What SPV Fees Can Cost You at 2x, 3x and 10x

Doing the mental math on VC returns in SPVs is harder than it looks. There are often multiple fees, sometimes layers of fees, and it's tough to play out scenarios when your attention is on the only question that feels urgent: is this a great investment?

So most people default to the simple version. If the valuation doubles, I'll about double my money. That's wrong for a few reasons. The first is dilution, which I'm skipping for this article. The second is fees, which is what I want to focus on here.

For simple math, let's assume you make a $100k investment into a venture SPV.

Here's the fee stack you'll see across most of the syndicate market today:

• Carried interest: 20%

• Fund admin / deal fees: 1–10%

• Management fee: 0–20%

Those last two numbers are the ones people gloss over, because they sound small next to the carry. They aren't. Fees come out of your principal — they never buy shares — so they don't just cost you the fee, they cost you the fee and every dollar of return that money would have earned.

Let's run the numbers.

The three structures

Three fee versions of the same investment:

I picked structure A and B because they're typical of what I see in the market. In some cases I see north of 20% in total fees, between access fees, management fees and fund admin. But I wanted to use cases I come across more consistently. Structure C I picked partially due to what we’re building… more on that later. 

The main mechanic that gets overlooked here is that cash fees come out of principal. In Structure A, a $100k investment only $88,000 buys stock. If the company doubles, that's a 2x on $88,000 — not on $100,000. 

What a 2x and a 3x actually pay you

Same company. Same entry price. In Structure A your 3x is a 2.31x. 

In dollars of profit: $131,200 (Structure A) vs $180,000 (Structure C). That's nearly $49,000 of gain you didn't get.

Here's several outcomes, side by side. Shorthand in the columns is carry / fund admin / all in management fee.

On the one deal in your portfolio that actually works — the 10x that's supposed to pay for everything else — the structure costs you $186,000 on a $100,000 check in Structure A versus C.

Three things the table is telling you

1. You start underwater.

In Structure A, the company has to return 1.14x just to get you back to even — $100,000 of capital divided by the $88,000 that actually bought stock. Structure B needs 1.04x. Structure C needs 1.00x. Every SPV that exits at a modest markup — and a lot of them do — is a losing trade for the LP in a high-fee wrapper, even though the underlying investment did nothing wrong. Funds partially solve some of this by recycling. SPV managers typically don't.

2. Fee drag is worst, proportionally, at the outcomes that happen pretty often.

At 1.5x, Structure C pays you 76% more profit than Structure A. At 10x, it's 30% more. The percentage gap narrows as outcomes get bigger, because carry starts to dominate the fixed fee drag.

Which means the damage concentrates in the 1x to 3x band — and that band is where most SPVs that don't go to zero end up. The modest outcome is the common outcome, and it's the one this structure punishes hardest.

3. The absolute dollars go the other way.

$12,000 of fees is $12,000 at 1x. At 10x it's $120,000 of forgone proceeds — plus the carry differential on top. Small percentage, big band. Both ends hurt; they just hurt differently.

Now do it twenty times

You don't pay these fees once. You pay them on every deal, including the ones that go to zero.

So: twenty SPVs, $100k each, $2m in. One returns 20x, two return 5x, three return 2x, four give you your money back, ten go to zero. Across the portfolio that's a 2x gross — an okay run in venture, carried by one winner.

Same twenty picks in all three structures. Only the paperwork changes. Structure A nets you 1.50x. Structure B, 1.63x. Structure C, 1.85x.

You caught the 20x and you still take home somewhere between 1.50x and 1.85x depending on which wrapper you signed. In dollars of profit, $1.0m versus $1.7m — a $694,000 spread on identical decisions.

The version I didn't model

I kept these scenarios somewhat conservative on purpose. But 20% (and more) management fees exist, and I've seen them. Add 4% fund admin on top and only $76,000 of your $100,000 ever buys stock.

That structure needs a 1.32x just to return your capital. At a 3x it pays you around 2x — your triple is a double. These get subscribed anyway, because the allocation is hot and LPs are price takers when it is. Few negotiate fees on a deal they're worried about getting cut from.

Here’s where I defend fees

Fees are not theft. Someone has to form the entity, file the K-1s, manage the cap table position, chase the company for updates, and hold the line for eight years. Fund admin is a real cost with a real invoice attached, those are real resources, and a lead who sources genuinely proprietary allocation is worth paying for.

Managers also need to live, and they need to hire if you expect any real support behind the deal. They can't run on carry alone. Realistically none of that carry gets paid for years, and a manager who can't fund the work in the meantime becomes a worse manager for their LPs.

So the question was never whether to pay. It's what you're paying for, and how much of it comes out before your money ever buys a share.

But there is a different model that’s rarely seen in the market - Structure C

Structure C wasn't a hypothetical. It's part of what we've been building for almost a year, and it's part of the reason I ran these numbers in the first place.

I can't share much today. More in a few weeks. 

✍️ Written by Zachary and Alex