When Fundrise's Innovation Fund (VCX) listed on the NYSE on March 19, 2026, its net asset value was $18.97 a share. Within days, retail investors momentarily bid the price above $400. That's more than 20X NAV, a premium so rich it made 1999 look responsible.

I was locked up through the whole thing. The lockup was originally set for September 14, 2026, then moved up to August 13. My shares hit my brokerage account on August 17. I wrote about the psychology of that wait in a companion post on what it's like to be a startup employee with a six-month lockup.

This post is about the math and the potential for what VCX could trade at in the future.

In July I published a NAV estimate built off the 3/31/2026 data. On August 28, the 6/30/2026 N-PORT hit EDGAR.

I'm revising my VCX NAV estimate down by about 13%-23% for year-end 2026, 2027, and 2028. Not because the method was wrong, but because three quarters of what moved the number was invisible until the filing landed. That includes the thing I have not seen reported anywhere: It seems as if VCX sold roughly a third of its Anthropic position during the quarter, at the old price, right before marking the rest of it up.

Note and TLDR: Note and TLDR: I published my original VCX NAV analysis on July 15, 2026, four months after the listing, deliberately waiting for the retail mania to die down before running the numbers. Those estimates were pro forma, built off the 3/31/2026 data, because that was all anybody had.

The 6/30/2026 N-PORT landed on August 28 and a lot changed in those three months. I've revised my NAV estimates for year-end 2026, 2027, and 2028 down between 17% – 23%.

My VCX NAV target is now $50 a share in 2027, down from $65. We are still in the early innings of the AI boom and the NAV keeps growing. What really moved the needle during the lockup, though, was retail demand bidding the price to multiples of NAV. That's the setup I'm willing to hold out for. Nothing is guaranteed.

Fundrise is a long-time sponsor of Financial Samurai, and I am a multiple six-figure investor in Fundrise. products. All views are my own.

Why My VCX NAV Estimate Came Down

The 6/30/2026 filing puts VCX's NAV at $21.70 a share. I had modeled roughly $26-$27 for July.

It is worth walking through what actually moved the number. Because 75% of the revision came from information that simply was not public in July.

There are four drivers. Here they are in order of size.

1) The rest of the portfolio did not move. 27.9 percentage points of NAV growth, or $5.29 a share.

This is by far the biggest driver, and it was unknowable. OpenAI was held flat to the dollar. Databricks flat. Epic Games flat. Fivetran flat. And Anduril was marked down 28.6%.

I had assumed those holdings would together add about 29 points of NAV growth, based on their publicly reported funding rounds. They added 1.2.

Private marks are not published between quarters. There was no way to know from the outside that a fund would leave four of its largest positions untouched while the headlines said their valuations were climbing.

On Anduril specifically, I wrote in July that secondaries were trading near double its last round. That was accurate reporting of the secondary market. The fund's valuation committee simply took a different view, and their view is the one that sets NAV.

2) Part of the Anthropic position was sold. 6.8 points. 

Nobody knew this until August 28. More on it below, because it is the most interesting thing in the filing. Thankfully, Anthropic is still VCX's #1 holding, which will be highly sought after by retail investors.

3) The Anthropic weight depends on which denominator you use. 6.5 points. 

This one is worth explaining, because both published figures turn out to be right.

I used 20.7%, which is what Fundrise showed on the fund's own holdings page in February when they planned to list. The annual report carries Anthropic at 16.5%. Two numbers from the same sponsor, four points apart.

They reconcile. Anthropic was 20.7% of the venture book and 16.6% of net assets. The gap is everything in the fund that is not a startup: cash, a Treasury money market fund, eight tranches of data-centre asset-backed bonds and a promissory note. Roughly $136 million, or a fifth of the fund. That is a wild card because we don’t know what management will do with this cash buffer at any given time.

So both figures are correct. They just answer different questions.

The catch is that NAV per share is calculated on net assets. Cash and bonds earn a coupon. They do not mark up when Anthropic raises. Apply a venture-book weight to a net-asset NAV and you overstate the result every time.

The split still applies today. As of 6/30, Anthropic is 22% of net assets and about 29% of the venture book. Use the 22% when you're modeling NAV. Use the 29% when you're asking how concentrated your AI bet really is.

4) The markup base. 4.9 points. 

I ran $350 billion pre-money Series G to $900 billion pre-money Series H and got a 2.57X markup. The fund booked 2.27X, which is what you get measuring from the $380 billion post-money instead.

This is a convention question rather than a factual error, and it is the one thing on this list I could have gotten right in July with the data I had.

Add it all up and one holding did 96% of the quarter's work. Anthropic contributed $97 million of a $101 million gain. Everything else, net of the Anduril writedown, was noise.

The takeaway is not that public-data modeling is hopeless. It is that this fund only tells you the numbers four times a year, and between those dates you are working with headlines that the valuation committee is under no obligation to agree with. Precision is impossible when modeling a closed-end NAV estimate.

The Anduril reduction is fascinating, and odd. Because Anduril did not raise a down round during this time. Fundrise may have acquired shares of Anduril above the latest funding round valuation, and then marked the shares down to the last funding round valuation. As a result, there should be upside to this valuation and contribution to VCX's NAV if Anduril raises at a much higher valuation.

Anduril's last round valuation was $61 billion on May 13, 2026. But in the secondary market, the shares are trading above a $100 billion valuation. So if there is another funding round, which there will likely be, expect a boost to VCX's NAV.

The Three-Step Method, In Short

I walked through the full mechanics in the July version of this post, so here's the compressed version.

First, calculate the dilution-adjusted markup for each holding. When a company raises, you can't just divide the new post-money by the old one, because new money dilutes existing shareholders including VCX. Use pre-money against the prior post-money.

Second, multiply each growth multiple by its portfolio weight. The formula is portfolio weight times markup multiple minus one. You subtract the one because the original position is already sitting inside the starting NAV. You only add the growth, not the whole new value.

Third, stack the contributions in dollars per share.

The method holds up. What it needs is the right denominator and a healthy respect for the fact that a fund does not have to mark a position just because the press reported a round. So I now anchor everything to the last NAV Fundrise actually reported and add only what I can verify, instead of building up from a stack of headlines.

VCX Quietly Sold A Third Of Its Anthropic Position

Here's the part that surprised me most.

At 3/31/2026, VCX held Anthropic through three separate vehicles worth about $112 million. By 6/30 one of them was gone. Roughly 139,000 shares, about $36 million, close to a third of the Anthropic exposure.

It went out at around $258 a share.

Weeks later, the fund marked the two remaining Anthropic vehicles at $589 a share for that same quarter end.

Had those shares been held, NAV would have printed about $22.99 instead of $21.70. That single position is $1.29 a share, or 5.9% of the entire fund's NAV. Perhaps the SPV sold unilaterally or Fundrise determined to de-risk its exposure in Anthropic or the HOF SPV that held Anthropic.

Why did it happen? Nobody knows, and I want to be careful here.

N-PORT does not disclose sale proceeds or the reason for a disposal. The vehicle was a third-party SPV, so it may have simply reached the end of its life and distributed on its own schedule rather than Fundrise choosing to sell. It could be deliberate risk management, trimming a position that had roughly doubled. It could be something else.

What I can rule out is forced selling. Shares outstanding did not change and the fund reported zero sales and zero repurchases across all three months. Nobody was redeeming.

Where The Money Went

VCX deployed $64.3 million during the quarter. About $54.3 million of it went into vehicles whose legal names identify no underlying company at all.

The fund says so itself, right in the schedule: the legal name of an investment vehicle may not identify the underlying company to which the fund has economic exposure.

Vehicles like that went from about 11% of net assets to roughly 22% in a single quarter.

So every company mapping in this post, mine or anyone else's, is inference from per-share arithmetic. It is not disclosure.

Private funds structure through SPVs for real reasons, including access. But it does mean the thing you're actually buying got harder to see, not easier, in the first full quarter after the fund went public.

The NAV You See Quoted Is Stale

Look at the monthly returns in the filing. April 0.00%. May 0.00%. June plus 14.39%.

There is no interim NAV anymore. The continuous offering closed in February and the tender program ended in March, so nothing forces a monthly strike. Everything books at quarter end.

Any VCX NAV you see quoted between quarter ends is interpolated. Your brokerage screen is showing you a number nobody calculated.

The next real mark is 9/30/2026, and it won't be public until late November. Which means if Anthropic prices in October, the IPO happens entirely inside a reporting blind spot.

But The $21.70 VCX NAV Is Already Two Months Stale

Here's the part that cuts the other way, and it matters more than the revision.

The 6/30 NAV is a photograph of June 30. It is now August 28, 2026. Anything that happened in between is not in that number, and almost everything that happened in between was good.

But I'm not going to hand-wave this. This fund marks to valuation events, not to sentiment. A funding round counts. A press report does not. So here is only what has actually triggered.

Databricks. On August 13 it closed $5 billion at a $190 billion valuation, up from $134 billion six months earlier, with revenue past a $7 billion run rate growing more than 80%. That is a real event on a position worth about 12% of the fund. Call it plus $1.02 a share.

SpaceX. It's public and marked daily. The fund's 6/30 mark implies about $130 a share and SPCX is near $141 now, though it has been a wild ride from $225 down to $105. Plus about $0.08 a share

Fees. 1.85% for two months. Minus about $0.08.

That puts my estimate of today's NAV at roughly $23.

Which is a modest number, and the reason it's modest is the whole point of this section.

Anthropic Has Not Been Re-Marked Since May

Anthropic is ~22% of the fund. Its carrying value has not moved a dollar since the Series H closed on May 28 at $965 billion, because there has been no new primary round to mark against.

Over that same stretch, the company's annualized revenue run rate went from $47 billion in mid-May to more than $65 billion at the end of July. Nearly 40% growth in ten weeks. It posted its first operating profit, filed confidentially, hired three banks, and investors are now reportedly targeting a $2 trillion IPO valuation.

None of that is in the NAV. Not one dollar of it.

That is not Fundrise being slow. That is how private fund accounting works. You mark to the last round until there is a new one, and Anthropic's next repricing event is the IPO itself.

Which means the repricing does not arrive gradually. It arrives in a single print. This is why private fund NAVs move in step functions instead of curves.

If Anthropic lists near the $2 trillion investors are targeting, its slice of VCX goes from $4.85 a share to roughly $9.25. NAV moves from about $23 to about $27 on that one holding alone.

And if demand is incredible and Anthropic trades to $3 trillion out of the gate, that slice becomes roughly $13.90 and NAV lands around $32. A $3 trillion Anthropic is 3.1 times the fund's current carrying mark, so a single position would be worth more than the entire fund's NAV was in June.

So the honest way to hold this is not to argue $21.70 is wrong. It's the right number for June 30. The argument is that the largest holding is frozen at a May valuation while the business behind it grew 40%, and the mechanism that unfreezes it is scheduled for the fourth quarter.

One caveat before you get carried away. A fifth of this fund cannot compound at all. Cash, Treasuries and data-centre bonds earn a coupon, not a markup. When Anthropic doubles, they sit there.

And You Won't See It When It Happens

Remember that the fund only strikes NAV quarterly now. The 9/30 mark won't be public until late November, 2026.

So if Anthropic prices in October 2026, there will be a stretch of weeks where the repricing has occurred, every VCX shareholder owns it, and the last officially reported NAV is still a number struck back in June.

That gap between what the fund is worth and what the last filing says it's worth is not a bug I'm complaining about. For anyone buying, it is the entire opportunity.

My Revised VCX NAV Estimates

Now let's look further out. Anthropic is 22.35% of the fund, worth $4.85 per VCX share at the current $965 billion carrying mark. The other $16.85 of NAV is everything else.

Hold everything else flat and Anthropic alone gets you this. Holding everything else flat is conservative, as names like Databricks, Anduril, and OpenAI are still growing aggressively.

Anthropic valuation	VCX NAV per share
$830 billion (today's mark)	$21.70
$2 trillion	about $28
$3 trillion	about $33
$5 trillion	about $44

Layering in my own assumptions for the rest of the portfolio, and subtracting the annual management fee that shareholders approved in the February 2026 proxy, here's where I land.

Why My NAV Target Is Now $50, Down From $65

My old target of $65 was a year-end 2027 NAV estimate with no premium attached. $50 is now more reasonable, and still plenty of upside.

Here's what a $50 NAV actually requires.

If the rest of the book	Anthropic must reach
stays flat	$6 trillion
grows 50%	$4.5 trillion
doubles	$3.0 trillion
triples	reached without any Anthropic gain

At $65, I needed Anthropic at $5.8 trillion even if the rest of the portfolio doubled. That's possible in due time, but unlikely for now.

At $50, a doubling of the rest plus Anthropic at $3 trillion gets there. $3 trillion is already my stated 2028 view, so the target and the thesis agree with each other.

One more thing worth saying plainly. $50 is not purely a NAV call. Against my more conservative base case where Anthropic is closer to $1 trillion market cap, reaching $50 in 2027 means collecting a 47% premium to NAV. In 2028 it's a 14% premium. Only in the blue sky case does $50 require no premium at all.

The Premium Is The Risk, Less So The Portfolio

Here's the dose of humility every VCX shareholder needs. The NAV rising 14.4% in a quarter and the share price falling more than 85% from its peak are both true at the same time.

VCX is a closed-end fund. There's no creation-and-redemption mechanism tethering price to NAV. Retail investors who paid 20X NAV weren't buying a portfolio. They were buying a lottery ticket on scarcity, since VCX was one of the only ways the public could own Anthropic, OpenAI, and SpaceX in a single ticker.

That scarcity is on a countdown clock. SpaceX is already public. OpenAI is targeting 2027. Anthropic has filed and will likely list at the end of 2026. Every IPO gives investors a way to buy each company directly, so the premium should erode over time.

The key word is over time, because sequencing matters. The longer VCX's holdings stay private, the longer the scarcity premium survives. This is why the Databricks CEO guiding toward a 2027 IPO instead of 2026 is good news for shareholders. Databricks is still the second largest holding at roughly 12% of the fund, and every year it stays private is another year VCX remains one of the only tickets to the show.

My math says the NAV keeps climbing, just from a lower base than I thought. It still says nothing about the premium.

Modeling In Mania Is Difficult

Never in my wildest dreams did I think VCX would go up 3X, 5X, 10X, 20X post listing, given I focus on fundamentals. As a FIRE investor without a steady paycheck, I cannot afford to be too wrong. But retail enthusiasm is now a variable investors have to carry in the model. We saw it with meme stocks in 2021 and the Reddit army has only grown since.

There's precedent for how high VCX can fly. It surpassed $380 in March, and over $250 in May. Anthropic is worth far more now than it was in the spring. Same retail intensity plus a bigger underlying asset means mania could bid VCX higher to a large premium again. Just not as extreme given there are more shares that can be sold.

The biggest catalyst is Anthropic going public now that the lockup has already expired. Sitting here in San Francisco, I'm 80% certain the demand for Anthropics IPO will be unprecedented. I got a glimpse of it during the spring 2026 selling season when home prices jumped 20% year over year as buyers moved ahead of expected liquidity events.

But I said something similar in July, and then the filing showed up and cut my estimate by -~20%. So take my conviction with the appropriate discount.

A Potential Ideal Scenario For VCX Shareholders

My ideal setup is perverse. As a long-term shareholder since 2023, trying to build a $500,000+ position in Anthropic alone before its IPO, I want VCX shareholders to read the 6/30 N-PORT, panic and dump shares.

That's my chance to accumulate more shares at a lower price before Anthropic's S-1 drops. I’m about $380,000 of the way to my $500,000 Anthropic position through VCX and traditional VC funds. Then the market learns the IPO is a go, the media covers the growth and the demand non-stop, retail mania returns, and the price runs well past my $50 target.

Buy the fundamentals. Get the potential frenzy for free.

The other X factor is the holdings nobody talks about. Databricks was once VCX's largest position at roughly 20% while Anthropic sat at about 5%. Weights move. One or two of the smaller names could run the way Anthropic did, and I wouldn't count OpenAI out. They are quietly making significant moves.

But whatever happens, I'm fine if VCX simply trades around a growing NAV for the next three years. It will probably never again fetch the 1,000% to 2,000% premium it did in March and May. But a 50% to 100% premium while everybody is clamoring for Anthropic allocation? That has precedent, and it's worth holding out for.

And if the frenzy never returns, so be it. I'm left owning the portfolio at around NAV, which is what I thought I was buying in the first place as a shareholder since the beginning of 2023.

Reader Questions

Do you own VCX, and if so, did you buy pre-listing at NAV, post-listing at a premium, or after the lockup expired? Does anyone have a read on why that Anthropic vehicle came off the books at the Series G price? And where do you think the premium settles once Anthropic actually trades?

Before making any investment, please do your own due diligence and only invest what you can afford to lose. Nothing here is specific investment advice for you. This is simply how I'm thinking about my own shares, and our circumstances, financial goals, and time frames are different. My calculations will not be exact and will change with time.

Invest At NAV

I just spent a whole post revising a number downward. You know what requires no revision at all? The price you paid to get in.

Investors who bought VCX at its $10 a $18.97 NAV pre-listing are up on fundamentals alone, even after I cut my estimate. Investors who paid a 10X premium are still deeply underwater owning the exact same portfolio. Same Anthropic, same OpenAI, wildly different outcomes. The only variable was entry price.

Fundrise has reportedly filed to launch VCX 2, though timing and final structure remain uncertain. If a sequel launches, the window that matters is the pre-listing one, where you buy at NAV like the first VCX's biggest winners did. Existing Fundrise investors get notified first, and you can open an account to get on the list. Free to sign up, and it beats setting a price alert and praying.

In the meantime, every fund on Fundrise's platform, from private real estate to private credit, transacts at NAV. No premium to overpay, no lockup supply wave, no mania variable to handicap. You get the fundamentals without the frenzy.

Disclosure: Fundrise is a long-time sponsor of Financial Samurai, and I am an investor in Fundrise products, including the Innovation Fund (VCX). All NAV projections in this post are my own estimates. The $21.70 figure is derived from the fund's own 6/30/2026 N-PORT filing. This is not investment advice. Please come up with your own VCX NAV estimates, because that is what you'll be investing against, not the past