Why Most VCs Are Indexed to Last Cycle

After years investing across venture and private equity - from backing AI moonshots to enterprise infrastructure plays, I’ve learned that the most expensive words in venture capital are “everyone else is doing it”. Yet here we are in 2025, watching the same tired playbook unfold again: 50.8% of all global VC funding went to AI-focused companies in Q4 2024, doubling from the same period in 2023.

The uncomfortable truth?

Most VCs are perpetually fighting the last war, optimizing for yesterday's opportunities while tomorrow’s category-defining companies germinate in the shadows.

Every venture partner worth their carried interest will tell you about pattern recognition - the holy grail of identifying winning founders and market dynamics before they become obvious. But there’s a dark side to this superpower: when your patterns are calibrated to the previous cycle’s successes, you become an index fund to last quarter’s winners.

I see this playing out in three destructive ways:

• The Zeitgeist Trap: VCs chase whatever sector is hot, creating an influx of capital that helps push forward new technology but also inflates valuations. Today’s AI frenzy mirrors the identical behaviour patterns we saw with crypto in 2021, consumer social in 2011, and clean tech in 2007. The technology changes; the herd behaviour remains constant.

• The Metrics Mirage: We’ve become obsessed with optimising for the wrong KPIs. ARR growth rates that made sense in a ZIRP environment suddenly look quaint when capital costs real money. Yet most partnership discussions still center around SaaS multiples from 2021, not the unit economics that will matter in 2007.

• The Consensus Premium: There’s a terrible herd mentality in the VC community where investors feel it’s better to do a bad investment that everyone else is doing than risk being wrong alone. This creates systematic misplacing - not just of individual companies, but of entire sectors.

Breaking free from last-cycle thinking requires fundamental changes to how we source, evaluate, and support companies:

Source Against Consensus: The best deals come from founders most VCs won’t take seriously. This means going to the technical conferences, not the networking events. Reading research papers, not pitch decks. Following the money flowing to PhD programs, not MBA programs.

Evaluate on First Principles: Strip away all the pattern matching and ask: What core human need is this solving that hasn’t been solved before? What’s structurally different about the world that makes this possible now but impossible three years ago? If your answer starts with “it’s like Uber for…” you’re probably indexed to last cycle.

Support for the Long Game: The companies that define the next decade won’t hit product-market fit in 18 months. They need patient capital that understands non-linear development cycles. This is where my hedge fund background provides crucial perspective - understanding that the highest IRR often comes from the longest hold periods.

2025 is shaping up to be a pivotal year for venture capital, with trends like continued rise of AI in investments and relevance, and hopefully an improved exit market with M&A and IPO slowly picking up is painting an optimistic picture. But optimism in venture is often the enemy of returns. The best investments are made when the market structure rewards patience over momentum.

We’re entering a period where the venture capital investment market is set to grow significantly to $364.19 billion in 2025, a growth of 20.7%. More capital chasing the same opportunities is a recipe for compressed returns - unless you’re willing to look where others aren’t.

The paradox of venture capital is that you need to be early enough to capture the full value creation, but not so early that you’re funding pure science experiments. The sweet spot is identifying secular shifts that are technically feasible today but economically viable in 3-5 years. Most VCs get this timing wrong because they’re looking backward at what just worked rather than forward at what’s about to work.

The venture capital industry suffers from a peculiar form of professional risk management: It’s safer to be wrong with everyone else than right by yourself. This creates systematic opportunities for those willing to endure the discomfort of contrarian positioning.

The next great venture outcomes won’t come from perfecting the last cycle’s playbook. They’ll come from having the intellectual honesty to admit that most of our pattern recognition is sophisticated retrospective analysis, and the courage to bet on what we can’t yet fully understand.

The question isn’t whether you’ll be wrong about some of these contrarian bets - you will be. The question is whether you’ll be wrong about all of them. In a portfolio construction game, that’s a risk worth taking.

Because in five years, today’s consensus will look as quaint as yesterday’s. And the VCs who recognised this first will own tomorrow’s category-defining companies.

The views expressed are those of the author and do not necessarily reflect the views of any investment firm or portfolio company.