Menlo’s Venky Ganesan warns venture’s valuation loop can break

When valuation becomes the signal

Ganesan does not think venture capital is broken. Some companies in Menlo Ventures’ portfolio are growing faster than anything he has seen in his career. At the same time, companies with no revenue, no product and founder teams small enough to fit in a conference room are raising billions at valuations between $10 billion and $50 billion.

Both things can be true. The mistake is treating them as if they can be judged by the same model.

Ganesan connects the problem to reflexivity, a concept George Soros developed from the 1980s onward. Weather does not respond to what people think about it. Markets do: prices change beliefs, and beliefs change prices. While the loop continues, it can look like ordinary progress.

Ganesan uses Anthropic as an example, noting that Menlo Ventures invests in the company. Investors watched the frontier lab’s valuation rise from $4 billion to $18 billion, then to $60 billion, $180 billion, $380 billion and nearly $1 trillion. They began treating that path as the normal trajectory for a new lab and valuing other companies against it, rather than against what those companies had built.

The latest round stopped being an output of analysis and became an input into the next valuation. Investors were no longer discounting future cash flows; they were discounting the valuation from the last round.

Two examples illustrate how quickly that loop can move:

Thinking Machines is not a perfect example of a zero-revenue company. The same reports put its annualized revenue at between $100 million and several hundred million dollars. They also say the proposed valuation is below the $50 billion the company was reportedly seeking at the end of last year.

That resistance is important. Even a reflexive market cannot move upward without friction.

Why investors keep dancing

The second idea comes from Chuck Prince, Citigroup’s chief executive in July 2007. He told the Financial Times that while the music was playing, his bank had to keep dancing. Four months later, he was out of a job.

Ganesan’s point is not that investors are irrational for staying in the market. It is that they have different reasons to continue:

The first group may underestimate the money it has at risk. The second may turn a missed opportunity into a realized loss. Reflexivity feeds both instincts, leaving every participant with a reason to keep dancing.

Position size is the real edge

The conventional venture strategy is to pick better companies: find the right new laboratory and make money from it. Ganesan considers that framing a trap. If price has become a signal, an investor can choose the right company and still be wrong about its value by a factor of ten.

His alternative is less glamorous and more consequential: position sizing.

The best public-market investors, Ganesan argues, pay as much attention to how much of an asset they own as to which asset they choose. He expects the strongest venture firms of the next decade to bring the same discipline to portfolio construction.

Each partner should be able to answer two questions:

I think this is the more useful part of Ganesan’s argument. A debate over whether a particular AI laboratory deserves a $40 billion or $50 billion valuation can become a distraction from the exposure created by owning too much of it. The number that matters to a fund is not only the valuation on paper, but how much damage a reset would do to the whole portfolio.

The missing discipline: getting out

For general partners, the practical response is a stress test that most models lack:

For limited partners, the exit decision matters as much as the entry. Trace Cohen told Ganesan that the hardest task in venture capital is knowing when you have already won.

An investor who entered at a $50 million valuation and now holds an interest valued at roughly $10 billion does not need to decide whether $10 billion is cheap. The immediate question is how much of the gain to keep.

Ganesan agreed, arguing that venture capital pays too little attention to exits. Limited partners who care about distributions, not just marked-up assets, should already be demanding that discipline from fund managers.

The announcement is quiet about one thing: who has the authority to sell, and on what terms, when everyone still wants to believe the next round will be higher. A portfolio can be carefully sized in theory and still remain trapped if its liquidity plan is only a valuation update.

A higher valuation is also a liability

For founders, the conclusion runs in the opposite direction. A latest-round valuation is not only an asset. It is an obligation that raises the bar for the next round and leaves less room for mistakes.

Founders who use the current market to build revenue and durable products will have something to stand on when the music stops. Those who treat the valuation itself as evidence of progress will have less protection.

Ganesan’s final warning is also his advice: the music will stop because it always does, so investors who keep dancing need to know where the chairs are. Prince never claimed the music would play forever. He only said he could not sit while it was playing.

The difference between those outcomes is not necessarily better deal selection. It may be deciding in advance how many chairs the portfolio needs.