How much private AI do you actually own?
You cannot buy OpenAI. What you can buy is a company that owns part of it, or a fund that does. Every list of those tickers you will find elsewhere stops there — it names the route and never says how much of it you get.
This page puts a number on each one: dollars of the private company per $10,000 invested, computed from the filing that discloses the stake, with the sentence it came from quoted underneath. Some of those numbers are large. Most are far smaller than the headline suggests, and two of them are not ownership at all.
Where the private-company prices come from. The per-share marks this page compares against are what regulated funds carry these companies at, taken line by line from SEC Form N-PORT. See all 46 of them, and how wrong those marks turned out to be for the ten companies that have since listed. The per-$10,000 arithmetic below, and what a fund's premium or discount to its own net asset value does and does not tell you, are worked through in Lesson 4 of the free Pre-IPO & IPO course.
Public companies that own a piece
The look-through column answers one question: buy $10,000 of this stock, and how many dollars of private-company stake come with it? It is the disclosed figure divided by the company's market cap, times $10,000 — so it assumes the market pays nothing extra for the stake, which is the conservative direction to be wrong in.
It is only computed for the two kinds of disclosure that represent money in the ground. A guarantee, a promise and a percentage with no dollar figure attached get no number, because giving them one would be arithmetic dressed up as a holding.
Salesforce (CRM)
$175.8B market cap · $213.65 on 2026-11-06The balance-sheet carrying value, remeasured to observable prices. The closest thing to a look-through valuation any of these filings contains.
The only single-company Anthropic carrying value in this file. Alphabet never names an investee, and Amazon's number is a combined Anthropic-and-OpenAI line; Salesforce states a dollar figure for Anthropic alone. The same filing elsewhere puts it at roughly 45 percent of the portfolio, up from 22 percent six months earlier — the position did not grow, the mark did.
“Strategic Investments As of July 31, 2026, our strategic investment portfolio consisted of investments in over 450 companies with a combined carrying value of $11.3 billion, including the Company's investment in Anthropic PBC ("Anthropic") which represented approximately $5.1 billion of the total strategic investments portfolio.”
The balance-sheet carrying value, remeasured to observable prices. The closest thing to a look-through valuation any of these filings contains.
The whole strategic-investment portfolio, and it CONTAINS the $5.1 billion Anthropic line above rather than sitting alongside it. Do not add the two. Spread across more than 450 companies, so the non-Anthropic half is diversified to the point of being a different asset class.
“Strategic Investments As of July 31, 2026, our strategic investment portfolio consisted of investments in over 450 companies with a combined carrying value of $11.3 billion, including the Company's investment in Anthropic PBC ("Anthropic") which represented approximately $5.1 billion of the total strategic investments portfolio.”
Amazon (AMZN)
$2.80T market cap · $259.3 on 2026-11-06The balance-sheet carrying value, remeasured to observable prices. The closest thing to a look-through valuation any of these filings contains.
The cleanest look-through number any of these companies discloses: a carrying value, remeasured to observable prices, in a sentence that names the issuers. It is a combined line, so it cannot be split between Anthropic and OpenAI without guessing — and we do not guess.
“As of December 31, 2025 and June 30, 2026, equity investments in private companies not accounted for under the equity-method, which primarily relate to nonvoting preferred stock in Anthropic and preferred stock in OpenAI, had a carrying value of $16.2 billion and $122.3 billion, with adjustments for observable changes in prices or impairments representing Level 3 fair value measurements recognized in "Other income (expense), net" on our consolidated statements of operations.”
What was actually paid in. A floor on what the stake is worth, never the value of it.
The original convertible-note position, since converted to nonvoting preferred. Shown because the gap between this and the carrying value above is the entire return on the trade.
“Non-Marketable Investments Anthropic - From Q3 2023 to Q4 2025, we invested $8.0 billion in convertible notes from Anthropic, which are classified as available-for-sale and reported at fair value with unrealized gains and losses included in "Accumulated other comprehensive income (loss)" and as Level 3 assets.”
A guarantee or a credit facility. Downside without ownership — the opposite shape of what someone buying the proxy for upside expects.
A facility that expires 30 months after an Anthropic liquidity event, so it is also a dated statement that Amazon expects one.
“Under this financing arrangement, in Q2 2026, we exercised our option to participate in subsequent Anthropic equity financings by investing $5.0 billion in Anthropic Series H nonvoting preferred stock, which reduced the amount available under the facility to $15.0 billion.”
Alphabet (GOOGL)
$3.50T market cap · $284.78 on 2026-11-06The balance-sheet carrying value, remeasured to observable prices. The closest thing to a look-through valuation any of these filings contains.
Alphabet never names a private investee in this filing. Search it for 'Anthropic' and the only hits are the word 'philanthropic'. So the widely repeated GOOGL-as-an-Anthropic-proxy trade cannot be sized from Alphabet's own disclosure, and we will not size it from a press report.
“As of June 30, 2026, the carrying value of our non-marketable equity securities accounted for under the measurement alternative was $124.3 billion, of which $87.9 billion was remeasured at fair value during the three months ended June 30, 2026 and was primarily classified within Level 2 of the fair value hierarchy at the time of measurement.”
Money promised and not yet paid, usually contingent on milestones. Exposure to a future decision, not to a current stake.
Milestone-contingent through 2030 and accounted for as an equity derivative. Not money spent, and not attributable to a named company.
“As of June 30, 2026, this amount includes $20.0 billion of future capital funding commitments with a private company contingent upon the achievement of specified operational and 24 financial milestones through 2030, which is accounted for as an equity derivative.”
A stake that has already listed. Included because a proxy that stops being a proxy is the most common way this kind of trade ends.
Sits in marketable equity securities now, with a further $14.1 billion restricted through the third quarter of 2027. This is what a private-exposure proxy looks like after the proxy stops being a proxy.
“… ts: Marketable equity securities (2) 14,126 0 14,126 Total $114,695 $129,082 $256,600 (1) Includes $80.0 billion of Space Exploration Technologies Corp.”
Coinbase (COIN)
$35.7B market cap · $134.76 on 2026-11-06The balance-sheet carrying value, remeasured to observable prices. The closest thing to a look-through valuation any of these filings contains.
Pinned to the annual report because the quarterlies restate this only inside a fair-value table with no stated units. The filing adds that these are positions under 20 percent with no significant influence, which is as close as any issuer here comes to saying outright that it is a passive minority book.
“As of December 31, 2025 and 2024 , our strategic equity investments in privately held companies were $623.0 million and $374.2 million , respectively.”
Intel (INTC)
$299.1B market cap · $56.28 on 2026-11-06The balance-sheet carrying value, remeasured to observable prices. The closest thing to a look-through valuation any of these filings contains.
A named, dollar-denominated stake in a private company that has nothing to do with AI model labs, which is why it is here: it is the counterexample to the assumption that private exposure inside a public company means exposure to the names in the headlines. Equity-method, so the carrying value moves with Altera's results rather than with a funding round.
“The $3.2 billion value of our non-marketable equity method investment in Altera is classified within equity investments in the Consolidated Balance Sheets at December 27, 2025, and was recognized as a non-cash investing activity within the 2025 Consolidated Statements of Cash Flows .”
Oracle (ORCL)
$362.0B market cap · $116.1 on 2026-11-06The balance-sheet carrying value, remeasured to observable prices. The closest thing to a look-through valuation any of these filings contains.
Included to settle the question rather than to support the trade. Oracle is widely called an OpenAI proxy on the strength of its cloud contracts, but a contract is revenue, not ownership: the entire non-marketable investment book is $2.4 billion, it names nobody, and it mixes equity with debt.
“Non-Marketable Investments Our non-marketable equity securities and debt investments totaled $2.4 billion and $2.3 billion as of August 31, 2026 and May 31, 2026, respectively, and substantially all of the balance is included in other non-current assets in the accompanying condensed consolidated balance sheets and is subject to periodic credit losses and impairment reviews.”
Microsoft (MSFT)
$3.94T market cap · $530.61 on 2026-11-06A disclosed percentage interest with no dollar figure attached, so it cannot be converted into a per-$10,000 number.
The largest single disclosed claim on OpenAI in any public company's filings, and it comes with no dollar figure attached. Microsoft accounts for it under HLBV, so the balance-sheet carrying amount tracks OpenAI's book value rather than its market value, and cannot be used as a look-through price.
“We have an investment accounted for under the equity method that represents an approximate 25 % interest on an as-converted basis.”
What was actually paid in. A floor on what the stake is worth, never the value of it.
Cash funded, not value. Treat it as the floor: whatever the 25% interest is worth, Microsoft paid this for it.
“We have made total funding commitments of $13.0 billion related to our investment, of which $11.9 billion has been funded as of June 30, 2026.”
NVIDIA (NVDA)
$4.51T market cap · $186.64 on 2026-11-06A guarantee or a credit facility. Downside without ownership — the opposite shape of what someone buying the proxy for upside expects.
Not a stake. NVIDIA's disclosed OpenAI exposure in this filing is a lease guarantee cap — downside without ownership, which is the opposite shape of what someone buying NVDA for OpenAI exposure thinks they are getting.
“The campus will exclusively host our compute under 20-year leases to OpenAI, subject to limited exceptions, with our obligation capped at $105 billion in the aggregate, subject to certain conditions including SB Energy, the lessor, satisfying applicable ready-for-service conditions.”
Cisco (CSCO)
$370.0B market cap · $93.85 on 2026-11-06Money promised and not yet paid, usually contingent on milestones. Exposure to a future decision, not to a current stake.
Money promised to funds and companies Cisco does not name, not money spent. Cisco's carrying value is disclosed only as table rows without units, so the commitment is the one figure here that survives being quoted on its own.
“We have total funding commitments of $0.4 billion related to non-marketable equity securities.”
Funds you can buy that hold them directly
These are registered funds holding private companies we price from at least three independent funds. They are not chosen by us: the builder scans every N-PORT already cached for the marks page, keeps the registrants that have an exchange ticker, and publishes each one that clears the bar. Hand-picking them would let us quietly drop a fund whose numbers embarrassed us.
The premium column is the part most write-ups skip, and it is the part that decides whether the fund is a route or a trap. 2 of the 10 funds below have a NAV we can source from the filer's own tagged disclosure; the rest show why not, rather than a figure we made up.
| Fund | Covered names | Of net assets | Per $10,000 | Price | NAV | Premium / discount | Holdings as of |
|---|---|---|---|---|---|---|---|
| RVI Robinhood Ventures Fund I | 5 | 39.01% | $3,901 | $21.35 | — | the fund has not tagged a net asset value per share in XBRL. Several of these funds report NAV only inside a combined shareholder report covering three trusts at once, and picking the wrong column would publish another fund’s NAV as this one’s — so the figure is left to the reader and the filing is linked | 2026-06-30 |
| ARKVX ARK Venture Fund | 13 | 31.55% | $3,155 | — | — | not applicable — the fund is bought and sold at NAV, not on an exchange | 2026-07-31 |
| BSTZ BlackRock Science & Technology Term Trust | 8 | 30.86% | $3,086 | $24.30 | — | the fund has not tagged a net asset value per share in XBRL. Several of these funds report NAV only inside a combined shareholder report covering three trusts at once, and picking the wrong column would publish another fund’s NAV as this one’s — so the figure is left to the reader and the filing is linked | 2026-06-30 |
| BST BlackRock Science & Technology Trust | 12 | 21.22% | $2,122 | $43.45 | $54.57 2026-06-30 | the last filed NAV is 129 days older than the price, past the 120-day limit — the difference would measure the gap between the two dates more than any premium | 2026-06-30 |
| DXYZ Destiny Tech100 Inc. | 8 | 19.74% | $1,974 | $24.37 | $34.30 2026-06-30 | the last filed NAV is 129 days older than the price, past the 120-day limit — the difference would measure the gap between the two dates more than any premium | 2026-06-30 |
| BTX BlackRock Technology & Private Equity Term Trust | 7 | 17.67% | $1,767 | $7.56 | — | the fund has not tagged a net asset value per share in XBRL. Several of these funds report NAV only inside a combined shareholder report covering three trusts at once, and picking the wrong column would publish another fund’s NAV as this one’s — so the figure is left to the reader and the filing is linked | 2026-06-30 |
| BOT RoboStrategy, Inc. | 1 | 13.65% | $1,365 | $21.62 | $11.35 2026-09-30 | 90.5% premium | 2026-08-31 |
| VCX FUNDRISE INNOVATION FUND, LLC | 6 | 8.74% | $874 | $20.05 | $38.12 2026-09-08 | 47.4% discount | 2026-06-30 |
| PWRL Powerlaw Corp. | 2 | 3.66% | $366 | $11.06 | $13.14 2026-03-31 | the last filed NAV is 220 days older than the price, past the 120-day limit — the difference would measure the gap between the two dates more than any premium | 2026-06-30 |
| BCAT BlackRock Capital Allocation Term Trust | 7 | 2.34% | $234 | $13.53 | $14.90 2026-06-30 | the last filed NAV is 129 days older than the price, past the 120-day limit — the difference would measure the gap between the two dates more than any premium | 2026-06-30 |
What each fund is actually holding
A position reached through a single-purpose vehicle is unwrapped to the company underneath — DXYZ's largest private holding is filed as "Magnitude ANC III, LLC (economic exposure to Anthropic PBC Series B Preferred Shares)", and filing it under the LLC would hide the only fact that matters about it. Such a position gets no per-share price: its reported share balance counts units of the vehicle, not shares of the company.
The cross-fund median in the last two columns includes the fund's own mark, because excluding it would leave several companies with too few opinions to have a median at all. So a fund that lands exactly on it is labelled as setting it, not as agreeing with it.
| Company | Value held | Per $10,000 | Fund carries at | Cross-fund median | Difference |
|---|---|---|---|---|---|
| Databricks | $86.6M | $1,271 | $201.54 | $253.00 | -20.3% |
| OpenAI | $75.0M | $1,100 | $687.69 | $687.69 | sets the median |
| Revolut Ltd held through a vehicle, unwrapped — no per-share price | $49.8M | $731 | $1,398 | $1,929 | -27.5% |
| Stripe Global Holdings Inc | $29.5M | $433 | $63.00 | $65.98 | -4.5% |
| Canva | $25.0M | $367 | $1,646 | $1,496 | +10.0% |
| Company | Value held | Per $10,000 | Fund carries at | Cross-fund median | Difference |
|---|---|---|---|---|---|
| OpenAI | $67.6M | $571 | $678.11 | $687.69 | -1.4% |
| Stripe Global Holdings Inc | $53.4M | $451 | $70.82 | $65.98 | +7.3% |
| Anthropic | $49.6M | $419 | $556.98 | $777.53 | -28.4% |
| Tenstorrent Holdings Inc | $41.2M | $348 | $85.67 | $86.43 | -0.9% |
| Figure AI Inc | $30.3M | $255 | $160.00 | $194.86 | -17.9% |
| Crusoe Inc | $26.9M | $227 | $226.20 | $205.97 | +9.8% |
| Databricks | $26.4M | $223 | $262.31 | $253.00 | +3.7% |
| Freenome Inc | $22.5M | $190 | $11.15 | $2.96 | different series |
| Zipline International Inc | $20.6M | $174 | $63.01 | $72.11 | -12.6% |
| Epic Games | $13.4M | $113 | $400.00 | $377.90 | +5.8% |
| Apptronik | $10.2M | $86.30 | $36.80 | $36.86 | -0.1% |
| Revolut Ltd | $5.8M | $49.19 | $2,017 | $1,929 | +4.6% |
| Discord Inc | $5.6M | $47.31 | $26.51 | $23.51 | +12.8% |
| Company | Value held | Per $10,000 | Fund carries at | Cross-fund median | Difference |
|---|---|---|---|---|---|
| Databricks | $321.2M | $1,350 | $203.46 | $253.00 | -19.6% |
| Psiquantum | $125.5M | $528 | $46.11 | $46.52 | -0.9% |
| Anthropic | $96.9M | $407 | $589.01 | $777.53 | -24.2% |
| Sambanova Systems | $88.0M | $370 | $123.86 | $127.10 | -2.6% |
| GrubMarket Inc | $47.5M | $199 | $26.92 | $26.92 | sets the median |
| ByteDance Ltd | $32.6M | $137 | $334.24 | $352.46 | -5.2% |
| Astranis Space Technologies Corp | $20.2M | $84.75 | $26.01 | $19.23 | +35.2% |
| DataRobot | $2.7M | $11.32 | $2.93 | $1.91 | +53.5% |
| Company | Value held | Per $10,000 | Fund carries at | Cross-fund median | Difference |
|---|---|---|---|---|---|
| Anthropic | $96.9M | $513 | $589.01 | $777.53 | -24.2% |
| Databricks | $85.6M | $453 | $203.46 | $253.00 | -19.6% |
| Anduril | $73.1M | $387 | $68.95 | $73.50 | -6.2% |
| Psiquantum | $40.4M | $214 | $45.91 | $46.52 | -1.3% |
| Sambanova Systems | $33.5M | $177 | $120.07 | $127.10 | -5.5% |
| OpenAI | $30.0M | $159 | $687.69 | $687.69 | sets the median |
| GrubMarket Inc | $19.1M | $101 | $26.92 | $26.92 | sets the median |
| Canva | $10.7M | $56.84 | $1,145 | $1,496 | -23.5% |
| ByteDance Ltd | $5.3M | $28.07 | $334.24 | $352.46 | -5.2% |
| Discord Inc | $3.0M | $16.13 | $239.61 | $23.51 | different series |
| Patreon | $2.4M | $12.46 | $13.18 | $12.16 | +8.3% |
| DataRobot | $715,128 | $3.79 | $2.89 | $1.91 | +51.6% |
| Company | Value held | Per $10,000 | Fund carries at | Cross-fund median | Difference |
|---|---|---|---|---|---|
| Anthropic held through a vehicle, unwrapped — no per-share price | $235.7M | $1,442 | — | $777.53 | — |
| OpenAI held through a vehicle, unwrapped — no per-share price | $35.0M | $214 | — | $687.69 | — |
| Databricks held through a vehicle, unwrapped — no per-share price | $19.6M | $120 | — | $253.00 | — |
| Revolut Ltd | $16.6M | $102 | $2,028 | $1,929 | +5.1% |
| Astranis Space Technologies Corp part direct, part through a vehicle | $7.5M | $45.88 | $19.23 | $19.23 | sets the median |
| Redwood Materials | $4.2M | $25.72 | $36.89 | $43.24 | -14.7% |
| Stripe LLC held through a vehicle, unwrapped — no per-share price | $3.3M | $20.27 | — | $70.01 | — |
| Discord Inc | $788,018 | $4.82 | $33.11 | $23.51 | +40.8% |
| Company | Value held | Per $10,000 | Fund carries at | Cross-fund median | Difference |
|---|---|---|---|---|---|
| Anthropic | $76.0M | $617 | $589.01 | $777.53 | -24.2% |
| Psiquantum | $73.7M | $598 | $48.29 | $46.52 | +3.8% |
| OpenAI | $23.0M | $187 | $687.69 | $687.69 | sets the median |
| Databricks | $21.4M | $174 | $203.46 | $253.00 | -19.6% |
| Patreon | $8.2M | $66.83 | $13.18 | $12.16 | +8.3% |
| Discord Inc | $7.8M | $63.55 | $239.61 | $23.51 | different series |
| Anduril | $7.7M | $62.24 | $68.95 | $73.50 | -6.2% |
| Company | Value held | Per $10,000 | Fund carries at | Cross-fund median | Difference |
|---|---|---|---|---|---|
| Apptronik | $37.7M | $1,365 | $36.70 | $36.86 | -0.4% |
| Company | Value held | Per $10,000 | Fund carries at | Cross-fund median | Difference |
|---|---|---|---|---|---|
| Databricks | $23.3M | $299 | $190.00 | $253.00 | -24.9% |
| Epic Games | $19.2M | $247 | $446.42 | $377.90 | +18.1% |
| Figure AI Inc | $10.0M | $129 | $194.93 | $194.86 | sets the median |
| Canva | $9.6M | $124 | $1,646 | $1,496 | +10.0% |
| Anduril | $5.2M | $67.52 | $68.95 | $73.50 | -6.2% |
| Stripe LLC | $617,652 | $7.95 | $63.00 | $70.01 | -10.0% |
| Company | Value held | Per $10,000 | Fund carries at | Cross-fund median | Difference |
|---|---|---|---|---|---|
| Stripe LLC | $25.3M | $361 | $63.00 | $70.01 | -10.0% |
| Waymo LLC | $360,000 | $5.14 | — | $164.32 | — |
| Company | Value held | Per $10,000 | Fund carries at | Cross-fund median | Difference |
|---|---|---|---|---|---|
| Fanatics Inc | $12.4M | $78.81 | $98.13 | $98.08 | +0.1% |
| ByteDance Ltd | $10.7M | $68.14 | $334.24 | $352.46 | -5.2% |
| Databricks | $8.2M | $52.37 | $203.46 | $253.00 | -19.6% |
| Psiquantum | $1.7M | $11.07 | $48.29 | $46.52 | +3.8% |
| Sambanova Systems | $1.7M | $10.83 | $129.41 | $127.10 | +1.8% |
| Epic Games | $1.2M | $7.92 | $440.85 | $377.90 | +16.7% |
| OpenAI | $814,221 | $5.18 | $687.69 | $687.69 | sets the median |
The reality check: Space Exploration Technologies Corp. since it listed
Everything above is a way of owning something before it lists. This is what happened the last time one of these companies did. SPCX priced at $135 — the figure parsed out of its own 424B4 prospectus for the lock-up radar, not a press report — and has traded for 103 sessions since.
Three things in those numbers are worth more than any projection. The first-day pop of +19.2% went to whoever was allocated at the offer price, which was not a retail buyer. Anyone who bought at the 2026-06-16 close is down -36.5% four months later. And the offer price itself was above every mark the funds on our accuracy page carried it at — which is the pattern in all ten completed listings, not a one-off.
Its lock-up expires 2026-12-09. Whether that matters is a question we publish a dated, unrewritable answer to before the fact rather than after — that is what the radar is for.
What this page cannot get you to
20 of the 46 companies we price have no route on this page at all — Xsight Labs Ltd, Lightmatter, Physical Intelligence, Checkr Inc, Nuro, Sila Nano, SB Technology, Mcdermott International Ltd, Cellink Corp, Relativity Space Inc, Bending Spoons Spa, Checkout Payments Group Ltd, Workrise, Kardium, Lilac Solutions, Corporate Travel Management Lt, Kobold Metals, Gusto, Avz Minerals Ltd, Avadel Pharmaceuticals PLC. Every fund holding them is private, institutional, or files no N-PORT we can reach. The honest answer for those names is that there is no public route, and saying so is more useful than routing you to a ticker with a rounding error of exposure.
A fund has to be more than a rounding error to appear. 15 registered funds were scanned and withheld. 4 of them hold nothing we price. The other 11 do hold a covered name, but below the 1% of net assets it takes to call the fund a route rather than a coincidence:
- PCN PIMCO CORPORATE & INCOME STRATEGY FUND — 0.25% of net assets (Databricks)
- PFL PIMCO INCOME STRATEGY FUND — 0.13% of net assets (Databricks)
- PFN PIMCO Income Strategy Fund II — 0.28% of net assets (Databricks)
- PGP PIMCO Global StocksPLUS & Income Fund — 0.19% of net assets (Databricks)
- PDX PIMCO Dynamic Income Strategy Fund — 0.57% of net assets (Databricks)
- PDO PIMCO Dynamic Income Opportunities Fund — 0.25% of net assets (Databricks)
- NBXG Neuberger Next Generation Connectivity Fund Inc. — 0.09% of net assets (Anduril)
- ECAT BlackRock ESG Capital Allocation Term Trust — 0.29% of net assets (Anthropic)
- PAXS PIMCO Access Income Fund — 0.29% of net assets (Databricks)
- PCM PCM FUND, INC. — 0.72% of net assets (Databricks, Mcdermott International Ltd)
- SOR SOURCE CAPITAL /DE/ — 0.74% of net assets (Mcdermott International Ltd, Epic Games)
None of this is advice, and two of the routes are worse than they look. A proxy's private stake is a fraction of a percent of what moves its share price; buying CRM for its private book means owning its entire operating business first and the stake as a rounding error. A closed-end fund at a premium can lose money on the fund while the underlying company gains. Both of those are arithmetic, not opinions, and neither is a recommendation.
Method
- Proxy figures. Each disclosure is pinned to an accession and an anchor phrase. At build time the filing is re-fetched, the anchor must appear in it verbatim, the published quote is the sentence around the anchor, and the stated figure must appear inside that sentence. A record failing any check is published as unverified with no number. The check caught a real error on its first run.
- Look-through. Disclosed value ÷ market cap × $10,000, computed only for cash invested and carrying value. Market cap is the one quote-feed figure used, and only for the proxies — large caps with stable share counts.
- Vehicle premiums use the filer's own tagged NAV (
us-gaap:NetAssetValuePerSharefrom XBRL), never a quote feed's. Two alternatives were tried and both produced wrong numbers: market cap ÷ N-PORT net assets was off by 32 points for DXYZ because the feed's share count is a third below the filing-implied one, and parsing NAV out of the N-CSR HTML fails because BlackRock files one combined report covering three trusts with identical text, so column selection would publish another fund's NAV. A NAV more than 120 days from the price is refused rather than shown. - Single-purpose vehicles are unwrapped to the company underneath, and an unwrapped position gets no per-share price: its reported share balance counts units of the wrapper, not shares of the company. Ignoring that published $5.1 million a share for a company priced at $37 elsewhere on this site.
- Cross-fund comparisons stop at 100%. Preferred stock comes in series and a late series can be worth several times an early one, so beyond that gap the two figures are not describing the same security and the difference is labelled rather than published.
The whole snapshot, including every refusal, is free at /api/private-exposure.