On September 22, 2026, Kelly Schricker, who runs Decile Labs and oversees Decile Group's venture education programs, and Max Harris, who leads the LP Institute, hosted a public session on investing in venture capital as a limited partner. Most of the people Max works with built their wealth somewhere other than venture, so a good part of the hour went to the thing that surprises new LPs most. A venture fund usually looks like a failure for its first few years, and that's the plan working.
Watch the full session here: Investing in Venture Capital as a Limited Partner
This article covers the J curve specifically. For who LPs are and how fund commitments work, read How to Invest in Venture Capital Funds as a Limited Partner.
Neither speaker is a financial adviser, and none of this is personal investment advice.
The short version
- The J curve is the shape a venture fund's returns trace over time. Down first, then up, with the climb arriving late.
- Years one to three are the dip. Companies shut down, others stall, and your updates read like bad news because they are.
- Years four to seven are the climb, when the companies that are working start working loudly.
- The dip is what the power law looks like from inside a 10-year commitment, rather than a sign the manager picked badly.
- The curve is compressing. Companies hit milestones faster than they used to, and AI is a large part of why.
What the J curve actually is
Draw the letter J. That's the shape.
A fund takes capital, deploys it into startups, and for the first stretch the only news that arrives is negative. Some portfolio companies fold. Others go quiet. Nobody's had an exit yet, so there's nothing on the other side of the ledger. The line goes down.
Then, somewhere around year four, the companies that found product market fit start to compound. Valuations mark up. A few get acquired. The line turns and climbs.
Kelly framed it in the session as two phases. The first one to three years are the dip, and years four to seven are the climb. Some companies plateau along the way, and the returns, when they come, come late.
Why the dip happens
The J curve is downstream of the power law, so it helps to take them together.
Venture returns don't form a bell curve. Most investments in a fund are expected to return nothing, and one or two are expected to carry the entire portfolio. That's the power law, and we've written about it separately in The Power Law in VC.
Run that forward and the timing falls out of it. Failures surface early, because a startup that isn't working runs out of money fast. Successes surface late, because compounding takes years and an exit takes longer. So the bad news is front-loaded by design and the good news is back-loaded by design.
An LP who doesn't know this reads year two as evidence the manager can't pick. An LP who does know it reads year two as the portfolio behaving normally.
Fees deepen the dip
One mechanic the session covered makes the early curve steeper. Management fees come out from day one.
The venture standard is 2 and 20. The 2% management fee covers salaries, office, software and the rest of the operating line, and it's charged whether or not anything has exited yet. The 20% is carried interest, and it usually doesn't start until the fund has returned all of the original capital to its LPs.
So in year two you're paying fees against a portfolio that hasn't returned anything. That's part of why the line dips rather than just flattening.
Kelly made a related point on the carry side. Because the manager doesn't earn carry until you've been made whole, both sides want the same thing. Raise $5 million, return $10 million, and the first $5 million goes back to investors before the manager takes 20% of what's left.
Venture is a milestone business, and that limits the damage
Kelly offered a comparison that explains why the dip doesn't keep going down forever.
If you're financing an apartment building and the money runs out before the doors are installed, somebody still has to buy doors. The commitment follows you. Venture doesn't work that way. Each round buys a company its next milestone, and a company that misses the milestone usually doesn't get the next round.
The investors change at each stage too. Whoever writes the first $250,000 check isn't the firm writing the Series B. A seed investor is judging a team and an insight. A growth investor is reading a P&L and a data room. So a struggling company generally isn't coming back to its earliest backer for a rescue, and your fund isn't throwing good money after bad.
Managers do sometimes stay in. Pro rata rights let a fund hold its ownership, and some funds run an explicit follow-on strategy, backing a company at pre-seed and again at seed or Series A. That's a choice the manager makes, not an obligation the structure imposes.
The curve is getting shorter
The old rule of thumb was ten years to an IPO. Kelly said that window's compressing, and gave the reason.
Companies are hitting milestones faster than they used to. A startup doesn't need a large engineering team to build something real any more, which lowers the cost of reaching each milestone. AI is doing a lot of that work. Founders have more capability per dollar than they did five years ago, so the distance between rounds shrinks.
The effect on the J curve is that it narrows. The dip doesn't last as long, and the climb starts sooner. Max noted the mild irony that this is happening at the same moment the secondary market has matured to solve the liquidity problem the long curve created.
If you need liquidity before the climb
A ten-year lock is a real constraint, and there's now a market around it.
Direct secondaries, LP-led secondaries where an LP sells its position in a fund, and GP-led deals including continuation funds all exist. Selling early usually means selling at a markdown, which is the price of not waiting.
Secondaries also move in and out of fashion. They've been busy recently, largely because investors want exposure to companies they can't reach directly.
It's also worth remembering that an IPO isn't the only way a position turns into cash. With public listings slower than many LPs would like, acquisitions carry a lot of the return.
Entry point and the manager you back
If a fund has $25,000 to put into a company, it would rather buy at a $1 million valuation than at $100 million. Same money, far more ownership, far more room above it.
That's a large part of the case for backing emerging managers. A firm writing early checks is buying at entry points the established funds have priced themselves out of, and it takes fewer outcomes to return the fund. Kelly pointed to Lowercase Capital as the example to go read about, a fund of roughly $7.5 to $8 million that produced returns nobody has matched since, while noting she couldn't recall the exact figures.
Press coverage favors the largest funds partly because a 2% fee on a very large fund pays for a PR team. Emerging managers investing in their own regions rarely get that coverage, and they're often buying at the entry points that produce the multiples. If you want to understand how those managers find their first backers, our recap of You Already Know Your First LPs covers it from the GP side.
How to sit with the dip
- Before you commit, decide what you'll do in year two when the shutdown notices arrive. If a run of them would change your behavior, size the commitment smaller.
- Ask the manager which stage they invest at. A first-check fund has a longer, deeper dip than one writing Series A checks.
- Ask whether they run a follow-on strategy and how much of the fund is reserved for it. Reserves change how the curve looks.
- Treat early markdowns as information about the portfolio, not about the manager. Judge the manager on sourcing, access and reasoning instead.
- If your horizon is genuinely shorter than ten years, say so at the start and ask about secondaries, rather than discovering the constraint in year three.
- Don't read a single mark, up or down, as the answer. The power law means one position decides most of the outcome, and you don't know which one yet.
Questions people ask
What is the J curve in venture capital? It's the shape a fund's returns trace over its life. Early losses and fees pull the line down, then exits from the few companies that work pull it up, producing a J.
How long does the J curve last? Historically the dip ran one to three years with the climb through years four to seven. That's been compressing, because companies are reaching milestones faster.
Why does a venture fund lose money at first? Failures show up early and exits show up late, and management fees are charged throughout. Nothing has been returned yet, so the line goes down before it goes up.
Is the J curve a bad sign? On its own, no. It's the expected path. What's worth watching is whether the manager's reasoning holds up and whether the portfolio is reaching its milestones, rather than what the marks say in year two.
Watch the full session
Investing in Venture Capital as a Limited Partner, recorded September 22, 2026 with Kelly Schricker and Max Harris. More sessions on the VC Lab YouTube channel.
If backing funds is where you want to be, our recap of what to expect from the LP Institute covers how that program runs, and the LP Institute itself is free to join. Cohort 7 is open.
Launch your own fund with VC Lab
Venture capital, done right, is one of the most powerful tools we have for solving hard problems. That conviction is why Decile Group gives its programs away. VC Lab is a free 14-week accelerator that has helped launch 1,003 venture firms across more than 90 countries, with no fees and no equity, because the industry gets better when more people can get into it rather than fewer.
If you'd rather run the fund than back it, Apply to VC Lab at govclab.com. Managers already raising funds two through four should look at the Emerging Institute, and anyone learning the industry from the ground up should start with Venture Institute.