The headline is that valuations are soaring — the top 5% of seed rounds jumped 177.6% from 2025 to 2026, faster than at any point in the past decade. The 95th percentile seed round now clears $200M. The 90th is pushing $98M. Even the median sits at $29M.

A lot of people have written smart takes on this data. But there’s one question I haven’t seen many people ask, and it’s the one that keeps me up at night: where on this curve do you live? And has that changed over time?

Let me make it crystal clear up front that the market isn’t perfectly efficient, so at no point am I assuming that a company at a median valuation is necessarily worse than one at a top-quartile valuation. If anything, the very top might be the most overvalued slice of the whole market. The 95th percentile is often where consensus founders meet consensus ideas, and everyone pays up for the privilege of being in the same round. So that may be exactly where you don’t want to invest.

But I think anyone who’s done this for a long time has been pulled towards investing in fundamentally different companies than they used to — mostly without noticing. Our sense of what counts as an acceptable valuation may have gone up. But I highly doubt it has gone up nearly as much as valuations themselves have. Like a frog in slowly boiling water, an investor can slip from buying primarily top 25th percentile-priced companies to the 50th percentile or lower without ever registering the change. I think that has a lot of implications.

If you’d asked me what NextView’s valuation philosophy was 10 years ago, I’d have said we were price-aware but willing to meet the market. I would have guessed our typical investment was priced somewhere around the second quartile of market valuation, with some exceptions that I think were clearly at top-decile prices at the time. There were certainly plenty of people who thought our deals were expensive, even if, in retrospect, they were absurdly cheap.

Now think about where the typical institutional seed or pre-seed fund is trying to live today. My sense is that it’s at the third or fourth quartile in terms of valuation. This isn’t intentional, but it’s the result of the ownership target tail wagging the dog. Check size and target ownership are often locked in first, and valuation discipline gets reverse-engineered to hit an arbitrary ownership number.

What this can lead to is a firm that used to invest in companies priced at the top 25th percentile now investing at the median or bottom pricing quartile without explicitly deciding to do so. It’s not like you are trying to invest in different companies, but your portfolio model ends up pulling you towards different parts of the market.

In an inefficient market, buying below the median is often where the returns are. But even as the market has skewed toward more consensus thinking in recent years, it’s hard to argue that it’s less efficient than it used to be. There is far more capital and far more players chasing opportunities than there were 10 years ago, and a very large number of companies successfully raise early-stage funding. When that much money floods a space, pricing tends to get more efficient, not less. More capital and more allocators should mean fewer diamonds in the rough, and it should mean the diamonds that do exist get priced more accurately.

Which brings me to the actual bet. Every investor is making one, whether they say it out loud or not.

If you’re honest with yourself and you know you used to buy at the top 25th percentile and now systematically buy at the bottom 25th, you’re implicitly claiming one of two things. Either your judgment and your ability to spot outliers have improved dramatically relative to everyone else’s. Or the segment of the market you’ve moved into is meaningfully more inefficient than it was before. Maybe that’s true. But it’s worth saying out loud, so that it’s an explicit bet and not just something that happened to you because the market moved.

Conversely, if you’re still investing at the same valuation percentile you always did, you’re making the opposite bet. You’re saying that as much as prices have gone up, the upside has expanded enough to more than justify paying it — even as the spread between the top decile and everything else has blown wide open. This is the argument I do hear articulated often, but mostly by the large multi-stage firms.

There’s a subtler cost, too, and I think it’s underrated. If you live in one valuation band long enough, your lens for quality starts to erode. If you don’t regularly see companies priced in the top 5, 10, or 25th percentile, do you actually know what those look like anymore? Yes, many of these companies are just overvalued, but I think it’s safe to assume some high-priced companies are genuinely excellent and are more than worth the high price.

When you mostly see and spend real time with companies that fit your buy box, you slowly lose the fidelity to evaluate the things sitting far outside it. And I think that’s a big problem. It’s why I think it’s a mistake for many seed funds to ignore YC companies just because “valuation expectations are crazy.” That may be true, but their ability to help propel excellent companies repeatedly cannot be ignored, and one needs to stay close enough to those types of companies to spot their commonalities.

And it cuts the other way too. Funds that live in the top end of the valuation curve lose their feel for what a great pre-seed looks like. They can’t tell the difference between a real gem priced at $10M pre and something priced that way because it’s adverse selection. From a distance the two may look identical. You only learn to tell them apart by seeing enough of both.

I’m trying not to make a value judgment here. Both living at the high end of the valuation spectrum and living at the low end could yield excellent results. But what I’ve found is that just looking at things through this lens is clarifying, and a little uncomfortable. Because there was a stretch where I thought I was living on one part of the curve, when in fact, I was living somewhere else. When I realize that the curve had shifted, it forced me to reckon with what bet I was willing to make, and refactored my mental model in pretty big ways.