Portfolio Tracker: What They Hold, and What Changed

Auto-generated 2026-09-21 12:52 (prices as of close 2026-09-21). Tracks each trader's disclosed portfolio: positions they still hold are marked to market, positions they exited stop being priced and move to a history table, and every new full-book disclosure is diffed against the previous one to catch adds and exits. Options legs, premium income, and undisclosed positions remain invisible to us. Not an audit. Dossiers: TJ · Papa G · Comparison.

Papa G / 10-Minute Trader

His own stated cost bases, read aloud position by position. Only one full-book walkthrough exists (2026-03-31), so his book cannot be diffed the way TJ's can; later entries are individual disclosures, not a complete book. Week of 2026-09-12: still no second full-book walkthrough, so still no diff. But three names he had never disclosed before turned up in his Sep 9 and Sep 10 streams, the first additions since March: LRCX at a stated $393 basis and down over 20%, NFLX at a stated $82 basis held since roughly February, and ORCL, which he owns but has never priced, so it sits in the not-priceable list. All three are his own book, not viewer questions, checked against his five-questions ritual. Carried over from last week and still open: he said on 09-01 that his Kohl's shares were called away at $18, which confirms the 08-26 buy-write was assigned, but he never said whether the original 4,000-share lot from 03-31 is still there. KSS is left open at 4,000 rather than guessing.

Week of 2026-09-19: still no second full-book walkthrough, so still no diff. Checked all 12 streams posted 09-15 through 09-21 (auto-sub transcripts pulled and grepped for position language). Nothing new to log. Two things worth naming: (1) he restated the QCOM loss again, near-verbatim to the 08-31 admission: "Uhoh, I lost money on QCOM. I did. Very rarely do I lose money, but I did on QCOM. Just got away from me. We're going to say no to QCOM this week" (stream txJoXxVKs6M) - same admitted loss, no new number, not logged as a separate event. (2) Several streams repeat his LRCX-393 and generic dividend-yield examples ('If I buy 1,000 shares at 560...', 'If I buy Kohl's at this price, I own it...') almost word for word; these are explicitly hypothetical teaching examples ('let's say hypothetically...', 'if I buy...'), not new trades, and were not logged as positions. ORCL is mentioned in nearly every stream (mostly viewers asking about their own ORCL, not his) but he still never states his own cost basis, so it stays unpriceable.

Only one full-book disclosure on file (2026-03-31), so no add/exit diff is possible yet. A second full snapshot enables this. Partial, single-name disclosures since then are folded into the holdings table but cannot reveal an exit.

Currently held (8)

Held but not priceable: ORCL (held, but his cost basis was never stated, so it cannot be marked. He owns it and is short the $145 calls against it: "I have had this for a while. It went up, it went down, and when it came down, I sold a strike price below my cost basis... I am in at the 145s. Well, I am in for MORE than that, but I sold the 145, so I have to buy that 145 back this week, and I kind of do not want to pay for it." That fixes his basis only as ABOVE $145, with no upper bound, which is why there is no ref price. Sources: streams xi_EXftHKBw (Sep 9 session, ORCL at $161 to $163) and bMYYzfoBBRY (Sep 10 session, ORCL at $152, short calls in the money, "I want this baby to keep coming down"). ORCL closed $150.28 on 2026-09-11. A stated basis would make this priceable, so it is worth listening for.)

Short-term buy-writes, not tracked as holdings: PEP @ $143.00 (2026-07-30), NTR @ $71.00 (2026-07-30)

ˢ Share-count notes: LRCX, Share count NOT reliably disclosed. He hedges it himself: "What do I have like 500 shares? 500 shares times three". The only hard number is the 3 covered calls he sold, which implies at least 300 shares. Left null rather than inventing 500 from a sentence he was unsure of. | NFLX, Share count never stated.

TJ WheelDeal

Baselines are the price shown when he last disclosed each position. Share counts only where he stated them. Measures movement since disclosure, not his profitability, his P&L is dominated by option legs we cannot see. Week of 2026-09-12: a fresh tastytrade Positions screenshot landed on 09-11, the first full book since 08-28. It confirms the two exits already logged from his own statements (HOOD, NBIS) and adds nothing new, so the book is unchanged on the week. It also settles the BMNR question the other way: BMNR is still on the book, as 1,500 Jan-2028 $25 calls with zero shares, so the stale flag is gone and it is carried as held at 0 shares. Every ref price was re-baselined to the 09-11 screenshot, which is why the Change column reads near zero: the baseline and the price are one day apart. The dollar column restarts from this disclosure, it is not a loss. One correction of his arithmetic: his 09-11 Northstar Bag slide prices PLTR at $187.21 when PLTR closed $167.23, overstating that slide's $28.4M bag by about $899,100; the refs here use his Positions screenshot instead, which is right.

Week of 2026-09-19: a second fresh full-book Positions screenshot landed 09-19 (X post 2101365132739625239), 8 days after 09-11. The book changed: NBIS reappears, nothing else added or dropped, so the 9 names carried since 08-28 are unchanged and NBIS is a straight re-add. He closed the original NBIS put 09-01 (booked +$25K) and reopened a bigger one, -800 Jun-16(2027) $100 strike puts, cost $1.79M, by 09-18, confirmed by a matching single-name crop the night before the full book (same $105,330 P/L Open in both). NBIS is carried 'open' again at 0 shares, the same convention already used for BMNR: no shares, but a live option book. That same 09-18 post (a 4-image 'Backstop Premium Cash Grab' teaching thread using SPCX/NBIS/RIOT/CLSK as live examples) also showed his RIOT and CLSK backstops rolled up in strike and size (both now -5,000 contracts, Jan-21 2027 expiry, $15 and $12 strikes respectively), still zero shares either way. All ref prices below are re-baselined to the 09-19 screenshot's Ask column; it has no share-count or avg-cost columns, so share counts stand at their 09-11 levels. SEPARATELY, on 09-21 (today) he posted a screenshot from a new tool, "TW Live Book", labelled "Account 1 of 10" for "60 - Family Office": Net Liquidity $19,319,118, P/L Day $927,600, P/L YTD (Last Close) $1,540,678, for THAT ONE SUB-ACCOUNT ONLY, with its own 4-name, 400,000-share Northstar Bag table (SOFI 250,000, ENPH 100,000, PLTR 40,000, MU 10,000; no MSTR or SPCX). This is explicitly one of roughly 10-11 accounts (his X bio says he "manage[s] our family office that consists of 11 accounts in all"), not his aggregate book, so it is logged as a partial snapshot and NOT diffed against the full-book Positions screenshots or the self_reported aggregate net-liquidity trend: doing either would read as a $3.7M net-liq jump or an MSTR/SPCX exit that never happened. Worth watching, since it is the first time he has shown a per-account breakdown; if he posts more of these for other accounts they could eventually explain how the aggregate book is built up.

PORTFOLIOPortfolio change: 2026-09-11 → 2026-09-19

• 🟢 Added (1): NBIS
• Held through both: BMNR, CLSK, ENPH, MSTR, MU, PLTR, RIOT, SOFI, SPCX

Currently held (7)

Held but not priceable: SPCX (Private SpaceX exposure, no public quote. Held but cannot be marked. 20,000 shares as of the 09-11 Northstar Bag, unchanged since 08-24. His own mark rose from $142.24 on 09-02 to $151.14 on 09-11 to $152.65 on 09-19; all three are his own broker marks, not traded prices. P/L YTD on his 09-11 Positions screenshot reads +$539,602.81; on the 09-19 full-book screenshot it reads +$631,280.92 (P/L Opn $62,055.00). The same-day 4-image "Backstop Premium Cash Grab" teaching post (X post 2101159470835314823) also shows the SPCX option backstop itself: -250 Nov-20 $110 puts and -1,500 Sep-17 $100 puts, combined cost $1,256,555.00, combined P/L Open $61,835.00 (close to the $62,055.00 P/L Opn on the full-book row, small gap likely a snapshot-timing difference).)

Exited, no longer tracked (4)

ˢ Share-count notes: PLTR, Share count and avg cost from the 09-11 "MANUFACTURE THE WIN" Northstar Bag (X post 2098513190594228659, slide 3/3), which totals 416,000 shares across 6 names. Ref price is taken from the tastytrade Positions screenshot posted 9 minutes later (X post 2098515365173030957), not from the Bag slide, because the Bag slide misprices PLTR. PLTR 42,500 -> 45,000 at an avg cost of $133.46. ARITHMETIC DOES NOT CLOSE: his 09-11 moves table covers only Sep 9 to Sep 11 and lists no PLTR buy, so the 2,500-share add happened in the undisclosed Sep 3 to Sep 8 window. PRICE DISCREPANCY, HIS ERROR, NOT OURS: the Bag slide prints PLTR "PRICE TODAY $187.21" and values the stake at $8,424,450 (45,000 x 187.21). PLTR actually closed $167.23 on 2026-09-11 (yfinance) and his own Positions screenshot from the same afternoon shows Last 167.21 / Ask 167.61. The Bag therefore overstates PLTR by about $899,100 and the "VALUE TODAY $28.4M" headline by the same amount; the real bag is about $27.5M. Every other name on the Bag slide matches the Positions screenshot to the cent. | SOFI, Share count and avg cost from the 09-11 "MANUFACTURE THE WIN" Northstar Bag (X post 2098513190594228659, slide 3/3), which totals 416,000 shares across 6 names. Ref price is taken from the tastytrade Positions screenshot posted 9 minutes later (X post 2098515365173030957), not from the Bag slide, because the Bag slide misprices PLTR. SOFI unchanged at 250,000, avg cost $18.56. He calls it "the biggest disappointment" of the book on 09-11; it is $1.23 a share underwater against his own basis and its P/L YTD on the Positions screenshot is -$220,108.58. | MSTR, Share count and avg cost from the 09-11 "MANUFACTURE THE WIN" Northstar Bag (X post 2098513190594228659, slide 3/3), which totals 416,000 shares across 6 names. Ref price is taken from the tastytrade Positions screenshot posted 9 minutes later (X post 2098515365173030957), not from the Bag slide, because the Bag slide misprices PLTR. MSTR unchanged at 35,000, avg cost $105.29. | MU, Share count and avg cost from the 09-11 "MANUFACTURE THE WIN" Northstar Bag (X post 2098513190594228659, slide 3/3), which totals 416,000 shares across 6 names. Ref price is taken from the tastytrade Positions screenshot posted 9 minutes later (X post 2098515365173030957), not from the Bag slide, because the Bag slide misprices PLTR. MU 5,500 -> 6,000 at an avg cost of $947.34. Arithmetic closes: his 09-11 moves table lists "Bought MU shares (Fri) 500 @ ~$976 (now 6,000)". | ENPH, Share count and avg cost from the 09-11 "MANUFACTURE THE WIN" Northstar Bag (X post 2098513190594228659, slide 3/3), which totals 416,000 shares across 6 names. Ref price is taken from the tastytrade Positions screenshot posted 9 minutes later (X post 2098515365173030957), not from the Bag slide, because the Bag slide misprices PLTR. ENPH 50,000 -> 60,000 at an avg cost of $39.25. ARITHMETIC DOES NOT CLOSE: the 09-11 moves table covers Sep 9 to Sep 11 only and lists +2,500 Tue @ ~$36.40 and +2,500 Fri @ ~$37.12, which is 55,000 from a 50,000 base. The missing 5,000 shares were bought in the undisclosed Sep 3 to Sep 8 window. Screenshot count outranks buy-post arithmetic, so 60,000 stands. He calls ENPH the deliberate build: "we would actually WELCOME a rung getting called away", coverage on two rungs. | SPCX, Share count and avg cost from the 09-11 "MANUFACTURE THE WIN" Northstar Bag (X post 2098513190594228659, slide 3/3), which totals 416,000 shares across 6 names. Ref price is taken from the tastytrade Positions screenshot posted 9 minutes later (X post 2098515365173030957), not from the Bag slide, because the Bag slide misprices PLTR. SPCX unchanged at 20,000, avg cost $127.96. He marks it at $151.14, his own mark, not a traded price; he calls it "the biggest surprise" of the year.

What he says the account is worth

His own published account figures, logged verbatim each time he posts one. This is the only account-level total that exists for him: our table above measures share legs only, his number is the whole account including every option leg and all premium. They are not two versions of the same measurement and should never be netted against each other.

A trailing ~ on a net-liquidity figure means it was read off a chart label rather than published as a number.

Note on the 2026-09-11 entry: Slide labelled "Weekly recap, week ending Friday, September 11, 2026"; Sep 11 2026 really is a Friday, so the label checks out (unlike his 08-22 slide). Headline "Loaded ENPH & MU all week, now positioned to cover-and-deliver": net liq $15,609,571, P/L on Friday +$495,851, YTD +$741,967, 416,000 shares owned, theta +$46,138/day which he annotates "= $1,384,140/month if it holds". Buying power used not stated. BIGGEST YTD ON FILE AND THE FIFTH ZERO-CROSSING: -$1,362,117 (08-21) -> +$459,164 (08-27) -> +$259,051 intraday (09-01) -> -$296,557 at the 09-01 close -> -$272,395 (09-02) -> +$741,967 (09-11), a +$1,014,362 swing in nine days. Net liq is also a fresh high, +$1.01M on the 09-02 read and +126% off the $6.92M Jun 24 low. The YTD figure still cannot be reconciled without a Transactions tab; see dossier E1b.

⚠️ His YTD number and his broker's YTD column disagree, and the gap is unexplained. The per-underlying P/L YTD column in his 08-17 Positions screenshot sums to roughly +$882,000; the graphic dated four days later reports -$1,362,117 for the same year. Both are his. Neither should be treated as the account's true 2026 result until he posts a Transactions tab. See TJ dossier E1b and E1d.

Watchlist: floated on X, not in the book (2)

Names he talked up publicly but has never been seen holding. Priced from the day he floated it, to test one claim: that his commentary actually leads his trades.

Trend across runs

Full history: tracker_history.csv. A change in the # held column means the portfolio itself changed that week.

⚠️ Do not read these rows as a running P&L. Each row measures movement from that run's reference prices, and the references are reset every time a fresh full-book disclosure arrives. A run right after a new screenshot therefore starts from roughly zero and shows only a few days of movement, which is not comparable to the prior run's figure. The # held column is the column that carries week-over-week meaning.

Reading this honestly

• This tracks portfolios, not tickers. A name is priced only while they still hold it. Once a disclosure shows it gone, it moves to the exited table and stops being marked.
• Exits are inferred from disclosures, so they lag. We only learn a position closed when the next full-book screenshot omits it, which is why the weekly run hunts for new screenshots.
• Papa G's rows measure his actual claim, that drawdowns don't matter because premium keeps arriving. This is what the buy-and-hold leg is worth against his own stated bases.
• TJ's rows measure movement since disclosure, not profitability.
• Neither includes premium collected, the income side of both strategies. Share legs only. See TJ dossier E1b for why even his own screenshots can't separate the two.
• Watchlist rows are not positions. They are names he floated publicly with no share count or trade disclosed. They exist to test one claim: that his X commentary leads his actual trades.
• Share counts marked ˢ carry a sourcing note, printed above the table. Some come from a screenshot he posted, some were adjusted from his dated daily buy posts. Screenshot counts are the more reliable of the two and override buy-post arithmetic when they disagree.
• Update: python3 tracker.py && bash deploy.sh. Runs weekly, see WEEKLY_TRACKER_RUN.md.