: the rest of the reset (the Income tab's book), into up days.
Every order is yours: the Desk's Portfolio (TWS) tab or TWS. Nothing on this page places one.
Why derisk first
Pro: it secures the $1M for about $27k, mostly in liquid shares and LEAPS. The thin, expensive legs (GLXY, COIN, ASST) that carry most of the full reset's friction wait.
Pro: it cuts what pays least for its risk: the crypto holdings earn about $2k/mo of calls on about $830k of forward Max Loss, GLD earns 6%/yr on Max Loss, and MU in Main alone loses about $258k at S&P −20%.
Pro: it keeps $64k/mo of the calls, enough for your $30-40k without touching the reserve while the second stage happens.
Pro: Joint's margin loan is repaid by its own crypto sales.
What it leaves
Con: another October still costs about $500k: COIN, GLXY, IREN, SNDK, DELL, NVDA, ASST and CRCL stay until the second stage.
Con: the glory path shrinks to the 500 MARA + 250 BMNR, 250 more at BMNR $30 sleeve, and MARA is out for three days around its earnings.
Con: about 30 positions in four accounts remain; complexity only falls in the second stage.
Con: it sells after the fall; a crypto bounce next week reaches you only through the sleeve.
Your income book: HBMax. DELL 3, COIN 8, MCD 10, XOM 13, SPCX 8, NVDA 7, CI 5, LLY 1, MU 1, ABT 12, JPM 4, AXP 3, PEP 12. Max Loss $866,171, held as cash in a $940,000 account ($503k invested, $437k idle); every short put hedged and rolled. Liquidation: none in the six-attack battery to −95%. Income after hedge rolls: $22,548/mo on the book basis. Plan on $12,521 to $17,398 (the scanner pick to the max-income pick, each held to expiry, in the backtest below); my best estimate at your own deltas and real rates is about $23,300 (about $10,900 of it from eight names with no record of yours); the 90-day rate on the same names would be about $10,853.
HBMax · Max Loss $866,171 · account $940,000
Name
Type
Contracts
Spot
LC / SP / HP
Invested
Max Loss
Notional
Calls / mo book
Calls / mo your rate
Beta
DELL
DIAMOND
3
417.75
280 / 410 / 230
$41,250
$95,250
11% of book
$125k
$2,970
$1,836
2.51
COIN
PIONEER
8
140.77
105 / 150 / 55
$19,192
$95,192
11% of book
$113k
$2,916
$1,733
2.97
MCD
BALANCED
10
264.60
200 / 250 / 220
$59,640
$89,640
10% of book
$265k
$1,641
$857
-0.01
XOM
BALANCED
13
179.66
150 / 175 / 145
$38,012
$77,012
9% of book
$234k
$715
$715
-0.72
SPCX
BALANCED
8
141.51
65 / 105 / 90
$60,016
$72,016
8% of book
$113k
$4,043
$1,430
3.18
NVDA
BALANCED
7
209.20
145 / 195 / 160
$44,590
$69,090
8% of book
$146k
$1,390
$625
1.94
CI
BALANCED
5
294.70
190 / 250 / 220
$51,825
$66,825
8% of book
$147k
$778
$382
-0.01
LLY
BALANCED
1
1408.22
900 / 1270 / 1120
$50,600
$65,600
8% of book
$141k
$1,314
$582
0.20
MU
BALANCED
1
880.03
430 / 630 / 520
$44,560
$55,560
6% of book
$88k
$3,820
$1,461
3.40
ABT
BALANCED
12
110.03
92.5 / 110 / 85
$20,580
$50,580
6% of book
$132k
$809
$408
-0.02
JPM
BALANCED
4
340.39
280 / 340 / 280
$24,076
$48,076
6% of book
$136k
$1,023
$527
0.75
AXP
BALANCED
3
321.38
220 / 320 / 270
$30,066
$45,066
5% of book
$96k
$594
$276
0.94
PEP
BALANCED
12
140.54
125 / 135 / 120
$18,264
$36,264
4% of book
$169k
$930
$416
-0.24
HBMax
87
$502,671
$866,171
$1.91M
$22,943
$11,248
How to write the calls
Each Friday sell the scanner's balanced pick, about 5-8% above the stock, 5-7 days out. Not your usual further-out strike: it earns about half.
Hold it to expiry, even when it is challenged. Ignore Rolls' ROLL NOW on an income fortress. In the last half hour of expiry day (4:30-5:00am SGT in US winter time): out of the money, let it go or buy it back for a penny; in the money, buy it back at about intrinsic and sell next week's pick. Above CC-SS the long call has already paid for that loss.
Exceptions: earnings inside next week (sell nothing into the print; a roll across it is the fallback), the stock below CC-SS (FIGHT's job), a deep in-the-money call on a dividend payer the day before the ex-date (buy it back that day).
The upside is not this book's job. It belongs to the convexity sleeves; capture a big run by rebuilding the fortress (Rolls), not by parking calls far out.
Weekly calls Jun-Oct 2026 on the HBMax names, priced on the scanner ledger's real implied vol (the eight buffer names have no ledger quotes, so theirs is realized vol x the cross-name ratio), sold at 97% of model and bought at 103%, earnings weeks skipped, CC-SS not modelled (strat_backtest.py --tape). On your own log, the 30 calls you bought back early at a loss since June cost $96k; at expiry they would have cost $17k.
How the weekly call is written and handled
HBMax calls / mo
Scanner pick (~5-8% out, 0.22 delta), held to expiry
$17,518
the plan
Max-income pick (~2-5% out, 0.32 delta), held to expiry
$22,396
more income, about twice the buybacks
Scanner pick, Rolls' credit roll (up and out, within 30 days) when challenged
$13,273
what ROLL NOW leads to
Scanner pick, bought back as soon as it is challenged
$10,156
the worst: pays the call's time value at its dearest
Your usual distance (further out, 0.12 delta), held to expiry
$9,092
about half the scanner pick
Appendixthe alternative book, the income yardstick, every name, stress tests, backtests and feedback
Your pick: HBMax, DELL, SPCX, MU, NVDA and COIN for income, JPM, AXP, MCD, PEP, XOM, LLY, ABT and CI as the non-tech buffer, inside $875k of Max Loss (owner, 3 Nov): DELL 3, COIN 8, MCD 10, XOM 13, SPCX 8, NVDA 7, CI 5, LLY 1, MU 1, ABT 12, JPM 4, AXP 3, PEP 12. Max Loss $866,171 held as cash (account $940,000); every attack in the stress battery comes back NONE. Calls: $22,943/mo at the INCOME_BOOK basis and $11,248 at your 90-day rate; replayed on the real paths, $30,648/mo on the two-year average. Every premium is taken at a normal week's volatility: the 13 Nov calls straddle earnings for MCD (x0.69) and CI (x0.48). XOM is carried at about $50 a contract a month after hedge, its real-vol replay, not the pooled ratio. Best estimate at the deltas you actually sell and your real rates: about $23,300/mo, of which about $10,900 comes from eight names with no call record of yours. The alternative built on your own record (SPCX 10, AMZN 6, COPX 17, RKLB 20, COIN 6, SNOW 3, CRWV 17, DELL 2, MU 1, MSTR 4, SHOP 3) has semis at 16% of Max Loss, beta 2.67, and pays $25,467 at the INCOME_BOOK basis, $11,937 at your rate and $47,076/mo on the two-year replay. The full comparison, every test and the name-by-name screen are on the Income HB750 page.
First: the income yardstick, checked against your own book
The INCOME_BOOK prices income at about twice what your fortresses have actually earned. Its basis is each name's call yield at the time you wrote (the trade log since 1 Jun) divided by the model's, 0.52 pooled. Measured instead per contract over every day of the last 90 (Coverage INSIGHTS: calls closed plus open marked to close, idle weeks and buybacks at a loss included), your book earned 0.32 of the model's kept premium on the names both cover. Book-wide: $64,732/mo of calls on $4.84M of stock, 1.34%/mo, or 25%/yr on $3.17M of Max Loss before hedge rent. HBMax assumes about 32%/yr on Max Loss after rent, at the INCOME_BOOK basis.
Per contract, every name in the universe. Book ratio = the INCOME_BOOK's. Your 90-day ratio = income per contract over the window as a yield on the average close, over the model's kept yield. Used = yours shrunk toward the pooled 0.32 with 10 pseudo-trades, never above the book ratio; a name you have not traded takes the pool. Net = kept calls minus hedge rent. Beta = 250 days vs SPY. Sorted by what each $ of Max Loss earns at your rate.
Name
Max Loss / ct
Book ratio
Your 90-day ratio
Trades (90d)
Used
Net / ct / mo book
Net / ct / mo your rate
Yield on ML book
Yield on ML your rate
Beta
Your realized % / yr on ML
MU
$55,560
0.77
0.27
39
0.32
$3,820
$1,461
83%
32%
3.40
22%
SNDK
$108,450
0.68
0.44
7
0.32
$5,721
$2,598
63%
29%
4.12
61%
CRWV
$4,017
0.52
0.02
4
0.32
$184
$82
55%
24%
3.67
2%
SPCX
$9,002
0.78
0.51
13
0.32
$505
$179
67%
24%
3.18
49%
DELL
$31,750
0.47
0.81
6
0.32
$990
$612
37%
23%
2.51
98%
RKLB
$4,384
0.52
0.59
16
0.32
$138
$82
38%
23%
3.88
20%
META
$28,535
0.40
no history
0
0.32
$740
$526
31%
22%
1.41
–
COIN
$11,899
0.52
0.49
30
0.32
$364
$217
37%
22%
2.97
35%
HOOD
$5,602
0.62
0.56
7
0.32
$222
$93
47%
20%
3.17
55%
ORCL
$7,810
0.52
no history
0
0.32
$232
$129
36%
20%
2.28
–
CAT
$27,595
0.52
no history
0
0.32
$930
$420
40%
18%
1.90
–
NOW
$6,445
0.52
no history
0
0.32
$172
$94
32%
17%
0.74
–
XBI
$7,697
0.52
no history
0
0.32
$195
$111
30%
17%
1.08
–
LMT
$21,150
0.52
no history
0
0.32
$588
$304
33%
17%
-0.09
–
MSTR
$9,632
0.52
0.16
4
0.32
$245
$129
31%
16%
3.19
8%
CRWD
$19,510
0.52
no history
0
0.32
$423
$247
26%
15%
1.55
–
INTC
$5,861
0.52
0.25
8
0.32
$124
$71
25%
15%
3.38
13%
AVGO
$16,660
0.52
no history
0
0.32
$393
$202
28%
15%
2.21
–
AMZN
$14,876
0.69
0.18
7
0.32
$443
$177
36%
14%
1.44
16%
GOOG
$11,355
0.47
0.18
28
0.32
$264
$133
28%
14%
1.39
21%
PEP
$3,022
0.52
no history
0
0.32
$78
$35
31%
14%
-0.24
–
GLD
$14,515
0.48
0.13
1
0.32
$300
$164
25%
14%
0.90
6%
JPM
$12,019
0.52
no history
0
0.32
$256
$132
26%
13%
0.75
–
SNOW
$22,985
0.52
no history
0
0.32
$420
$240
22%
13%
1.41
–
CRM
$12,696
0.52
no history
0
0.32
$240
$125
23%
12%
0.42
–
MCD
$8,964
0.52
no history
0
0.32
$164
$86
22%
11%
-0.01
–
ASST
$1,858
0.52
no history
0
0.32
$34
$18
22%
11%
3.52
–
IREN
$1,892
0.26
0.16
74
0.26
$18
$18
11%
11%
4.09
19%
XOM
$5,924
0.23
no history
0
0.23
$55
$55
11%
11%
-0.72
–
NVDA
$9,870
0.52
0.35
7
0.32
$199
$89
24%
11%
1.94
29%
LLY
$65,600
0.52
no history
0
0.32
$1,314
$582
24%
11%
0.20
–
ABT
$4,215
0.52
no history
0
0.32
$67
$34
19%
10%
-0.02
–
GLXY
$1,841
0.52
0.27
11
0.32
$34
$15
22%
10%
4.21
18%
MSFT
$21,660
0.52
no history
0
0.32
$411
$173
23%
10%
0.99
–
TSLA
$17,480
0.52
no history
0
0.32
$419
$137
29%
9%
2.19
–
COPX
$5,240
0.52
0.49
8
0.32
$73
$36
17%
8%
2.43
26%
AXP
$15,022
0.52
no history
0
0.32
$198
$92
16%
7%
0.94
–
SHOP
$12,195
0.52
no history
0
0.32
$132
$71
13%
7%
1.86
–
CI
$13,365
0.52
no history
0
0.32
$156
$76
14%
7%
-0.01
–
DDOG
$20,375
0.52
no history
0
0.32
$229
$109
13%
6%
1.40
–
NFLX
$2,851
0.52
no history
0
0.32
$44
$15
18%
6%
0.19
–
NET
$22,375
0.52
no history
0
0.32
$211
$90
11%
5%
1.65
–
QCOM
$9,237
0.52
0.72
2
0.32
$68
$29
9%
4%
2.15
12%
UBER
$5,166
0.52
no history
0
0.32
$25
$10
6%
2%
0.81
–
AMD
$23,915
0.53
no history
0
0.32
$414
$25
21%
1%
3.30
–
APP
$16,180
0.52
1.16
8
0.32
$47
-$26
3%
-2%
2.31
14%
PLTR
$10,965
0.52
no history
0
0.32
-$157
-$209
-17%
-23%
1.98
–
Why I trust the 90-day number more, and where it may be too low.
It counts the weeks a call was not on, the buybacks at a loss and strikes written far out to avoid capping. Those are real habits and a new book inherits them.
It may be pessimistic for a fresh book: today's fortresses are mostly underwater, so CC-SS sits far above spot and the calls pay little. A new at-the-money fortress can write closer. But RetireInc, built fresh in late May, still realized only 8-20%/yr on Max Loss per name.
Your best names on your own record were DELL (98%/yr on Max Loss), SNDK, HOOD and SPCX (49-61%). Those are high beta, which is why the high-beta books look relatively better at your rate.
INCOME_GAPS found about $14k/mo your current book leaves on the table with no convexity cost (buybacks at a loss, idle weeks, strikes too far). Closing that gap is the lever for $50k.
The books
From three fortress_v28 --live runs on the 30 Oct close, with the HBMax tech names re-pinned on the 2 Nov RTH run and its eight buffer names on after-hours runs of 3 Nov (Jan 2029 LEAPS), every contract written, whole contracts, account = Max Loss / 0.93, weekly premiums at a normal week's volatility (earnings weeks corrected). Income_HB750 is your reviewed book. HB750 proven: at least 60% of Max Loss in names with a strong record in your own book since January, look-alikes only if their calls pay when replayed the way you write, semis at most 25%, crypto at most 15%, at most 12% per name, Max Loss at most $750k.
Measure
HBMax
HB750 proven
Max Loss
$866,171
within $1M
$758,847
within $1M
Account (NLV)
$940,000
$820,000
Invested / idle cash
$503k / $437k
$437k / $383k
Calls / mo, book basis
$22,943
$25,467
Calls / mo, your 90-day rate
$11,248
$11,937
Idle cash interest / mo
$1,232
$1,078
Max Loss per $1 of monthly calls
$37.8
your rate $77.0
$29.8
your rate $63.6
Names
13
11
Avg pairwise correlation (180d)
+0.04
+0.28
Beta (dollar-delta weighted, 250d)
0.84
2.67
Daily swing (1 sd, % of account)
1.87%
3.71%
Hedge rent / yr
$55,217
$42,062
Liquidation (6 attacks to −95%)
NONE
NONE
Worst simulated NLV (to −95%)
$59k
at -66%
$101k
at -66%
NLV at S&P -10% (cockpit betas)
$578k
-39%
$550k
-33%
NLV at S&P -20% (cockpit betas)
$302k
-68%
$319k
-61%
NLV at S&P -30% (cockpit betas)
$144k
-85%
$175k
-79%
NLV at S&P −20% (measured betas)
$475k
-49%
$268k
-67%
Sep-Oct, your rate
-0.5%
-19.5%
Sep-Oct, model writing every week
+4.3%
max drawdown -4.1%
-10.8%
max drawdown -25.0%
Cost to open (quarter spread)
$13,352
$15,868
Holdings
HBMax · Max Loss $866,171 · account $940,000
Name
Type
Contracts
Spot
LC / SP / HP
Invested
Max Loss
Notional
Calls / mo book
Calls / mo your rate
Beta
DELL
DIAMOND
3
417.75
280 / 410 / 230
$41,250
$95,250
11% of book
$125k
$2,970
$1,836
2.51
COIN
PIONEER
8
140.77
105 / 150 / 55
$19,192
$95,192
11% of book
$113k
$2,916
$1,733
2.97
MCD
BALANCED
10
264.60
200 / 250 / 220
$59,640
$89,640
10% of book
$265k
$1,641
$857
-0.01
XOM
BALANCED
13
179.66
150 / 175 / 145
$38,012
$77,012
9% of book
$234k
$715
$715
-0.72
SPCX
BALANCED
8
141.51
65 / 105 / 90
$60,016
$72,016
8% of book
$113k
$4,043
$1,430
3.18
NVDA
BALANCED
7
209.20
145 / 195 / 160
$44,590
$69,090
8% of book
$146k
$1,390
$625
1.94
CI
BALANCED
5
294.70
190 / 250 / 220
$51,825
$66,825
8% of book
$147k
$778
$382
-0.01
LLY
BALANCED
1
1408.22
900 / 1270 / 1120
$50,600
$65,600
8% of book
$141k
$1,314
$582
0.20
MU
BALANCED
1
880.03
430 / 630 / 520
$44,560
$55,560
6% of book
$88k
$3,820
$1,461
3.40
ABT
BALANCED
12
110.03
92.5 / 110 / 85
$20,580
$50,580
6% of book
$132k
$809
$408
-0.02
JPM
BALANCED
4
340.39
280 / 340 / 280
$24,076
$48,076
6% of book
$136k
$1,023
$527
0.75
AXP
BALANCED
3
321.38
220 / 320 / 270
$30,066
$45,066
5% of book
$96k
$594
$276
0.94
PEP
BALANCED
12
140.54
125 / 135 / 120
$18,264
$36,264
4% of book
$169k
$930
$416
-0.24
HBMax
87
$502,671
$866,171
$1.91M
$22,943
$11,248
HB750 proven · Max Loss $758,847 · account $820,000
Name
Type
Contracts
Spot
LC / SP / HP
Invested
Max Loss
Notional
Calls / mo book
Calls / mo your rate
Beta
SPCX
BALANCED
10
141.51
65 / 105 / 90
$75,020
$90,020
12% of book
$142k
$5,054
$1,788
3.18
AMZN
VAULT
6
240.80
125 / 230 / 185
$62,256
$89,256
12% of book
$144k
$2,661
$1,063
1.44
COPX
BALANCED
17
71.96
50 / 73 / 45
$41,480
$89,080
12% of book
$127k
$1,247
$613
2.43
RKLB
DIAMOND
20
55.06
40 / 50 / 20
$27,680
$87,680
12% of book
$110k
$2,752
$1,648
3.88
COIN
PIONEER
6
140.77
105 / 150 / 55
$14,394
$71,394
9% of book
$84k
$2,187
$1,300
2.97
SNOW
BALANCED
3
306.08
180 / 250 / 140
$35,955
$68,955
9% of book
$92k
$1,261
$720
1.41
CRWV
BALANCED
17
55.16
20 / 45 / 35
$51,289
$68,289
9% of book
$94k
$3,129
$1,392
3.67
DELL
DIAMOND
2
417.75
280 / 410 / 230
$27,500
$63,500
8% of book
$84k
$1,980
$1,224
2.51
MU
BALANCED
1
880.03
430 / 630 / 520
$44,560
$55,560
7% of book
$88k
$3,820
$1,461
3.40
MSTR
BALANCED
4
130.51
50 / 80 / 65
$32,528
$38,528
5% of book
$53k
$980
$514
3.19
SHOP
BALANCED
3
157.49
75 / 110 / 70
$24,585
$36,585
5% of book
$47k
$397
$213
1.86
HB750 proven
89
$437,247
$758,847
$1.06M
$25,467
$11,937
Stress battery (no liquidation)
module_risk_cockpit's TIMS-race crossover in 1% steps to −95%, the INCOME_BOOK's own battery: margin modelled as the scan loss x 1.22, excess liquidity must stay above zero. A fully hedged book with Max Loss under its cash cannot cross.
Each book rebuilt on the 28 Aug close at the same moneyness and tenor-to-go as its 30 Oct structure, same contracts, volatility held still at the pinned leg IVs, revalued on every close with the cockpit's option model. Two income lines: your rate = the 90-day income per month, accrued; model writing = one weekly call per contract every Friday at the pinned income call's distance out, never below that day's CC-SS, held to expiry. The model line is an upper bound: it never sits out a week, and October's fall let most calls expire.
Book
Account
Structure P&L
Calls, your rate
Return, your rate
Calls, model weekly
Return, model
Max drawdown (model)
Weeks a call hit CC-SS
Worst day
HBMax
$940,000
-$30,063
$23,280
-0.5%
$68,169
+4.3%
-4.1%
42 of 117
-4.1%
2026-10-30
HB750 proven
$820,000
-$186,665
$24,706
-19.5%
$96,123
-10.8%
-25.0%
24 of 99
-9.8%
2026-10-30
My feedback on the construction
Plan on your 90-day rate, aim for the book basis. At your hit rate HBMax earns about $11,248/mo, and the two-year replay shows $30,648, which holds to expiry and cannot see how you manage. The book basis ($22,943) is the model's premium at the trade-level keep of the INCOME_BOOK. Track it monthly; the quickest wins on your own record are writing every week and holding to expiry instead of buying back at a loss (holding did better 39 of 54 times).
The low-volatility names are back, as a buffer rather than an earner. XOM, PEP, LLY and the others earn much less on calls than the tech names: XOM's and PEP's calls roughly pay for their hedges. They earn their place by moving differently (average pair correlation +0.03) and by their structure gains when they rally.
Diversification is thin by nature here. The mega-caps pay about 22-36% a year on Max Loss against 73-90% for AMD, MU, PLTR and SPCX, so each slice moved into them costs income. HBMax: $866k Max Loss, $22,943/mo book basis, beta 0.84, 13 names. HB750 proven: $759k Max Loss, $25,467/mo book basis, beta 2.67, 11 names.
Earnings weeks inflate weekly premiums. Price a weekly call only after the name has reported: on Friday PLTR's 6 Nov call cost about seven normal weeks. The same check belongs on any fortress_v28 run done in an earnings week.
Every hedge is rented (77-112 day puts, rolled). The rent ($55,217/yr) is already out of the income. Never let one lapse: the no-liquidation result assumes they are renewed.
Earnings before you build: SPCX 3 Nov, MCD 4 Nov, CI 5 Nov, NVDA 17 Nov, DELL 24 Nov, MU 23 Dec, JPM 14 Jan, ABT 28 Jan, AXP 29 Jan, PEP 2 Feb, COIN 11 Feb. Build each leg after its print; never sell income into earnings.
Why it is less correlated
When one stock has a good or bad day, does the other tend to do the same? Daily returns over the last 180 trading days to 30 Oct, from −1 (always opposite) through 0 (unrelated) to +1 (always together). SPCX has only about 98 days. Share of daily risk is how much of the book's normal daily move comes from each position (dollar delta through the covariance), the INCOME_BOOK page's method.
DELL
COIN
MCD
XOM
SPCX
NVDA
CI
LLY
MU
ABT
JPM
AXP
COIN
+0.14
MCD
-0.20
-0.03
XOM
-0.16
-0.26
-0.00
SPCX
+0.27
+0.23
+0.03
-0.24
NVDA
+0.32
+0.32
-0.27
-0.31
+0.26
CI
-0.13
+0.06
+0.23
+0.07
-0.02
-0.12
LLY
-0.16
+0.05
+0.35
-0.04
-0.11
-0.19
+0.25
MU
+0.37
+0.27
-0.34
-0.25
+0.17
+0.46
-0.18
-0.12
ABT
-0.28
-0.03
+0.38
-0.00
-0.05
-0.11
+0.13
+0.29
-0.26
JPM
+0.19
+0.16
+0.09
-0.20
+0.01
+0.09
+0.21
+0.08
+0.03
+0.05
AXP
+0.02
+0.31
+0.14
-0.31
+0.10
+0.13
+0.14
+0.02
+0.03
+0.11
+0.55
PEP
-0.24
-0.15
+0.53
+0.09
-0.10
-0.30
+0.26
+0.26
-0.43
+0.34
-0.02
+0.06
Move together: AXP and JPM (+0.55), PEP and MCD (+0.53), MU and NVDA (+0.46), ABT and MCD (+0.38), MU and DELL (+0.37), LLY and MCD (+0.35).
Move opposite: PEP and MU (-0.43), MU and MCD (-0.34), AXP and XOM (-0.31), NVDA and XOM (-0.31), PEP and NVDA (-0.30), ABT and DELL (-0.28).
Average pair +0.04, against +0.27 for the previous book. Risk-weighted +0.04 against +0.24. Daily swing $18k against $30k.
The tech names still carry the daily risk: DELL, SPCX and MU are 51% of it, against 61% before. COIN and ASST were the tightest pair (+0.76), and ASST is out; JPM and AXP are the closest buffer pair, so together they are one financial-sector weight.
Sudden crashes are different. On the market's sharpest days most names fall together, so this helps most in slow or AI-led falls; the hedge puts and Max Loss protect you in a fast crash.
Name
Contracts
Max Loss
Share of Max Loss
Share of daily risk
Avg correlation to the others
Reading
DELL
3
$95,250
11.0%
21%
+0.01
carries more risk than its size
COIN
8
$95,192
11.0%
20%
+0.09
carries more risk than its size
MCD
10
$89,640
10.3%
5%
+0.08
diversifier: far less risk than its size
XOM
13
$77,012
8.9%
-4%
-0.14
offsets the rest of the book
SPCX
8
$72,016
8.3%
19%
+0.05
carries more risk than its size
NVDA
7
$69,090
8.0%
9%
+0.02
roughly in proportion
CI
5
$66,825
7.7%
4%
+0.08
diversifier: far less risk than its size
LLY
1
$65,600
7.6%
4%
+0.06
diversifier: far less risk than its size
MU
1
$55,560
6.4%
11%
-0.02
carries more risk than its size
ABT
12
$50,580
5.8%
2%
+0.05
diversifier: far less risk than its size
JPM
4
$48,076
5.6%
4%
+0.10
roughly in proportion
AXP
3
$45,066
5.2%
4%
+0.11
roughly in proportion
PEP
12
$36,264
4.2%
1%
+0.02
diversifier: far less risk than its size
Your convexity: $646k in long calls, and that is the whole Max Loss. Nothing in it can force a sale. The crypto sleeve (500 MARA + 250 BMNR, $203k, with 250 more BMNR only if it reaches $30), the IBIT calls that replace your 10 BTC ($125k, freeing $643k) and three bets (MU $110k, SPCX $101k, LLY $107k). IREN ($43k) comes out of Neville's own cash on top of this, so it is in no total. On top, a 25% weekly buffer: each Friday the weekly call nearest 0.20 delta on a quarter of each long call (MU skipped, LLY on 3 of 5), about $8,283/mo on the book basis, $3,017 on last year's real path.
Hold each weekly to expiry, even when challenged; never buy back mid-week at 3-4x. Out of the money at expiry: penny buyback, sell the next one.
In the money at expiry, in a rally: buy it back at intrinsic and pause that name's calls (or drop to about 10%) while the rally runs. Do not roll out: the rolled call stays short below your LEAPS strike, into the move you own them for. One +25% MARA week costs about $4.8k to buy back against about $34k gained on the LEAPS.
25% is the default, 50% the ceiling, as a temporary income lever when crypto is flat or falling. Past 50% a doubling week leaves the sleeve close to flat.
Keep the weeklies in a portfolio-margin account: they sit below the LEAPS strikes, so Reg-T treats them as call spreads.
More convexity from the freed cash
Your ask (2026-11-01): $300-400k of the freed cash into small bets on IREN, one each on MU and SPCX, and an AI-biotech bet; ASST and TEM dropped. Your changes (2026-11-02): XBI replaced by 5 LLY; SPCX to the 175C; IREN to 45x the 70C, and only if you secure additional funds. Your change (3 Nov): IREN to 50x the Jan '29 50C, from Neville's own cash, outside the plan's budget. Calls only, so cost = Max Loss. MU, SPCX and LLY break even at about 1.5-1.6x the stock (your 2-3x rule); IREN is further out of the money for convexity. Odds = the market's chance (implied vol) of finishing above break-even. P&L at expiry with the stock at each multiple of its 30 Oct close. It comes out of the BTC swap's freed cash (Overview).
Bet
Stock (option bid / ask)
Cost = Max Loss
Break-even
Odds
Open interest
Stock 1.5x
Stock 2x
Stock 3x
Stock 5x
Why
5x MU Dec '28 1200C
$884.74
bid 204.00 / ask 220.50
$110,250
mid $106,125
1.61x
19%
1,666
-$47k
$174k
$617k
$1.50M
memory for AI; beats the 1400C at 1.5-2x and matches it at 3x; you kept it over the 1000C at your targets
31x SPCX Dec '28 175C
$134.72
bid 31.65 / ask 32.50
$100,750
mid $99,433
1.54x
20%
1,877
-$17k
$192k
$610k
$1.44M
your strike (2 Nov), the plan's budget: pays at your $200 target, where the old 200C pays nothing
5x LLY Dec '28 1900C
$1417.54
bid 200.00 / ask 213.60
$106,800
mid $103,400
1.49x
20%
52
$6k
$361k
$1.07M
$2.49M
your swap for XBI (2 Nov), 5 contracts, weekly calls on 3; thin, 52 open interest
Total
$317,800
mid $308,958
-$57k
$727k
$2.30M
$5.43M
Outside the budget: 50x IREN Jan '29 50C
$23.95
bid 8.35 / ask 8.55
$42,750
not in the total
2.44x
10%
3,355
-$43k
-$43k
$66k
$306k
your strike and size (3 Nov); outside the plan's budget
The bets at your targets
Value at expiry if the stock is at your price then: low, high and moonshot. LLY: $2,620 is its ten-year 33.6% a year carried to Dec '28, $4,200 is that rate doubled; it has no moonshot target, so it stays at $4,200 in that column's total.
Bet
Cost = Max Loss
Break-even
Low target
High target
Moonshot
5x MU Dec '28 1200C
$110,250
$1,420.50
$200k
at $1,600, profit $90k
$400k
at $2,000, profit $290k
$650k
at $2,500, profit $540k
31x SPCX Dec '28 175C
$100,750
$207.50
$78k
at $200, profit -$23k
$388k
at $300, profit $287k
$698k
at $400, profit $597k
5x LLY Dec '28 1900C
$106,800
$2,113.60
$360k
at $2,620, profit $253k
$1.15M
at $4,200, profit $1.04M
–
Total
$317,800
$638k
$1.94M
$2.50M
Outside the budget: 50x IREN Jan '29 50C
not in the total
$42,750
$58.55
$250k
at $100, profit $207k
$500k
at $150, profit $457k
$750k
at $200, profit $707k
This is one bet, not three. AI and growth rally together and fall together, LLY less so; in a 2022-style bust they can all go to near zero at once. With the crypto sleeve ($203k) and the IBIT calls ($125k), calls that can expire worthless total $646k, before IREN.
Double-ups: HBMax already writes calls on MU and SPCX fortresses. Dropped: ASST (a fourth bitcoin bet); TEM (owner 2026-11-01).
Realistic good cases: MU at 3x is about $2654, a ~$3T company after a huge run, so 2x is the one to plan on. LLY at $2,620 is its own ten-year rate; $4,200 is that rate doubled, about 3x the stock.
Inside your $300-400k range: $318k. Room left in it: up to $82k.
Execution: limit orders near the mid. LLY is the thin one: 5 contracts is 10% of its open interest and the spread is about 7%, so work it patiently. Quotes are the 30 Oct close: recheck on the day.
The 25% weekly buffer on IBIT and the bets
Your ask (2026-11-01): a 25% income sleeve as a buffer for the IBIT calls and the bets. The crypto sleeve's rule: each Friday sell the weekly call nearest 0.20 delta on a quarter of each long call, close it the next Friday with a penny buyback if out of the money, skip earnings weeks. Premium at the 30 Oct bid; 61 weeks to end-2027 on 4,000 paths at the weekly's volatility. Paths: no view (the average price stays flat), last year's realized volatility, down (each name halves), rally (each name doubles). One rally week = the short weeklies' loss net of the LEAPS' gain that week. Last 12 months = the same rule on the real Friday closes.
Long calls
Covered
Weekly call sold
Premium / week
Per month gross
No view to end-2027
At last year's volatility
Down path (halves)
Rally path (doubles), typical
Last 12 months, real path
One +50% week
One +100% week
300x IBIT Dec '28 90C
75 of 300
46 C, 6 Nov
0.18 delta · 6% above · $0.23
$1,725
$7,501
$3k
-$31k
$21k
-$52k
$17k
stock 0.70x
-$78k
-$146k
5x MU Dec '28 1200C skip
1 of 5
952.5 C, 6 Nov
0.20 delta · 8% above · $7.65
$765
$3,327
$1k
-$32k
$6k
-$13k
-$36k
stock 3.96x
-$8k
-$17k
31x SPCX Dec '28 175C
8 of 31
144 C, from 13 Nov
0.21 delta · 7% above · $1.25
$1,000
$4,349
$294
-$3k
$8k
-$19k
$5k
stock 0.84x
-$10k
-$20k
5x LLY Dec '28 1900C
3 of 5
1475 C, 6 Nov
0.21 delta · 4% above · $4.25
$1,275
$5,544
-$47k
-$39k
-$6k
-$158k
-$67k
stock 1.65x
-$65k
-$104k
Total
$4,765
$20,721
-$43k
-$106k
$30k
-$241k
-$81k
-$160k
-$287k
What it earns: $20,721 a month gross, $277k to end-2027 if every call expires worthless. That is the ceiling: the rallies take a lot of it back.
With no view it loses about $43k at these bids, nearly all of it on LLY, whose weekly bid is 60% of its mid, and there is no volatility premium to harvest here. On MARA the options price bigger moves than the stock makes; on these names last year it was the other way round: MU moved 75% a year against 59% implied and IBIT 47% against 43%. At last year's volatility the buffer loses $106k.
What it buys: if the bets fail and every name halves by end-2027, about $30k, roughly 7% of the $442k these calls can lose. What it costs: on the path you own them for (each name doubles by end-2027) it loses about $241k typical, and one +50% week across all of them at once costs $160k.
Run it on IBIT and SPCX at 25%, LLY on 3 of 5 (your choice, 2 Nov); skip MU; none on IREN. MU is one contract for $3,327 a month, it went 4.0x last year (its weeklies lost $36k on the real path) and its realized volatility runs well above implied. Without MU: $17,394 a month gross, -$44k expected with no view, -$73k at last year's volatility, -$45k on the last 12 months. As on the crypto sleeve, 25% is the ceiling, and pause a name's weeklies once it is up about 50% from here: from there you are on the rally path, where they cost the most.
LLY on 3 of 5 is above the ceiling, on purpose. Its long call is far above the weekly, so past the weekly's strike each short loses $1 per $1 while the 1900C gains about 50 to 70 cents: at 3 of 5 a big rally leaves the position about flat, and at 5 of 5 it loses. The rule above lost on LLY this year; your own handling (pausing, closing early) is what this leans on, and there is no LLY record of yours yet. Close a challenged call rather than roll it, and work it near the mid: 4.25 / 9.95 on the 1475 C, modelled at the bid.
Earnings: SPCX reports 3 Nov, so its first weekly is 13 Nov (priced here at the ex-earnings volatility: its 6 Nov quote is 58% with the print in it, 57% without). MU skips the week of 23 Dec. IBIT has none. LLY reports around 4 Feb (an estimate from its past Februaries: confirm the date).
Margin: under portfolio margin (Main's calibration) the weeklies add nothing (model $248k with them, $255k without), because each is covered by its LEAPS. In a Reg-T account they would lock up the gap between the two strikes, $507k, which undoes most of what the BTC swap freed: keep these in a portfolio-margin account.
Appendixthe BTC swap strikes, the MARA / BMNR buffer model, every sleeve size, blends, the Sep-Oct replay and feedback
Your sleeve: 500 MARA + 250 BMNR, 250 more at BMNR $30 with weekly calls on 25%. 500x MARA Dec '28 25C @ $1.34 + 250x BMNR Jan '29 40C @ $5.45. Cost $203,250 at the ask ($195,375 at the mid) on the 30 Oct close, and that cost is the whole Max Loss until the add: nothing can force a sale. Runway to Dec '28 / Jan '29, about 25 to 27 months (no Jan '28, as you asked). At the CRYPTO page's targets, with the add bought on the way up, it is worth $2.25M at MARA $40 / BMNR $70, $5.25M at $70 / $100 and $9.25M at $100 / $150, 56% of IDEAL's, against the current book's $3.99M, $8.38M and $13.70M. Break-even at expiry MARA $30 / BMNR $55. On top, each Friday sell the weekly call nearest 0.20 delta on 125 MARA and 62 BMNR (a quarter of each leg), the CRYPTO_SLEEVE page's rule: earnings weeks skipped, closed the next Friday with a penny buyback if out of the money.
The BMNR rule: 250 now, 250 more at $30
The price of the other 250 BMNR 40C Jan '29 if BMNR first reaches $30 by each date: the leg's 30 Oct volatility (74%) held, today's ask over mid. Against buying all 500 now for $340k.
BMNR reaches $30 by
Price per call
Cost of 250
Total sleeve paid
Against 500 now
Never
–
$0
$203k
saves $136k
Dec '26
$10.87
$272k
$475k
costs $136k more
Jun '27
$8.80
$220k
$423k
costs $84k more
Dec '27
$6.61
$165k
$368k
costs $29k more
Jun '28
$3.87
$97k
$300k
saves $39k
Above about BMNR $40 the payoff is the same as 500 now, because the add is already bought. What differs is what you paid, and that gap does not grow however high BMNR goes.
The bad case is a whipsaw: BMNR touches $30, you add, and it fades below $40. The sleeve can then lose its full paid cost, up to about $475k rather than $340k.
The money: $136k stays in Neville as cash. If the add costs more than that, the rest comes from Joint's reserve, so budget for it before you buy.
It is your decision on the day: a new ticket, limit near the mid, into a stock that has just jumped about 50%. Volatility may be higher then (at 85% the Jun '27 add costs about $34k more). Weekly calls stay on 62 BMNR until the add, then 125.
Which strike on the day. Today Jan '29 stops at 40, but higher strikes should be listed by then, and the strike is your call. Every number on this page uses the 40C as the cost. These are the Jan '29 calls with BMNR at $30 on 30 Jun 2027, volatility held at 74% across strikes (higher strikes usually cost a little more than this), value at expiry.
Strike
Price
250 cost
Against the 40C
250 at BMNR $70
250 at $100
250 at $150
Same money as 250x 40C buys
At $70
At $100
At $150
Passes the 40C above
40C
$8.80
$220k
–
$750k
$1.50M
$2.75M
250
$750k
$1.50M
$2.75M
–
50C
$6.82
$171k
saves $49k
$500k
$1.25M
$2.50M
322
$644k
$1.61M
$3.22M
BMNR about $85
60C
$5.39
$135k
saves $85k
$250k
$1.00M
$2.25M
408
$408k
$1.63M
$3.67M
BMNR about $92
Your 10 BTC: swap for IBIT calls
Sell the 10 BTC you hold outside IBKR ($767k at about $76.7k) and buy IBIT Dec '28 calls that pay only if BTC does really well by 15 Dec 2028. IBIT $43.36 on the 30 Oct close (CBOE), about 0.000565 BTC per share (the 0.25%/yr fee shrinks it to expiry). Every row spends about $125k; P&L at expiry includes interest on the cash left over. Your pick (2026-11-02, after the strike remodel below): 300x Dec '28 90C, $124,500, frees $643k; your first pick, 240x 80C, is shown beside it.
BTC equiv.
Cost = Max Loss
Cash freed
Break-even BTC
1.5x ($115k)
2x ($153k)
2.5x ($192k)
3x ($230k)
4x ($307k)
5x ($384k)
Hold 10 BTC
–
$767,434
–
–
$384k
$767k
$1.15M
$1.53M
$2.30M
$3.07M
144x IBIT Dec '28 60C
8
$125,280
$642k
$117k (1.5x)
-$12k
$299k
$609k
$920k
$1.54M
$2.16M
187x IBIT Dec '28 70C
11
$125,290
$642k
$132k (1.7x)
-$79k
$225k
$628k
$1.03M
$1.84M
$2.64M
240x IBIT Dec '28 80C first pick
14
$124,800
$643k
$148k (1.9x)
-$79k
$72k
$589k
$1.11M
$2.14M
$3.18M
300x IBIT Dec '28 90C your pick
17
$124,500
$643k
$165k (2.1x)
-$78k
-$78k
$456k
$1.10M
$2.40M
$3.69M
225x IBIT Jan '29 80C
13
$124,875
$643k
$148k (1.9x)
-$77k
$64k
$549k
$1.03M
$2.00M
$2.97M
The least cash that matches holding 10 BTC
BTC at expiry
10 BTC gains
Least calls that match it
Cash
$200k
$1.23M
379 x 75C
$224k
$300k
$2.23M
294 x 90C
$122k
$400k
$3.23M
244 x 90C
$101k
$600k
$5.23M
213 x 90C
$89k
Every strike from 65C to 90C pays about the same at 3x for the same cash. The strike only moves the payoff between 2x and 5x+: the 80C breaks even at about 2x (your first rule, "2-3x or give up"), the 90C (your pick since 2 Nov, for BTC $200-250k by mid-2028 and more in an impulsive leg) pays nothing at 2x and the most above 4x, the 70C pays at 2x and gives up the top.
Below about $160k BTC on 15 Dec 2028 (the strike) the calls are worth nothing, where holding would still be up $834k; with the interest on the freed cash you break even at about $165k. Between there and about $300k they trail holding; above that they lead.
Liquidity: 300 of the 90C is 0.7% of its 40,072 open interest (spread 3.80 / 4.15); 240 of the 80C 0.7% of 33,814. Jan '29 pays about 5% less for the same cash and is thinner. Sell the BTC and buy the calls in the same session, in clips near the mid.
The freed $643k, less the bets below and plus your spare cash, goes into the index and the reserve (Brought in, Overview); it earns about $1,811/mo at 3.38% until used. It is not in the reserve the crash grids and replays use, because today's book does not hold the BTC either.
Strike choice if BTC reaches $200-250k between Dec '27 and Jun '28
Your premise (2026-11-02): BTC reaches $200k-250k sometime from late 2027 to mid 2028, and could go much further in an impulsive leg. Every listed IBIT call from Dec '27 to Jan '29 with at least 500 open interest, bought with the same $125k at the ask (30 Oct close, CBOE). Your window = BTC $200k, $225k or $250k reached on 1 Dec '27, 15 Mar '28 or 30 Jun '28, along two paths: steady (log-linear from today) or flat then one leg up over the last four months. The calls are sold that day at Black-Scholes, 45% IV, less 3% for the spread; a call that expired on the way up pays its intrinsic value then (nothing on the flat-then-up path). A quarter late = the same levels on 30 Sep '28. The impulsive columns use the flat-then-up path. Holding excludes interest; the calls exclude the interest on the freed cash (about $23-36k).
Strike in BTC
Open interest
Worst in your window
Average in your window
Best in your window
A quarter late (worst)
BTC $300k by 30 Jun '28
BTC $350k by 30 Jun '28
BTC $400k by 30 Jun '28
BTC $450k by 30 Jun '28
BTC $500k by 30 Jun '28
BTC only $150k by Dec '28
Hold 10 BTC
–
–
$1.23M
$1.48M
$1.73M
$1.23M
$2.23M
$2.73M
$3.23M
$3.73M
$4.23M
$733k
186x Dec '28 70C
$125k
7,200
$677k
$953k
$1.22M
$653k
$1.68M
$2.19M
$2.70M
$3.20M
$3.71M
$141k
211x Dec '28 75C
$133k
5,322
$693k
$1.01M
$1.32M
$658k
$1.82M
$2.40M
$2.97M
$3.55M
$4.13M
$72k
240x Dec '28 80C first pick
$142k
33,814
$705k
$1.07M
$1.42M
$655k
$1.98M
$2.63M
$3.29M
$3.94M
$4.60M
-$22k
300x Dec '28 90C your pick
$160k
40,072
$686k
$1.13M
$1.57M
$590k
$2.22M
$3.04M
$3.85M
$4.67M
$5.49M
-$125k
225x Jan '29 80C
$142k
1,941
$672k
$1.01M
$1.33M
$622k
$1.85M
$2.47M
$3.08M
$3.69M
$4.31M
-$29k
271x Jun '28 70C expires 16 Jun '28
$124k
4,726
$706k
$1.32M
$1.78M
-$125k
$1.88M
$2.45M
$3.01M
$3.56M
$4.11M
-$68k
362x Jun '28 80C expires 16 Jun '28
$142k
2,961
$622k
$1.47M
$2.09M
-$125k
$2.19M
$2.95M
$3.70M
$4.44M
$5.16M
-$125k
Dec '28 80C and 90C are tied at the bottom of your window (BTC $200k on 30 Jun '28: $705k vs $686k). Everywhere above it the 90C pays more: $95k more at $250k, $245k more at $300k and $895k more at $500k. It loses only where BTC stalls near $150k: there the 80C is about flat (-$22k) and the 90C loses everything (-$125k).
Jun '28 pays the most when the timing is right (the 80C makes up to $2.09M), but it expires on 16 Jun '28, right at the end of your window. If the run comes a quarter late, or the impulsive leg only starts in June, it is worth nothing. Dec '28 still pays $590k or more a quarter late.
Dec '27 and Jan '28 expire in the middle of your window. The best of them still loses $88k if the run comes in 2028.
Against holding the BTC: in your window holding makes $1.23M to $1.73M, the calls $686k to $1.57M plus about $30k of interest on the freed cash. The calls pull ahead only in the impulsive leg (about $300k and up). That gap is the price of capping the downside at the premium.
IV on the hit day moves the Dec '28 numbers by about $0.1M: at 35% the 80C leads, at 60% the 90C does. The ranking inside Dec '28 barely changes.
The table sells on the day BTC first reaches the level. Keeping a Dec '28 call after that keeps the impulsive-leg upside, with less time value at risk.
The 25% weekly buffer
The CRYPTO_SLEEVE page's weekly-call model (weekly_income.py: Friday 0.2-delta weeklies at the 30 Oct bids, 4,000 paths to end-2027), for 125 MARA + 62 BMNR covered. Premium and the last 12 months are summed per contract; the path rows scale the page's 125 / 105 row by the premium (BMNR's share is larger here, so they are close, not exact).
Result
What it means
Premium collected
$2,178 a week, $9,471 a month gross
every call expires worthless: $122k to end-2027
Last 12 months, real path
+$84k
a falling year (MARA $18 to $7): the best case for selling calls
No view, to end-2027
$6k typical / -$12k expected
roughly break-even: the premium pays for the rallies
No view at last year's calmer vol
$24k expected
the volatility-premium case (MARA realized 85% vs 115% implied)
Down path (prices halve)
$8k typical / -$7k expected
a small cushion, not a hedge
IDEAL path (MARA $40 / BMNR $70 by end-2027)
-$114k typical / -$278k expected
1-in-10 bad -$737k, against the sleeve's own gain there
One rally week
+25%: $6k / +50%: -$16k / +100%: -$51k
the short calls' loss net of the long calls' gain that week
The long-dated sizes
MARA 25C Dec '28 + BMNR 40C Jan '29 only, the Crypto Options Comparison page's sizes, re-priced on the CBOE close of 30 Oct (MARA $6.77, BMNR $20.18), which is cheaper than that page's 29 Oct quotes. Values at expiry with the pair reached. Weekly columns at 25% cover of each size. Reserve = the reset's cash reserve with that sleeve. The last row, for comparison only, is the same strikes on Jan '28 at half of IDEAL 50%'s budget.
Sleeve
Legs (bid / ask)
Cost = Max Loss
MARA 40 / BMNR 70 book $3.99M
MARA 70 / BMNR 100 book $8.38M
MARA 100 / BMNR 150 book $13.70M
Weekly calls / mo gross
Weekly calls IDEAL path, typical
S&P −10%
S&P −20%
Sep-Oct replay
Reset reserve
IDEAL over $0.5M
1000x MARA Dec '28 25C
1.20 / 1.34 · 11% of OI
840x BMNR Jan '29 40C
5.10 / 5.45 · 12% of OI
$591,800
mid $570,100
$4.02M
+1%
$9.54M
+14%
$16.74M
+22%
$26k
-$315k
-$297k
-$490k
-43%
-$325k
IDEAL 75%
750x MARA Dec '28 25C
1.20 / 1.34 · 9% of OI
630x BMNR Jan '29 40C
5.10 / 5.45 · 9% of OI
$443,850
mid $427,575
$3.02M
-24%
$7.16M
-15%
$12.55M
-8%
$19k
-$236k
-$223k
-$368k
-43%
-$175k
IDEAL 60%
600x MARA Dec '28 25C
1.20 / 1.34 · 7% of OI
500x BMNR Jan '29 40C
5.10 / 5.45 · 7% of OI
$352,900
mid $339,950
$2.40M
-40%
$5.70M
-32%
$10.00M
-27%
$15k
-$189k
-$177k
-$292k
-43%
-$82k
500 MARA + 500 BMNR your size once the add is bought
500x MARA Dec '28 25C
1.20 / 1.34 · 6% of OI
500x BMNR Jan '29 40C
5.10 / 5.45 · 7% of OI
$339,500
mid $327,250
$2.25M
-44%
$5.25M
-37%
$9.25M
-32%
$15k
-$177k
-$171k
-$281k
-41%
-$68k
IDEAL 50%
500x MARA Dec '28 25C
1.20 / 1.34 · 6% of OI
420x BMNR Jan '29 40C
5.10 / 5.45 · 6% of OI
$295,900
mid $285,050
$2.01M
-50%
$4.77M
-43%
$8.37M
-39%
$13k
-$157k
-$149k
-$245k
-43%
-$24k
Jan '28, half budget expires Jan '28
480x MARA Jan '28 25C
0.55 / 0.67 · 6% of OI
400x BMNR Jan '28 40C
2.70 / 2.86 · 2% of OI
$146,560
mid $140,480
$1.92M
-52%
$4.56M
-46%
$8.00M
-42%
$12k
-$151k
-$89k
-$131k
-54%
$128k
Value at expiry along the tandem ladder
MARA / BMNR
$11 / $22
$20 / $35
$25 / $45
$30 / $55
$40 / $70
$50 / $80
$70 / $100
$100 / $150
$150 / $225
$200 / $300
IDEAL
$0
$0
$420k
$1.76M
$4.02M
$5.86M
$9.54M
$16.74M
$28.04M
$39.34M
IDEAL 75%
$0
$0
$315k
$1.32M
$3.02M
$4.39M
$7.16M
$12.55M
$21.03M
$29.50M
IDEAL 60%
$0
$0
$250k
$1.05M
$2.40M
$3.50M
$5.70M
$10.00M
$16.75M
$23.50M
500 MARA + 500 BMNR
$0
$0
$250k
$1.00M
$2.25M
$3.25M
$5.25M
$9.25M
$15.50M
$21.75M
IDEAL 50%
$0
$0
$210k
$880k
$2.01M
$2.93M
$4.77M
$8.37M
$14.02M
$19.67M
Jan '28, half budget
$0
$0
$200k
$840k
$1.92M
$2.80M
$4.56M
$8.00M
$13.40M
$18.80M
Jan '28 at half the budget: same payoff, a shorter clock
500 MARA + 500 BMNR vs Jan '28, half budget (480x MARA Jan '28 25C @ $0.67 + 400x BMNR Jan '28 40C @ $2.86). Model values, each leg at its 30 Oct implied vol held still; a Jan '28 leg is worth nothing after 21 Jan 2028.
500 MARA + 500 BMNR
Jan '28, half budget
Cost = Max Loss
$339,500
$146,560
MARA $40 / BMNR $70 at expiry
$2.25M
$1.92M
Prices flat, end-2027
$126k
$0
Prices double, end-2027
$739k
$121k
Prices triple, end-2027
$1.68M
$845k
MARA $40 / BMNR $70 by end-2027
$2.82M
$1.93M
MARA $40 / BMNR $70 only in June 2028
$2.52M
$0
Crypto falls 50% now
$71k
$17k
Jan '28 costs half and pays almost the same if the full move lands by 21 Jan 2028. A partial rally pays a fraction of the long-dated sleeve, and a move that arrives in 2028 pays nothing. Pick it only if you are confident on timing as well as direction.
Blends: part long-dated, part Jan '28
IDEAL 50%'s 500 MARA / 420 BMNR split between the long-dated legs and Jan '28 (same strikes, 30 Oct asks). Model values, each leg at its 30 Oct implied vol held still. "Re-buy" = the cost on 21 Jan 2028 to replace the expired Jan '28 contracts with Jan '29 calls, back to full size.
Sleeve
Legs
Cost = Max Loss
MARA 40 / BMNR 70 at expiry
40 / 70 by end-2027
40 / 70 only in June 2028
21 Jan 2028 down 50%: left
21 Jan 2028 flat: left
21 Jan 2028 doubled
S&P −20%
Sep-Oct replay
500 MARA + 500 BMNR your size once the add is bought
500x MARA Dec '28 25C 500x BMNR Jan '29 40C
$339,500
$2.25M
$2.82M
$2.52M
$10k
$116k
$715k held
-$281k
-41%
IDEAL 50%
500x MARA Dec '28 25C 420x BMNR Jan '29 40C
$295,900
-$44k vs that
$2.01M
$2.53M
$2.26M
$9k
$100k
$617k held
-$245k
-43%
Blend 75 / 25
380x MARA Dec '28 25C 320x BMNR Jan '29 40C 120x MARA Jan '28 25C 100x BMNR Jan '28 40C
$261,960
-$78k vs that
$2.01M
$2.41M
$1.72M
$7k
$76k
$473k held
+ $153k to re-buy
-$219k
-44%
Blend 50 / 50
250x MARA Dec '28 25C 210x BMNR Jan '29 40C 250x MARA Jan '28 25C 210x BMNR Jan '28 40C
$224,760
-$115k vs that
$2.01M
$2.28M
$1.13M
$4k
$50k
$315k held
+ $321k to re-buy
-$191k
-47%
Blend 25 / 75
120x MARA Dec '28 25C 100x BMNR Jan '29 40C 380x MARA Jan '28 25C 320x BMNR Jan '28 40C
$187,560
-$152k vs that
$2.01M
$2.14M
$540k
$2k
$24k
$157k held
+ $489k to re-buy
-$163k
-50%
Jan '28, half budget
480x MARA Jan '28 25C 400x BMNR Jan '28 40C
$146,560
-$193k vs that
$1.92M
$1.93M
$0
$0
$0
$12k held
+ $613k to re-buy
-$131k
-54%
IDEAL 38%, long-dated
380x MARA Dec '28 25C 320x BMNR Jan '29 40C
$225,320
-$114k vs that
$1.53M
$1.93M
$1.72M
$7k
$76k
$470k held
-$187k
-43%
The blend is a dial, not a sweet spot. Every Jan '28 contract swapped in saves about half its long-dated price now and costs about four times that saving if, on 21 Jan 2028, crypto has doubled but not reached the targets: Blend 50 / 50 saves $115k today and needs $321k to get back to full size then.
When you are wrong, the saving is smaller than it looks. Your pick keeps about a year of time value: on 21 Jan 2028 with prices flat it is still worth $116k, so walking away loses $223k against Blend 50 / 50's $175k.
The same money as Blend 50 / 50 on long-dated calls only (IDEAL 38%, long-dated) pays less if you are right on time ($1.93M vs $2.28M at MARA $40 / BMNR $70 by end-2027) and more if you are late ($1.72M vs $1.13M in June 2028). The blend is a bet on timing; the smaller long-dated sleeve is a bet on direction.
Sep and Oct replayed
The sleeve rebuilt on 28 Aug at the same moneyness and dollars, volatility held still (every size moves the same in percent). MARA fell 37% and BMNR 15% over the window. The weekly calls are not in this line.
Feedback on the sleeve
The long-dated choice is right. Jan '28 lands three weeks after the CRYPTO page's end-2027 horizon; a late cycle would expire it. Dec '28 / Jan '29 buys about a year of slack for about twice the premium per contract.
A smaller sleeve means a smaller payoff, at every price. $2.25M at MARA $40 / BMNR $70 against the book's $3.99M. That is the trade you are choosing: $203k now (up to about $475k after the BMNR add) at risk instead of today's $526k if crypto goes to zero.
The weekly calls are a buffer for a flat or falling market, and a tax on the rally you own the sleeve for. With no view they are about break-even (-$12k expected); at last year's calmer volatility they earn about $24k; on the IDEAL path they cost about -$114k typical. At 25% cover that tax is about a tenth of the sleeve's own gain there, which is why 25% is the right ceiling. I would not go above it.
My suggestion: pause them once the rally is under way, say MARA above about $15 or BMNR above about $40. From there you are on the IDEAL path, where they cost the most: one +50% week costs $16k and one +100% week $51k at this cover.
Earnings: MARA reports 4 Nov and BMNR 20 Nov. No MARA weekly next week; the first BMNR weekly can go on 6 Nov.
Build in clips, and swap, do not stack. 500 MARA 25C is 6% of open interest and 250 BMNR 40C 4%. Buy the sleeve in the same sessions you close today's MARA / BMNR legs.
Hold vs reset, side by side
Hold today's book
Reset now
Account value now
$2,555,181
$2,975,154
after about $125k of friction
Speed: loss at S&P −10%
-35%
$9.84M beta-weighted, 3.7x NLV (cockpit)
-21%
Max Loss vs NLV
Main 1.08x, Neville 0.95x, Joint 0.92x, RetireInc 0.76x
$866k fortress Max Loss = 0.93x its own account; crypto Max Loss = its cost
Cash / margin loan
Joint −$382k loan
no loan; $500k reserve
S&P -10%
$1.65M (-35%)
$2.34M (-21%)
S&P -20%
$1.03M (-60%)
$1.84M (-38%)
S&P -30%
$811k (-68%)
$1.52M (-49%)
S&P -40%
$851k (-67%)
$1.32M (-55%)
S&P +10% (before calls cap)
+$818k
+$723k
Forced selling
No on all four today (Repair, 30 Oct); two accounts were At risk on 21 Oct
$22,943 book basis / $11,248 at your rate (calls after rent)
Sep-Oct 2026
-27.7% (actual)
-5.1% (replay)
Crypto if MARA $40 / BMNR $70
$3.99M (CRYPTO page)
$2.25M
with the 250 BMNR added at $30
Crypto if both go to zero
about −$526k (30 Oct partial-reset pricing)
−$203k
about −$368k to −$475k if the add is bought first
Leverage
Frozen on 3 Nov: built on the 2 Nov HBMax (nine names) and the old book's legs, which are now closed. New HBMax beta: about 1.06.
Exposure divided by all your money on each side: the IBKR accounts, spare cash and the 10 BTC for hold, the full plan for the reset. The reset is shown before the BMNR add; buying it lifts Neville and the total. S&P-weighted is the Risk Cockpit's measure, delta times spot times each name's beta: 1.0x is like holding all your money in the S&P. Volatility held still in the felt rows.
Hold today's book
Reset
All your money
$3.39M
$3.27M
once done, after friction
S&P-weighted, with weekly calls
3.11x
$10.54M
the calls on today
about 2.19x
$7.16M
estimate, calls not on yet
S&P-weighted, before weekly calls
3.50x
$11.86M
2.49x
$8.12M
Dollar delta, no beta, with weekly calls
1.38x
$4.68M
about 1.24x
$4.07M
Notional: shares and long calls at full value
2.36x
$7.99M
2.25x
$7.35M
Felt at S&P -10%: share lost per 1% of S&P
3.0x
2.4x
no calls on
Felt at S&P -20%: share lost per 1% of S&P
2.6x
2.1x
no calls on
Felt at S&P -30%: share lost per 1% of S&P
2.0x
1.7x
no calls on
Felt at S&P +10%: share gained per 1% of S&P
2.7x
2.8x
before calls cap it
Short puts in the S&P-weighted number
$5.14M
$565k
The reset cuts S&P-weighted leverage by about 30%, 3.11x to about 2.19x, but it barely de-gears the dollars: delta goes from 1.38x to about 1.24x. Most of the cut comes from swapping high-beta names for the index.
The kind of leverage changes more than the amount. Short puts fall from $5.14M to $565k of the S&P-weighted number, and the margin loan goes. About 44% of the reset's exposure ($3.56M of $8.12M) is long calls in Neville, where the most you can lose is the $646k paid, with no margin call.
In a real fall the gap is smaller: 2.4x against 3.0x at S&P −10%, with no reset calls on yet; weekly calls would help the reset a little. On the way up the reset gains more until calls cap it.
Why the table above looks wider: its 3.6x against 2.1x at S&P −10% counts the IBKR money only, before the BTC swap. Selling the BTC for IBIT calls and the bets adds leverage to the reset, while the BTC and spare cash dilute hold's.
Hold's 3.7x: that is the cockpit on the IBKR accounts alone, from its 30 Oct 16:30 run. These legs (Repair, 2026-11-02 22:34) give 3.70x on that basis, and 3.11x with the BTC and spare cash.
By account: each against its own money. Neville at 4.55x is high because it holds only long calls; Joint at 1.22x is right at the 1.2x ceiling suggested for it.
Money
S&P-weighted
Dollar delta
Notional
Hold: Main
$1.20M
3.77x
1.47x
2.70x
Hold: Joint
$590k
3.98x
1.76x
3.36x
Hold: Neville
$460k
3.77x
1.50x
2.65x
Hold: RetireInc
$308k
2.77x
1.42x
2.57x
Hold: Outside IBKR: 10 BTC + spare cash
$833k
1.30x
0.92x
0.92x
Reset: Main: income book
$790k
3.16x
1.30x
1.58x
Reset: Joint: index + cash
$1.70M
1.22x
1.15x
1.41x
Reset: Neville: crypto, IBIT, bets
$782k
4.55x
1.97x
4.75x
If it bounces
Frozen on 3 Nov: this tab was built on the 2 Nov HBMax (nine names) and the old book's legs, which are now closed, so it was not re-run for the new HBMax.
The reset done by next week, then the market moves. Every name by its Risk Cockpit beta (MARA and BMNR about 3x, IBIT 1.4x), volatility held still; gains from each side's own start (the reset starts about $125k lower, its friction). The reset is the full plan with the 10 BTC sold for the IBIT calls, the bets and the rest of the index; hold keeps the 10 BTC.
Hold
Reset
Reset vs hold
Now (Reset: once done)
$3.39M
$3.27M
-$121k
S&P +5%
+$500k
NLV $3.89M
+$445k
NLV $3.71M
with calls +$399k, NLV $3.67M
-$55k
NLV -$176k
with calls -$101k, NLV -$222k
S&P +10%
+$927k
NLV $4.31M
+$914k
NLV $4.18M
with calls +$717k, NLV $3.98M
-$12k
NLV -$133k
with calls -$209k, NLV -$330k
S&P +15%
+$1.33M
NLV $4.71M
+$1.40M
NLV $4.67M
with calls +$1.05M, NLV $4.32M
+$79k
NLV -$42k
with calls -$271k, NLV -$392k
S&P +20%
+$1.72M
NLV $5.11M
+$1.94M
NLV $5.21M
with calls +$1.45M, NLV $4.71M
+$224k
NLV +$103k
with calls -$275k, NLV -$396k
S&P +25%
+$2.12M
NLV $5.51M
+$2.52M
NLV $5.78M
with calls +$1.87M, NLV $5.14M
+$397k
NLV +$276k
with calls -$249k, NLV -$370k
NLV here counts all your money so both sides compare: the IBKR accounts ($2.56M today) plus the $66k spare cash and the 10 BTC ($767k). The big number in each cell is the gain; "Reset vs hold" shows the gap in gain, then the gap in NLV, which also carries the reset's $125k friction.
Hold keeps the calls it has written today; the top Reset line has no weekly calls on yet. "With calls" stops each income name at about +8%, roughly where a weekly call would, so the income book makes at most about $109k. The first line flatters the reset and the second is harsh on it (no premium, no rolling): the truth sits between them, and the bigger the move, the wider the gap. Moves this large are a straight line on beta (MARA about +76% at S&P +25%), a rough guide only.
By account
Each cell is the value with the gain under it. Index LEAPS volatility eases about 0.5 points per +10%, as the rest of this page does; everything else is volatility held still. Hold's spare cash and 10 BTC sit outside the IBKR accounts; on the reset they are already moved in.
The index: $929k of shares and $266k of LEAPS in Joint.
Now
+5%
+10%
+15%
+20%
+25%
Shares
$929k
$979k
+$50k
$1.03M
+$99k
$1.08M
+$149k
$1.13M
+$198k
$1.18M
+$248k
LEAPS
$266k
$321k
+$54k
$379k
+$113k
$441k
+$175k
$506k
+$240k
$574k
+$308k
Index total
$1.20M
$1.30M
+$104k
$1.41M
+$212k
$1.52M
+$323k
$1.63M
+$438k
$1.75M
+$555k
Index return
–
+8.7%
+17.7%
+27.0%
+36.6%
+46.4%
Joint return, with the $500k cash
–
+6.1%
+12.5%
+19.1%
+25.8%
+32.7%
The index over the years: the $1.20M sleeve with the cash left out and no calls sold.
SPY/QQQ
3 years
5 years
10 years
18% a year
$2.29M
shares only $1.96M
$3.55M
shares only $2.74M
$10.58M
shares only $6.26M
16% a year
$2.10M
shares only $1.87M
$3.06M
shares only $2.51M
$7.86M
shares only $5.27M
14% a year
$1.92M
shares only $1.77M
$2.63M
shares only $2.30M
$5.80M
shares only $4.43M
10% a year
$1.59M
shares only $1.59M
$1.92M
shares only $1.93M
$3.11M
shares only $3.10M
The actual path from 31 Oct 2016
$2.01M
shares only $1.83M
$4.28M
shares only $3.13M
$8.02M
shares only $5.37M
How it runs: every year the LEAPS roll to about 2 years out at the same distance from the price and go back to 22% of the sleeve; the excess buys shares, 60/40. Each roll pays a quarter of the spread each way. Volatility stays at today's, and dividends are left out of every line.
The last 10 years: 60/40 made +17.5% a year with dividends and +16.2% on price alone (SPY +15.3%, QQQ +20.2% with dividends). The LEAPS only see the price move, so the actual path is the fairest read: about $8.02M at 10 years. It reaches its 5-year mark just before the 2022 fall, which is why that cell looks so strong.
The road matters less than the return: ending at the same 18% price by a bumpy road gives $10.32M to $11.03M at 10 years. If you never trim and every dollar stays in LEAPS, 10 years is about $17.05M, but LEAPS grow to 71% of the sleeve.
The LEAPS need the return: at about 10% a year they add nothing over shares.
Reset, by account
Now
+5%
+10%
+15%
+20%
+25%
Main: income book
$790k
$943k
+$153k
$1.10M
+$306k
$1.25M
+$459k
$1.40M
+$608k
$1.55M
+$755k
Main, with calls at +8%
$790k
$897k
+$107k
$899k
+$109k
$899k
+$109k
$899k
+$109k
$899k
+$109k
Joint: index + cash
$1.70M
$1.80M
+$104k
$1.91M
+$212k
$2.02M
+$323k
$2.13M
+$438k
$2.25M
+$555k
Neville: crypto sleeve
$339k
$401k
+$62k
$469k
+$129k
$541k
+$202k
$654k
+$314k
$789k
+$449k
Neville: IBIT calls + bets
$442k
$569k
+$127k
$710k
+$267k
$863k
+$420k
$1.03M
+$584k
$1.20M
+$757k
Total
$3.27M
$3.71M
+$445k
$4.18M
+$914k
$4.67M
+$1.40M
$5.21M
+$1.94M
$5.78M
+$2.52M
Total, with calls
$3.27M
$3.67M
+$399k
$3.98M
+$717k
$4.32M
+$1.05M
$4.71M
+$1.45M
$5.14M
+$1.87M
Hold, by account
Now
+5%
+10%
+15%
+20%
+25%
Main
$1.20M
$1.41M
+$210k
$1.58M
+$386k
$1.74M
+$546k
$1.90M
+$707k
$2.07M
+$868k
Joint
$590k
$706k
+$116k
$814k
+$223k
$922k
+$332k
$1.03M
+$441k
$1.15M
+$557k
Neville
$460k
$536k
+$76k
$586k
+$126k
$624k
+$165k
$657k
+$197k
$688k
+$228k
RetireInc
$308k
$352k
+$44k
$391k
+$83k
$429k
+$121k
$466k
+$158k
$504k
+$196k
10 BTC
$767k
$822k
+$54k
$876k
+$108k
$930k
+$162k
$984k
+$216k
$1.04M
+$271k
Spare cash
$66k
$66k
+$0
$66k
+$0
$66k
+$0
$66k
+$0
$66k
+$0
Total
$3.39M
$3.89M
+$500k
$4.31M
+$927k
$4.71M
+$1.33M
$5.11M
+$1.72M
$5.51M
+$2.12M
The 22 Sep to 30 Oct move, reversed and again
Hold
Reset
Reset vs hold
Now (Reset: once done)
$3.39M
$3.27M
-$121k
Snap-back: every name to its 22 Sep close
+$1.28M
+$699k
-$583k
The same fall again
-$969k
-$392k
+$577k
NLV after snap-back
$4.67M
$3.97M
with calls $3.81M
-$704k
with calls -$856k
NLV after fall again
$2.42M
$2.88M
+$456k
A full snap-back favours holding. The names that fell hardest bounce hardest (MARA ×2.01, BMNR ×1.43, SPY ×1.02, QQQ ×1.08), and today's book owns far more of them: MARA +$307k, BMNR +$274k, MU +$87k, COIN +$77k.
The reset keeps most of it through the crypto sleeve (+$292k), the income book (+$214k), the IBIT calls and bets (+$102k) and the index (+$92k). If weekly calls stop each income name at about +8%, the income book makes +$61k instead.
A repeat favours the reset: it loses $392k against $969k. It gives up about $583k of the best case to save about $577k in the bad one.
Carry on with today's book
Pro: no friction (about $125k to $247k to unwind and rebuild) and no selling at the bottom of a sharp two-day fall. If the names you hold bounce, you keep every leg of it: +10% on the S&P is worth about $818k.
Pro: the hedges, rent and LEAPS time value are already paid. RetireInc, the cleanest read of the strategy, beat its own names on the way down.
Pro: it keeps the full crypto road with a floor: shares are worth something at any price.
Pro: it earns more today, about $52,146/mo net of rent against the reset's $11,248 at your rate.
Con: the cockpit levels you want to get away from: S&P −20% leaves about $1.03M (-60%), −30% about $811k. Main's Max Loss is above its NLV.
Con: it is one bet: high beta, semis and crypto. That bet cost $986k in two months while the S&P was flat, and the cockpit's S&P grid does not show it.
Con: Joint carries a $382k margin loan, and 0 of the 32 income fortresses are below design or not paying their hedge.
Con: it takes daily work across ~1,000 contracts and four accounts to keep it safe; that work is the risk when you are away.
Reset now
Pro: no liquidation path: the income account holds its Max Loss as cash, the crypto calls can only lose their cost, the index and cash have no margin call.
Pro: the deep end improves: S&P −20% leaves about $1.84M (-38%) instead of $1.03M (-60%). A crypto-only or high-beta crash, like this one, costs a capped amount.
Pro: a simple book: 13 fortresses, two calls, two ETFs with two LEAPS, and cash. A $500k reserve to add on a deeper fall.
Pro: crypto upside at a known cost: $2.25M at MARA $40 / BMNR $70 with $203k now (up to about $475k after the BMNR add) at risk, and runway to Dec '28 / Jan '29.
Con: about $125k of friction ($247k if filled at the natural), paid after the fall. And it is still a geared portfolio: S&P −10% costs about 21% of it.
Con: at your 90-day hit rate the income drops to about $11,248/mo. The book basis needs a better hit rate than you have achieved across a whole book; the reserve covers the gap.
Con: the crypto calls are worth $0 below MARA $25 / BMNR $40 at Dec '28 / Jan '29, and at this size they carry about 56% of IDEAL's payoff ($2.25M at MARA $40 / BMNR $70 vs the book's $3.99M).
Con: a big execution job: about 4,109 contracts to close and 1,011 to open, in RTH, without legging into a gap.
My recommendation: reset, in four stages
Week 1, crypto first (the biggest single risk): close every MARA / BMNR / IREN / GLXY / CLSK / MSTR / COIN position and buy the 500 MARA + 250 BMNR, 250 more at BMNR $30 sleeve in the same sessions. That clears Joint's margin loan and caps crypto at $203k (the 30 Oct partial-reset pricing: about $445k freed).
Weeks 1-2, the high-beta and not-paying fortresses: ENPH, CRWV, AMZN Joint (not paying their hedge), then the below-design ones. Close into up days, as combos, with the Desk's NUDGE.
Weeks 2-3, build HBMax as cash frees up, one name at a time, each after its earnings, every call written the first Friday.
Index in one go: right after the reset, as far as Joint's cash goes with the $500k reserve kept; the rest of the shares the day the BTC money lands. The reserve is never touched for the index.
Why not all at once: the fall was two days old on 30 Oct, and spreads on the illiquid legs (GLXY, COIN, ASST, QUBT, COPX) cost the most at the natural; $114k of the $210k natural-spread cost sits in just ten legs. Why not hold: the goal you set is the cockpit levels, and holding leaves them where they are.
Which account holds what (your split)
Account
NLV 30 Oct
Holds
Why
Main
$1,191,687
HBMax ($940k)
the only account big enough to hold the whole income book with Max Loss as cash
Joint
$578,126
the index ($929k in shares (738 SPY, 538 QQQ) + $266k in Dec '28 LEAPS (12x SPY 735C, 8x QQQ 675C)) + the $500k reserve
shares, paid-for LEAPS and cash, no margin loan: it cannot be margin-called; calls on the shares and the LEAPS in one place
Neville
$457,532
convexity: crypto sleeve $203k plus $136k cash for the BMNR add, IBIT calls and the bets
long calls only, each can lose only its cost; kept apart from the index
RetireInc
$307,185
closed
its cash moves to the other three
Accounts cannot share margin, so each job has its own. IBKR transfers between your accounts move cash, not positions.
Your index: $1.20M, with $500k kept in cash. $929k in shares (738 SPY, 538 QQQ) and $266k in at-the-money Dec '28 LEAPS (12x SPY 735C, 8x QQQ 675C), 60% SPY / 40% QQQ. About $2.07M of S&P-equivalent exposure. Bought in one go right after the reset: all the LEAPS and every share Joint can pay for with the reserve kept at $500k, the rest of the shares the day the BTC money lands (Monday plan tab). Calls on every lot and LEAPS (32 contracts), written your way: $14,910/mo on the book basis; your own SPY / QQQ record per contract gives about $13,509. Your way, not every Friday: 1-3% out, 5-30 days, close early for pennies, sit out the week after a run-up, write after dips; when challenged, roll up and out and wait (an index moves far less than IREN or MU). At S&P −20% the index loses about $325k; +10% adds about $212k.
How big the LEAPS slice should be
The $1.2M index and $500k of cash ($1.7M), the Index + Cash model (index_size.py): each extra $100k of LEAPS adds about $350k of exposure and costs about $27k in a year when the market goes nowhere. Your pick: $270k, the 1.2x swap. About $120k moves from shares into LEAPS, so the $1.7M runs about 1.2x S&P-equivalent with the $500k reserve untouched and no margin loan; each LEAPS can only lose what it cost, so a crash costs less than reaching 1.2x with more shares. Rolling 1-year results use every start day since 1999 (median / 1-in-20 bad, on the $1.7M).
LEAPS in the $1.2M
Exposure
Flat year
S&P −20%
S&P +20%
1-year: median / 1-in-20 bad
2008 trough
2022 trough
all shares
$1.33M
1.10x the $1.2M
$25k
-$265k
$265k
+10.4% / -20.1%
-$640k
-$340k
$150k
$1.73M
1.44x the $1.2M
-$17k
-$323k
$387k
+12.0% / -25.6%
-$705k
-$425k
$270k your pick
$2.07M
1.73x the $1.2M
-$51k
-$372k
$488k
+13.3% / -30.2%
-$760k
-$496k
$350k
$2.30M
1.91x the $1.2M
-$73k
-$404k
$555k
+14.2% / -33.1%
-$796k
-$543k
Calls on the index: your way, not every Friday
Your own record (Trade Query, every SPY / QQQ call closed) against the fixed rule (a call every Friday, held to expiry) over the same weeks, priced at the volatility your own fills got (implied vol over that day's VIX / VXN; index_calls_record.py). The difference is the discretion: on SPY a call was on only about half the time, skipping the weeks after run-ups, and most were closed early for pennies.
Trades
Contracts
Net
/ month
/ contract / month
Time a call was on
Your fills vs VIX / VXN
Fixed rule, 0.25 delta
Fixed rule, 0.15 delta
SPY
22 calls
2026-02-18 to 2026-10-27
11.5
$18,778
$2,277
$197
52%
0.68x
-$26,829
-$17,511
QQQ
8 calls
2026-02-06 to 2026-04-13
10.0
$16,280
$7,509
$751
92%
0.72x
-$28,818
-$21,168
The fixed rule since 2000
Weekly calls every Friday on the real SPY / QQQ path (index_calls_bt.py), priced at the day's VIX / VXN scaled to the call wing (the 30 Oct quotes: SPY 0.77x, QQQ 0.91x), sold at 97%, bought back at intrinsic at expiry; % of the position per year. Out-of-the-money index calls are priced close to fair and the index rises most years, so written mechanically they hand back most of the premium. That is why the rule above is discretion, not a calendar.
Call
Window
Premium / yr
Kept
Net / yr
Positive years
Worst year
SPY 0.25 delta
2000-2026
16.2%
5%
+0.8%
17 of 27
-9.2% (2020)
SPY 0.15 delta
2000-2026
8.4%
17%
+1.4%
19 of 27
-5.9% (2000)
QQQ 0.25 delta
2000-2026
24.3%
15%
+3.6%
20 of 27
-43.9% (2000)
QQQ 0.15 delta
2000-2026
12.7%
30%
+3.8%
22 of 27
-31.5% (2000)
SPY 0.25 delta
last 10y
13.9%
-12%
-1.7%
3 of 11
-9.2% (2020)
SPY 0.15 delta
last 10y
7.3%
-1%
-0.1%
5 of 11
-5.2% (2022)
QQQ 0.25 delta
last 10y
20.1%
7%
+1.3%
8 of 11
-6.7% (2023)
QQQ 0.15 delta
last 10y
10.5%
24%
+2.6%
9 of 11
-2.4% (2023)
Appendixthe 31 Oct study: $1.3M of index + cash, five mixes, crash and crisis tests, LEAPS shapes, the SPY / QQQ split
Your pick (5), now the plan: $800k cash, $350k shares + $150k ATM Dec '28 LEAPS, 70% SPY / 30% QQQ. 324 SPY shares ($245k) 8x SPY Dec '28 735C ($104k) 152 QQQ shares ($105k) 3x QQQ Dec '28 675C ($41k) S&P −20%: -$166k (my mix -$144k, option 1 -$113k); +20%: +$229k (my mix +$184k, option 1 +$113k); a flat year: -$10k (my mix +$5k); 2008 low: -$323k (my mix -$297k). Typical 12 months +8.4%, 1 in 20 -16.5%. It is the most geared of the cash-heavy mixes: the same 70/30 split as my mix, with $50k more in LEAPS and $50k less in shares.Every share lot and LEAPS writes the weekly 0.25-delta call (Total Income tab).
In plain words
This $1.3M is the only ballast in the reset. The rest of the plan (the HBMAX income book and the crypto sleeve, $1.12M together) loses about $781k at S&P −20% on the Risk Cockpit grid. So the question is less "which earns more" than "how much of the next fall do you want this money to absorb".
Option 1 vs option 2 is a straight dial: every $100k moved from cash to shares adds about $23k of loss at S&P −20% and the same gain at +20%, and costs about $3k a year of interest. Over every 12-month window since 1999 the median return is +6.5% (option 1) vs +9.5% (option 2); the worst is -17.3% vs -29.9%.
LEAPS change the shape, not just the size. Swapping a fifth of option 1's shares for at-the-money Dec '28 calls (3a) gains about as much as option 2 when the market rises (+$205k vs +$181k at +20%), loses less in a deep bear (2008: -$309k vs -$421k), and keeps $793k in cash. The price is the calls' time value: about $31k a year if the market goes nowhere.
My mix (4) is option 1 plus option 3, tilted to SPY: $800k cash, $400k shares + $100k ATM Dec '28 LEAPS, 70% SPY / 30% QQQ. In the two great bears it loses close to what option 1 loses (dot-com -$260k vs -$261k, 2008 -$297k vs -$255k) and far less than option 2 (-$483k, -$421k), with the upside of option 2 (+$184k vs +$181k at +20%) and $2,274/mo of interest on the cash. The 70/30 tilt is because your income book is already a Nasdaq bet: its names move 0.85 with QQQ.
Whole portfolio at S&P −20%: about -$894k with option 1, -$962k with option 2, -$925k with my mix (the rest of the reset included).
Model output from your own data and public prices, not licensed financial advice. Nothing on this page places an order.
Change in the $1.3M on an immediate S&P move. QQQ moves its measured 2-year beta to SPY (1.26x). LEAPS repriced on the Dec '28 smile from the CBOE close, with implied vol +3.5 points per 10% fall. The "whole reset" rows add the income book and crypto sleeve from the PortfolioReset grid (cockpit betas, vol held still).
S&P move (QQQ moves 1.26x)
1. $800k cash
2. $500k cash
3a. 1 + LEAPS
3b. 2 + LEAPS
4. Claude's mix
5. Your pick
+30%
+$170k
+$271k
+$313k
+$484k
+$281k
+$352k
+20%
+$113k
+$181k
+$205k
+$316k
+$184k
+$229k
+10%
+$57k
+$90k
+$99k
+$154k
+$90k
+$111k
-10%
-$57k
-$90k
-$84k
-$131k
-$76k
-$90k
-20%
-$113k
-$181k
-$158k
-$246k
-$144k
-$166k
-30%
-$170k
-$271k
-$220k
-$344k
-$202k
-$227k
-40%
-$226k
-$362k
-$275k
-$431k
-$253k
-$277k
Whole reset at -10%
-$517k
rest of book -$461k
-$551k
rest of book -$461k
-$545k
rest of book -$461k
-$591k
rest of book -$461k
-$537k
rest of book -$461k
-$550k
rest of book -$461k
Whole reset at -20%
-$894k
rest of book -$781k
-$962k
rest of book -$781k
-$939k
rest of book -$781k
-$1.03M
rest of book -$781k
-$925k
rest of book -$781k
-$947k
rest of book -$781k
Whole reset at -30%
-$1.10M
rest of book -$934k
-$1.21M
rest of book -$934k
-$1.15M
rest of book -$934k
-$1.28M
rest of book -$934k
-$1.14M
rest of book -$934k
-$1.16M
rest of book -$934k
A year from now
Model, on today's prices: cash at 3.38%, dividends after the 30% US withholding, LEAPS aged a year on the same smile.
1. $800k cash
2. $500k cash
3a. 1 + LEAPS
3b. 2 + LEAPS
4. Claude's mix
5. Your pick
Market flat for a year
+$30k
+$22k
-$537
-$24k
+$5k
-$10k
S&P −20% over a year
-$84k
-$160k
-$150k
-$257k
-$133k
-$165k
S&P +20% over a year
+$144k
+$204k
+$206k
+$295k
+$190k
+$221k
Interest + dividends / mo
$2,501
$1,804
$2,431
$1,723
$2,503
$2,467
Every 12-month window since Mar 1999
6,695 start days, 1999-03-10 to 2026-10-30: SPY and QQQ total return (dividends after withholding), cash at the 13-week T-bill rate less 0.5%, LEAPS bought at today's cost per dollar of index and marked with implied vol moving half as much as VIX / VXN. Returns are on the whole $1.3M.
1. $800k cash
2. $500k cash
3a. 1 + LEAPS
3b. 2 + LEAPS
4. Claude's mix
5. Your pick
Bad year (1 in 20)
-10.8%
-18.6%
-16.2%
-26.3%
-13.8%
-16.5%
Below-par year (1 in 4)
+1.8%
+2.2%
+0.2%
-0.1%
+0.3%
-0.4%
Typical year (median)
+6.5%
+9.5%
+8.2%
+11.9%
+7.5%
+8.4%
Good year (1 in 4)
+10.4%
+15.6%
+15.3%
+22.7%
+13.4%
+15.6%
Great year (1 in 20)
+16.6%
+25.1%
+27.0%
+40.4%
+23.1%
+27.5%
Worst 12 months
-17.3%
-29.9%
-21.2%
-35.5%
-20.4%
-22.7%
Chance of losing over 10%
6%
13%
13%
17%
12%
14%
Typical 2 years (median)
+11.6%
+16.7%
+13.9%
+20.0%
+12.5%
+13.6%
Worst 2 years
-20.1%
-35.8%
-23.7%
-40.5%
-20.4%
-22.4%
Bought at the top of the last five falls
Each mix bought on the day the S&P peaked, held with no changes for up to 3 years (to 30 Oct 2026 for 2025). Main number = the lowest it fell to; below it the month of the low and the result after 1 and 3 years. A LEAPS that expires is held as cash, never rolled.
Bought at the peak of
1. $800k cash
2. $500k cash
3a. 1 + LEAPS
3b. 2 + LEAPS
4. Claude's mix
5. Your pick
Dot-com
2000-03-24
-$261k
low 2002-10 · 1y -$179k · 3y -$233k
-$483k
low 2002-10 · 1y -$327k · 3y -$440k
-$304k
low 2002-10 · 1y -$239k · 3y -$280k
-$539k
low 2002-10 · 1y -$412k · 3y -$504k
-$260k
low 2002-10 · 1y -$190k · 3y -$236k
-$284k
low 2002-10 · 1y -$223k · 3y -$262k
2008
2007-10-09
-$255k
low 2009-03 · 1y -$177k · 3y -$56k
-$421k
low 2009-03 · 1y -$296k · 3y -$103k
-$309k
low 2009-03 · 1y -$232k · 3y -$150k
-$496k
low 2009-03 · 1y -$374k · 3y -$241k
-$297k
low 2009-03 · 1y -$218k · 3y -$148k
-$323k
low 2009-03 · 1y -$245k · 3y -$193k
Covid
2020-02-19
-$154k
low 2020-03 · 1y +$145k · 3y +$150k
-$247k
low 2020-03 · 1y +$232k · 3y +$223k
-$178k
low 2020-03 · 1y +$239k · 3y +$301k
-$280k
low 2020-03 · 1y +$369k · 3y +$448k
-$178k
low 2020-03 · 1y +$183k · 3y +$262k
-$191k
low 2020-03 · 1y +$218k · 3y +$330k
2022
2022-01-03
-$142k
low 2022-10 · 1y -$118k · 3y +$232k
-$234k
low 2022-10 · 1y -$201k · 3y +$284k
-$207k
low 2022-10 · 1y -$191k · 3y +$159k
-$328k
low 2022-10 · 1y -$307k · 3y +$173k
-$185k
low 2022-10 · 1y -$164k · 3y +$164k
-$217k
low 2022-10 · 1y -$200k · 3y +$128k
Tariffs 2025
2025-02-19
-$100k
low 2025-04 · 1y +$95k · 3y +$181k
-$164k
low 2025-04 · 1y +$124k · 3y +$244k
-$123k
low 2025-04 · 1y +$119k · 3y +$238k
-$197k
low 2025-04 · 1y +$160k · 3y +$328k
-$108k
low 2025-04 · 1y +$117k · 3y +$226k
-$122k
low 2025-04 · 1y +$131k · 3y +$254k
QQQ fell 83% from March 2000 to October 2002; that is why the dot-com row punishes QQQ-heavy mixes. In 2008 the calls bought at the peak were nearly worthless by March 2009, so 3a loses more than option 1 (its $100k of calls is almost all gone, where $100k of shares lost about half) but far less than option 2, which has the same upside. The loss on the calls can never exceed what was paid.
Which LEAPS for the leveraged slice
CBOE close of 30 Oct, Dec '28 expiry (2.1 years). The year-ahead columns are the option's own % change (for comparison, the shares: SPY -19% / +21%, QQQ -25% / +26% at −20% / +20%, dividends included).
LEAPS (Dec 15 '28)
Cost / share (mid + ¼ spread)
Bid / ask
Delta
Leverage (exposure per $)
Breakeven at expiry
Open interest
Market flat 1 year
Index −20% in a year
Index +20% in a year
SPY 600C
deep in the money
$227.12
223.50 / 228.32
0.82
2.7x
+9.4%
296
-12%
-63%
+55%
SPY 735C
at the money
$130.46
127.33 / 131.50
0.65
3.8x
+14.5%
1,257
-27%
-81%
+78%
SPY 820C
out of the money
$80.28
77.39 / 81.24
0.50
4.7x
+19.1%
823
-42%
-88%
+106%
QQQ 500C
deep in the money
$256.75
253.00 / 258.00
0.86
2.3x
+9.6%
229
-11%
-65%
+56%
QQQ 675C
at the money
$137.38
134.00 / 138.50
0.66
3.3x
+17.7%
321
-28%
-84%
+81%
QQQ 780C
out of the money
$83.25
79.50 / 84.50
0.49
4.1x
+25.0%
362
-44%
-89%
+103%
Deep in the money is not really leverage. A 0.82-delta call plus the cash it frees behaves like the shares with a put about 20% below: it costs about the put's price a year and gives back little extra upside. Good for "shares with a floor", not for "leveraged upside".
At the money is the honest leveraged slice: about 3.3-3.8x the exposure per dollar, all of it can be lost, and it loses roughly a quarter of its value in a flat year. In a 20% rally it makes nearly four times what the shares make.
Out of the money gives the most upside per dollar but needs SPY +19% / QQQ +25% by Dec '28 just to break even, and loses about 40% in a flat year.
Are they expensive? SPY's at-the-money Dec '28 vol is 19.5% against 12.8% realised over the last year, 16.4% over two and 19.2% since 1999; QQQ 24.4% against 20.3%, 21.9% and 26.9%. Priced at the long-run average, above the calm of the last two years. Not cheap, not stretched.
Why Dec '28 and not Jan '29 or Dec '27: Dec '28 has far more open interest than Jan '29 at these strikes (tighter fills), and over two years of life it loses less per year than Dec '27. Buy on a limit near the mid; the bid/ask on these is $3-5.
SPY or QQQ
Option 1 ($800k cash / $500k shares) at four splits.
SPY / QQQ
S&P −20%
S&P −30%
Dot-com low
Typical year
Bad year (1 in 20)
100 / 0
-$100k
-$150k
-$174k
+5.6%
-7.4%
70 / 30
-$108k
-$162k
-$226k
+6.2%
-9.3%
50 / 50
-$113k
-$170k
-$261k
+6.5%
-10.8%
0 / 100
-$126k
-$189k
-$349k
+7.2%
-13.9%
Going from all SPY to all QQQ adds about 1.6 points to a typical year and costs more in every fall (dot-com low -$349k vs -$174k). The bigger reason to lean on SPY is the rest of your book: the Income_HB750 names move 0.85 with QQQ and 0.79 with SPY over the last 180 days, and 0.31 with the part of QQQ that SPY does not explain. QQQ adds more of the same tech bet the income book already carries; SPY adds the banks, healthcare, energy and industrials the income book does not own. 70/30 keeps a Nasdaq tilt without doubling it.
S&P-equivalent exposure $791k (option 1: $566k, option 2: $905k). Cash $807k earning about $2,274/mo. Why this one: it keeps option 1's cash (the reset's shock absorber, and the reserve that covers months the income book does not pay), buys option 2's upside with the LEAPS instead of more shares, and leans to SPY because the income book already holds the Nasdaq. What it gives up: about $25k a year against option 1 if the market goes sideways, and about $31k more loss than option 1 at S&P −20%.
about $20k a year more than the at-the-money slice lost in a flat market
Same, all SPY
-$143k
+$188k
+$2k
+7.0%
-13.0%
-$314k
gives up a little upside for a little less dot-com risk; fine too
$750k cash / $450k shares / $100k LEAPS
-$154k
+$195k
+$3k
+8.0%
-15.0%
-$325k
more crash loss for half a point of typical return
$850k cash / $350k shares / $100k LEAPS
-$133k
+$173k
+$6k
+7.1%
-12.7%
-$269k
fine if you want more dry powder; less upside
Same + cash buys $100k shares at S&P -20% and -30%
-$144k
+$184k
+$5k
+7.6%
-13.5%
-$374k
buying into a falling market deepens the 2008 low; the reserve is not for the index
Doing it
Shares in thirds, as the reset plan already says: a third when the income book is built, the rest over the following weeks or at each further 5% S&P drop. The LEAPS in two clips, on down days, on a limit near the mid.
Where: Neville and RetireInc (the reset plan's "index + reserve" accounts). Shares and LEAPS are paid in full, so they need no margin and cannot be force-sold.
The cash can earn more. Idle USD at IBKR earns the benchmark less 0.5% (about 3.38% today); 13-week T-bills yielded 3.97% at the 30 Oct close. On $800k that gap is about $5k a year, for a ladder of bills you would roll every few months.
Ask your tax adviser about US estate tax. For a non-US person, US-listed ETFs such as SPY and QQQ are US-situs assets: the US can tax them at up to 40% above $60k at death. Ireland-listed equivalents (CSPX for the S&P 500, CNDX / EQQQ for the Nasdaq 100) are not US-situs and suffer 15% dividend withholding inside the fund instead of 30%. Your current US stocks and options carry the same estate exposure today, so this matters for the long-term shares more than for anything short-dated.
Every order is yours, in TWS or the Desk's Portfolio (TWS) tab. Nothing here places one.
Method and caveats
Prices: SPY 755.72, QQQ 690.49 (30 Oct close). LEAPS from the bundle's CBOE close quotes (quotes.json), bought at mid + a quarter of the spread; exits at model mid (no exit spread).
LEAPS model: Black-Scholes on the Dec '28 smile by moneyness, calibrated to each strike's mid, 3.7% rate, trailing dividend yield (SPY 1.19%, QQQ 0.51%). Scenarios add 3.5 vol points per 10% fall; history moves implied vol half as much as VIX (SPY) or VXN (QQQ; VIX x 1.25 before 2001).
History (fetch_hist.py, Yahoo): applies today's LEAPS cost per dollar to every past start date; when vol was cheaper or dearer the real cost differed. No rebalancing. Shares get dividends after 30% withholding; cash gets the 13-week T-bill less 0.5%.
Rest of the reset: the PortfolioReset grid for HBMAX and the 500 MARA + 500 BMNR crypto sleeve (analysis.json), cockpit betas, vol held still. That grid is an S&P move; a high-beta-only fall like October's hurts the rest of the book more than it shows.
Total income of the reset: $49,087 a month on the book basis. Plan on about $32k to $44k. HBMax $22,548, and the buffer around it $26,539: SPY/QQQ calls and dividends $15,616, crypto weekly calls $2,483, IBIT weekly calls $2,453, the bets' weekly calls $3,347, interest $2,640. Every call line is the model's premium at the bid times the same book haircuts (your trade-tape keep 71% x the INCOME_BOOK trade ratio 0.52), after hedge rolls and $0.49 a contract of commissions. The low end: HBMax on the scanner pick held to expiry ($12,521) plus the buffer at last year's real record ($19,871). The high end: the max-income pick ($17,398) plus the buffer at the book basis.
Monthly income, by piece
Last 12 months, real record (a check, not in the total): HBMax at your 90-day rate; the index calls at your own SPY / QQQ record per contract; MARA / BMNR calls at the share of premium the same rule kept on the real path over the last 52 weeks; IBIT and the bets at their real-path result per month (SPCX since June). Last year rose for the index (its calls gave the premium back) and fell for crypto (their calls kept it).
Piece
Capital
Call premium (model, at the bid)
Book haircut (keep x trade ratio)
Hedge rolls (put rent)
Commissions
Dividends / interest
Book income / mo
Last 12 months, real record / mo
HBMax income book
$940k
Max Loss $866k
$71,417
-$43,872
-$4,601
-$396
–
$22,548
$10,853
SPY / QQQ weekly calls, written your way
$1.20M
32 contracts (shares + LEAPS)
$40,716
-$25,669
–
-$136
$705
dividends
$15,616
$14,214
Crypto sleeve, 0.20-delta 7-day calls on 25%
$203k
125 MARA / 62 BMNR covered
$8,695
-$5,482
–
-$730
–
$2,483
$6,090
IBIT, 0.20-delta 7-day calls on 25%
$125k
75 of 300 IBIT 90C covered
$7,501
-$4,729
–
-$319
–
$2,453
$1,444
Bets (SPCX, LLY), 0.20-delta 7-day calls on 25% (LLY 3 of 5)
$318k
8 SPCX / 3 LLY covered; MU skipped
$9,173
-$5,783
–
-$43
–
$3,347
-$4,517
Cash: reserve + income account idle cash
$500k + $437k
–
–
–
–
$2,640
at 3.38%
$2,640
$2,640
Total
$137,501
-$85,535
-$4,601
-$1,624
$3,345
$49,087
$30,724
What to plan on
HBMax lines after hedge rolls and commissions; the plan rows from the Income tab's backtest (real implied vol, Jun-Oct 2026). The buffer's call lines lean opposite ways: index calls pay in flat or falling markets, crypto, IBIT and bet calls pay when those names fall, and in a broad rally every call line shrinks while the LEAPS and the index gain. Interest is the only certain line. Raising the crypto, IBIT and bet calls from 25% to 50% adds about $8,283/mo on the book basis.
Case
HBMax
Buffer
Total / mo
Scanner pick, held to expiry (the plan)
$12,521
$19,871
$32,392
Max-income pick, held to expiry
$17,398
$26,539
$43,937
Book basis (this page's total)
$22,548
$26,539
$49,087
Last 12 months, real record
$10,853
$19,871
$30,724
AppendixHBMax name by name, the SPY / QQQ calls, the crypto and bet weekly calls, and how far book is from what to expect
HBMax: name by name
Your picks of 3 Nov (hbmax_set.py), Max Loss at most about $875k. Premium = the weekly call the fortress_v28 run picked (about 2-3% out), earnings-corrected. Commissions = a sell and a close per contract per week, plus two legs per contract at each hedge roll (about every 100 days).
Name
Contracts
Max Loss
Premium / mo
Keep
Trade ratio
Kept
Hedge rolls
Commissions
Book income / mo
DELL
3
$95,250
$9,491
76%
0.47
$3,414
-$445
-$14
$2,956
COIN
8
$95,192
$8,509
68%
0.52
$3,025
-$109
-$36
$2,879
MCD
10
$89,640
$5,438
71%
0.52
$2,007
-$366
-$45
$1,596
XOM
13
$77,012
$8,686
70%
0.23
$1,400
-$686
-$59
$656
SPCX
8
$72,016
$7,406
77%
0.78
$4,416
-$373
-$36
$4,007
NVDA
7
$69,090
$5,520
68%
0.52
$1,957
-$566
-$32
$1,358
CI
5
$66,825
$2,749
71%
0.52
$1,014
-$236
-$23
$755
LLY
1
$65,600
$5,073
71%
0.52
$1,872
-$558
-$5
$1,309
MU
1
$55,560
$6,557
80%
0.77
$4,015
-$195
-$5
$3,816
ABT
12
$50,580
$2,782
71%
0.52
$1,027
-$218
-$55
$754
JPM
4
$48,076
$3,436
71%
0.52
$1,268
-$245
-$18
$1,005
AXP
3
$45,066
$2,205
71%
0.52
$814
-$219
-$14
$581
PEP
12
$36,264
$3,564
71%
0.52
$1,315
-$385
-$55
$876
HBMax
87
$866,171
$71,417
$27,545
-$4,601
-$396
$22,548
SPY / QQQ covered calls
$929k in shares (738 SPY, 538 QQQ) + $266k in Dec '28 LEAPS (12x SPY 735C, 8x QQQ 675C). Every round lot and LEAPS carries a call, written your way (owner 2026-11-02); modelled as the 7-day call nearest 0.25 delta (CBOE bid at the 30 Oct close). A call written against a LEAPS is a diagonal: the LEAPS backs it, so it needs no extra cash, and an assignment is met by selling or exercising the LEAPS. The LEAPS cover 20 contracts for $266k, against 12 for $929k of shares: that is the leverage, and it cuts both ways (S&P −20% costs the index $325k, Overview).
ETF
Holding
Contracts
Call
Bid / ask
Delta
IV
Premium / mo
Book kept
Commissions
Dividends
Book income / mo
Kept on the real path (last 52 weeks)
SPY
738 shares + 12x 735C Dec '28
$714k; 38 shares not covered
19
7 on shares, 12 on LEAPS
767C 6 Nov
2.67 / 2.74
0.25
15%
$22,060
$8,152
-$81
$534
$8,606
-9%
QQQ
538 shares + 8x 675C Dec '28
$481k; 38 shares not covered
13
5 on shares, 8 on LEAPS
711C 6 Nov
3.30 / 3.40
0.25
26%
$18,655
$6,894
-$55
$171
$7,010
-1%
Crypto weekly calls
The CRYPTO_SLEEVE page's rule on a quarter of the sleeve (500 MARA + 250 BMNR, 250 more at BMNR $30): the weekly call nearest 0.20 delta each Friday, closed the next Friday (penny buyback if out of the money), earnings weeks skipped (92% of weeks sold). Premium $8,695/mo, book $2,483 after $730 of commissions: at MARA's $0.08 bid the commission is about 6% of the premium. IBIT (the 300 90C bought with the BTC, Convexity tab): the same rule on 75, no earnings to skip, premium $7,501/mo, book $2,453 after $319 of commissions.
Name
Contracts covered
This week's call
Bid
MARA
125
8C
$0.08
BMNR
62
22.5C
$0.19
IBIT
75
46C
$0.23
SPCX
8
144C
$1.25
LLY
3
1475C
$4.25
How far book is from what to expect
HBMax at your own 90-day rate: $11,248/mo instead of $22,548.
Replays rank names; they do not forecast dollars. A mechanical replay shows $74.7k/mo against your real $63.5k/mo from June to October, and fails January to May. Your income comes mostly from closing winners early and selling into up days, which no fixed rule fully captures. Your own record covers DELL, SPCX, MU, COIN and NVDA; JPM, AXP, MCD, PEP, XOM, LLY, ABT and CI have none yet. Grade those eight after 60 to 90 days.
Index calls on the real path: the same 0.25-delta weekly kept -9% (SPY) and -1% (QQQ) of its premium over the last 52 weeks: in a rising year they hand most of it back.
Crypto calls: the CRYPTO_SLEEVE model expects about -$857/mo with no view and about -$8,160/mo if the IDEAL rally comes; the last year (a falling one) kept +78%. IBIT's calls: about $188/mo with no view, about -$3,712/mo if BTC doubles.
Opening the books (spreads and commissions) is a one-off cost, in the Overview's friction, not in these monthly lines.
Verdict for an income sleeve: fortress.
The income holds for two years. The fortress's CC-SS stays near the entry price as time passes, so the design weekly call stays sellable. A LEAPS's CC-SS climbs about half a point a month at a flat price and passes the design weekly after 4 to 18 months (most names 7 to 11), after which the scanner and FIGHT push the calls further out and the income falls. The LEAPS's $31,581/mo is a year-one number.
More at a flat price: $25,186 vs $21,465 a month once time value is counted.
Better in a drawdown: the FIGHT target stays fixed instead of running away, and the hedge put gains in a selloff and can be rolled down for cash.
Not losing everything: neither structure protects better in a crash (S&P −40%: -$873k vs -$838k). Liquidation: none by design: Max Loss $866k is under the $940k account, every short put is hedged, and the six-attack battery to −95% finds no crossing, as long as every hedge is rolled. The one line to know: the account falls under IBKR's ~$100k portfolio-margin minimum only at about a 42% S&P drop.
What you take on: short-put assignment on the names whose short puts sit closest to the money (JPM, AXP, ABT, PEP, XOM, MCD and NVDA) in a deep, long drawdown. When a short put's time value drops under about 3% of its strike, roll it down and out (sp_roll_fix.py) rather than wait.
HBMax as fortresses or as LEAPS
Same contracts = the HBMax count. Same Max Loss = each name keeps its share of the $866k. Same delta = contracts scaled so the S&P-weighted exposure roughly matches. Crash rows: the Risk Cockpit's betas, volatility held still.
Fortress (HBMax)
LEAPS, same contracts
LEAPS, same Max Loss
LEAPS, same delta
Contracts
87
87
102
114
Max Loss
$866k
$746k
$849k
$916k
Cash paid up front
$503k
$746k
$849k
$916k
Weekly calls, book basis
$27,545
$27,545
$31,758
$32,907
Hedge rent
-$4,601
0
0
0
Commissions
-$396
-$370
-$434
-$485
Interest on the unspent part of the $940k
$1,232
$546
$256
$67
Cash income / mo
$23,779
$27,720
$31,581
$32,489
Time-value carry at a flat price
+$1,407
-$8,453
-$10,116
-$10,649
Total at a flat price
$25,186
$19,267
$21,465
$21,840
S&P-weighted exposure (beta dollars)
$3.70M
$2.93M
$3.36M
$3.64M
S&P +10%
+$377k
+$307k
+$352k
+$381k
S&P -10%
-$376k
-$280k
-$320k
-$347k
S&P -20%
-$664k
-$514k
-$588k
-$636k
S&P -30%
-$810k
-$671k
-$766k
-$828k
S&P -40%
-$873k
-$736k
-$838k
-$908k
CC-SS over two years
CC-SS of the weekly call, as % above the stock price at the time, today / in 1 year / in 2 years. At entry both sit at about today's price: either structure is worth roughly what you paid, so you break even back at the entry price. Then they part: the fortress's short put loses time value about as fast as its long call, so its CC-SS stays put; a LEAPS has nothing to offset its decay, so its CC-SS rises. Same exit model as cc_scanner (today's mid as the anchor, the weekly's bid as the premium), hedge held at a constant tenor (it is rolled; its cost is the rent), volatility held still.
Name
Design weekly above spot
Flat price fortress
Flat price LEAPS
LEAPS CC-SS passes the weekly
Stock −20% fortress
Stock −20% LEAPS
Short-put time value left, −30%, in 1 yr
Stock bought if the put is assigned
DELL
+6.5%
-2% / -5% / -7%
-3% / +7% / +17%
12 mo
+23% / +18% / +17%
+22% / +33% / +46%
-0.0%
$123k
SPCX
+3.9%
+1% / 0% / -3%
0% / +5% / +9%
10 mo
+26% / +26% / +21%
+25% / +31% / +36%
19.9%
$84k
MU
+4.0%
0% / 0% / -3%
-1% / +4% / +8%
12 mo
+25% / +25% / +22%
+24% / +30% / +35%
22.4%
$63k
COIN
+6.6%
-1% / +1% / +1%
-2% / +10% / +24%
10 mo
+24% / +26% / +26%
+23% / +37% / +54%
8.8%
$120k
NVDA
+2.8%
0% / +2% / 0%
0% / +6% / +12%
6 mo
+26% / +27% / +26%
+25% / +33% / +40%
5.0%
$136k
JPM
+2.1%
0% / +2% / +2%
0% / +5% / +9%
6 mo
+25% / +27% / +27%
+25% / +31% / +37%
0.4%
$136k
MCD
+3.0%
0% / 0% / -1%
-1% / +2% / +5%
14 mo
+25% / +25% / +23%
+24% / +28% / +32%
0.8%
$250k
XOM
+3.0%
0% / +1% / 0%
-1% / +4% / +9%
10 mo
+25% / +26% / +25%
+24% / +30% / +37%
1.2%
$228k
LLY
+3.0%
0% / +1% / -1%
-1% / +5% / +9%
9 mo
+25% / +26% / +24%
+24% / +31% / +36%
5.5%
$127k
ABT
+2.7%
+2% / +3% / +2%
+1% / +6% / +11%
4 mo
+27% / +28% / +27%
+26% / +33% / +39%
0.9%
$132k
CI
+4.3%
0% / 0% / -2%
-1% / +3% / +6%
18 mo
+25% / +25% / +23%
+24% / +29% / +32%
5.5%
$125k
AXP
+2.7%
-1% / +1% / +1%
-1% / +4% / +9%
9 mo
+24% / +27% / +26%
+24% / +30% / +36%
0.9%
$96k
PEP
+2.5%
+1% / +1% / 0%
0% / +3% / +8%
10 mo
+26% / +26% / +25%
+25% / +29% / +35%
0.5%
$162k
Average
+3.6%
0% / +1% / -1%
-1% / +5% / +10%
+25% / +26% / +24%
+24% / +31% / +38%
$1.78M
FIGHT on each: the same engine runs both (fortress_fight handles LEAPS-only rows: no short put, no hedge, Max Loss = the call's cost); the rules are the same: sell above CC-SS, income floor = the larger of hedge rent and half of normal income, banked premium tracked per campaign.
The target. After a 20% drop both need about +25% to break even. The fortress's target holds there; the LEAPS's keeps rising (about +33% after a year, +42% after two), so each month its safe calls sit further out and pay less.
The cushion. In a drop the fortress's hedge put gains and can be rolled down for cash, which carries the lean months. A LEAPS has no hedge, so every FIGHT dollar comes from calls.
Assignment. A deep, long drawdown can bring short-put assignment on the low-volatility names (red cells): you buy the stock at the put's strike, with the hedge put still under it, so Max Loss is unchanged. A LEAPS cannot be assigned.
The income floor barely differs: in HBMax half of normal income already sets it for every name (DELL rent $159 against $1,341 of calls).
Banked premium. The LEAPS's rising CC-SS is decay already paid, and the calls more than cover it. But the tools gate on the pure recovery CC-SS (FIGHT shows the banked figure as information only), so they will still steer a LEAPS to safer, lower-paying strikes.
Appendixthe cash-vs-carry detail, name by name per contract, and choosing
LEAPS pay more cash each month, but once you count the time value they lose, the fortress comes out a little ahead. Same names and long-call strikes (Jan '29, at the ask), short put and hedge put dropped, the same weekly call written on each LEAPS. At the same Max Loss the LEAPS book pays $31,581/mo of cash against the fortress's $23,779, because there is no hedge rent and more contracts fit in the budget. But a deep in-the-money LEAPS still carries time value that melts to nothing by Jan '29 if prices stay flat: about $10,116/mo across the book, while in the fortress the short put's time value pays for the long call's (+$1,407/mo). Counted together at a flat price: $21,465 vs $25,186 a month. Close on day one; what separates them over two years is CC-SS, below.
Name by name, per contract
Strikes are the fortress's long call / short put / hedge put; the LEAPS keeps only the long call. Calls = the same weekly call's book income (model premium x tape keep x INCOME_BOOK trade ratio). Carry = what the structure gains or loses by Jan '29 if the stock ends at today's price, per month: the fortress's long call + short put (the hedge's cost is its rent), or the LEAPS's intrinsic value less its cost.
Name
Fortress strikes
Fortress delta
LEAPS delta
Fortress Max Loss
LEAPS Max Loss (= cost)
Calls / mo
Hedge rent / mo (fortress only)
Fortress carry / mo
LEAPS carry / mo
Contracts HBMax
Contracts LEAPS same ML
DELL
280 / 410 / 230
0.84
0.84
$31,750
$22,300
$1,138
-$148
+$155
-$321
3
4
SPCX
65 / 105 / 90
1.06
0.93
$9,002
$9,130
$552
-$47
+$12
-$56
8
8
MU
430 / 630 / 520
1.05
0.92
$55,560
$54,205
$4,015
-$195
+$43
-$346
1
1
COIN
105 / 150 / 55
1.14
0.82
$11,899
$7,450
$378
-$14
+$11
-$146
8
13
NVDA
145 / 195 / 160
1.06
0.86
$9,870
$9,245
$280
-$81
+$13
-$106
7
7
JPM
280 / 340 / 280
1.13
0.80
$12,019
$9,750
$317
-$61
+$6
-$140
4
5
MCD
200 / 250 / 220
1.14
0.86
$8,964
$8,235
$201
-$37
+$22
-$67
10
11
XOM
150 / 175 / 145
1.03
0.77
$5,924
$4,995
$108
-$53
+$9
-$76
13
15
LLY
900 / 1270 / 1120
1.06
0.89
$65,600
$66,400
$1,872
-$558
+$84
-$586
1
1
ABT
92.5 / 110 / 85
1.16
0.82
$4,215
$3,130
$86
-$18
+$4
-$52
12
16
CI
190 / 250 / 220
1.11
0.89
$13,365
$12,650
$203
-$47
+$8
-$82
5
5
AXP
220 / 320 / 270
1.16
0.90
$15,022
$13,650
$271
-$73
+$14
-$132
3
3
PEP
125 / 135 / 120
1.00
0.72
$3,022
$2,785
$110
-$32
+$4
-$46
12
13
Choosing
Fortress: about $3,721/mo more at a flat price and more upside per dollar of Max Loss (S&P +10%: $377k vs $352k). The price is three legs per name, hedge rolls every few months and a short put that can be assigned. No liquidation by design while every hedge is rolled (Max Loss under the cash).
LEAPS: one leg plus the weekly call, nothing to roll, nothing that can be assigned against you, no margin. At the same contracts the account needs $120k less (Max Loss $746k), which goes to the reserve, and it loses less in a crash (S&P −20%: -$514k vs -$664k). The crash rows hold volatility still, which understates the LEAPS's edge: a long call gains from a volatility spike.
The trap with LEAPS: the cash income looks $7,801/mo higher, but the LEAPS mark down by about that much as time passes. Judge it on the flat-price total, not the cash line.
Model numbers: the 30 Oct pinned quotes; SPCX's long call had no IV in the pin, so it is backed out of the mid (58%). Re-run: python3 leaps_alt.py && python3 gen_page.py.
Method, sources, caveats
Structures and premiums: the INCOME_BOOK's pinned fortress_v28 --live run on the 30 Oct close (v28_pin), its archetypes and keep rates, Jan 2029 LEAPS, rented 77-112 day hedges. HBMax's DELL, SPCX, MU, NVDA and COIN are on the 2 Nov RTH run (10:33 ET, v28_pin/fight_dashboard_hbmax_rth.html); JPM, AXP, MCD, PEP, XOM, LLY, ABT and CI are on after-hours runs of 3 Nov (v28_pin/fight_dashboard_hb2_*.html, re-run in market hours before placing), with MCD and CI premiums cut for earnings and AXP's scaled to the CBOE close (its run's call quote was stale). The rest of the page (close cost, index, crypto) stays on the 30 Oct close. The old book is closed, so its figures are frozen at the 30 Oct legs.
Income bases: book = the INCOME_BOOK's trade-level realized ratio (Sep, since 1 Jun); your rate = Coverage INSIGHTS' 90-day per-fortress income (31 Oct state), as a yield on the window's average close, shrunk to the pooled 0.32 with 10 pseudo-trades, capped at the book ratio.
Correlation and beta: Yahoo daily closes, 180-day covariance (pairwise overlap; SPCX has only ~100 days), 250-day beta to SPY.
Optimizer: integer iterated local search (opt.py), whole contracts; not proven optimal, but warm-started from every other book and stable across reruns.
Crash grid: module_risk_cockpit.scenario with its betas, volatility held still. Today's book from Repair's What-if legs (30 Oct 15:50 ET); the IBKR-matched cockpit row is the 30 Oct 16:30 SGT page.
Friction: CBOE delayed close quotes (Friday 30 Oct, fetched Saturday) against IBKR marks; after-close spreads on thin legs are wider than in RTH.
Nothing here places an order. Re-price every ticket in RTH; IBKR's What-if is the margin check.