Question: BOT (RoboStrategy) primary-source holdings verification — N-2 + N-CSRS confirm the humanoid basket, but a 434% premium to $7.34 NAV is the real signal Verdict: open

What we're asking

The 2026-05-26 BOT coverage decision added RoboStrategy to humanoid-robotics.json as a basket-platform layer, but flagged its claimed private holdings — Figure AI, Apptronik, Dyna, Path Robotics, CoCo Robotics, GMI Cloud — as unverified from primary docs. Those names came from a Discord conviction doc (Frank, 2026-05-16), not the filings. This is the gating primary-source step: do the SEC filings actually show that basket, and what are the NAV, expense ratio, and premium/discount?

Primary sources read (via SEC EDGAR, CIK 0002081119):

• N-CSRS 0001213900-26-053808 (filed 2026-05-08, portfolio as of 2026-02-28) — ea0288089-01_ncsrs.htm
• N-2 0001213900-26-055474 (filed 2026-05-13, registration statement) — ea0290107-01_n2.htm

What we found

Holdings — the claimed basket is real. All six conviction-doc names are confirmed in primary docs:

The N-2 (fresher, ~Mar-31) names additional portfolio companies beyond the six — Dexmate, REK Inc., plus SPV wrappers (e.g. DDGR LLC) — so the doc-claimed "6 + more" holds. No marketing/basket mismatch.

Financials (N-CSRS, period ending 2026-02-28):

• NAV/share $7.34 (up from $5.29 beginning-of-period); NAV total return −1.87%
• Net assets $146.2M; 92.19% invested (cost $134.8M), essentially all in private preferreds/SPVs
• Sector: Robotics ~90.8% (50.81% preferred + 40.01% SPV), Artificial Intelligence 1.37%, nominal Berkshire Hathaway common (0.00%)
• Realized expense ratio 1.96% (both including and excluding deferred-tax) for the period

Financials (N-2 prospectus): Management Fee 2.50%, Total Annual Expenses 3.22% (the 3.22% the conviction doc cited is the prospectus's stated annual figure; 1.96% is the realized semi-annual ratio — both correct in their context).

The disqualifying signal — premium, not discount. The N-2 states plainly: the market price represented a 434% premium to NAV per share as of March 31, 2026. This inverts the original "buy the CEF at a discount = buy the basket at a discount" arbitrage thesis. You are not buying $1 of private-robotics basket for $0.80; you are paying ~$5.30 for it.

Fresh tape (2026-06-30, research/market-engine/data/summaries/humanoid-robotics.json): BOT $39.75 — up ~54% from the stale $25.88 (2026-05-26) the prior caveat carried, and against the $7.34 NAV (N-CSRS Feb-28) that is a ~442% premium, wider than the N-2's March 434%. The premium is expanding, not compressing — the market is bidding the wrapper faster than the underlying private marks move. That is the entire risk in one line: a premium-mean-reversion event (a NAV re-mark, a lock-up unlock, or sentiment cooling on the private-humanoid names) hits the $39.75 wrapper far harder than the $7.34 basket.

Verdict + reasoning

Holdings: VERIFIED. All six claimed names confirmed against primary SEC filings (five directly in the Feb-28 N-CSRS; CoCo/Cyan and additional names in the May-13 N-2). The conviction doc's basket, NAV ($7.34), and premium claims all check out — a credibility data point for Frank as a candidate source (profile already built 2026-06-17, status candidate; this corroborates it — noted in the follow-up below, not a full re-eval).

Coverage: stays research_only / watch-only in humanoid-robotics. The verification removes the holdings caveat but hardens the investment caveat: a 434% NAV premium + 3.22% total annual expenses + wholly-illiquid private-preferred underlying = a poor vehicle, not an entry. It's the only public wrapper we have for the pre-IPO humanoid names (Figure, Apptronik, Dyna, Path, CoCo), which is genuine coverage value — but as a watch surface for the private-humanoid theme, not a buy. It fails the trend-hold screen decisively (premium mean-reversion risk dwarfs any basket appreciation; this is a lottery-adjacent structure the book avoids). The discount-arbitrage angle from the original decision is dead — reframe any future note around premium compression risk, not discount capture.

Follow-ups:

• Update humanoid-robotics.json caveat: holdings verified (N-2 + N-CSRS 2026-02-28), CoCo = Cyan Robotics; add the 434% premium / $7.34 NAV / 3.22% expense flag as the operative risk. (done in this pass)
• Watch the newer N-2/A 0001213900-26-068988 (filed 2026-06-16) for updated NAV / premium / holdings if the theme heats up — the premium is the number to re-read, not the basket.
• Frank source: holdings claims verified against primary docs → confidence-supporting; leave as candidate, revisit if a second substantive doc lands.