Why VCs Make "Bad" Bets on Purpose
- Aki Kakko
- 27 minutes ago
- 4 min read
In our previous articles, we explored the mathematics of the Power Law, the psychology of the Fear of Looking Stupid (FOLS), and the technological inevitability of the "Ghost in the Deal" (data-driven investing).
But there is one lingering question that math and technology cannot explain.
If the logic of "contrarian, data-driven farming" is so obvious, why do so many top-tier firms continue to herd into over-valued, hyped-up rounds that make no mathematical sense? Why do smart investors fight to write a check into an AI wrapper at a $500M valuation when they know the margins will collapse in six months? The answer lies in the operational model of the industry:
Venture Capitalists have two different customers, and they have conflicting interests.
The Entrepreneur: Needs capital to build a business.
The Limited Partner (LP): Provides the capital to the VC.
Most founders believe the VC’s job is to make a return on investment. In reality, for many General Partners (GPs), the primary job is Asset Accumulation. And sometimes, the best way to raise more assets is to make a bad investment.
The "Logo" Economy: Buying Marketing Materials
To understand the bad bet, you have to understand the Venture Capital business model. VCs are paid in two ways:
Carry (20%): A share of the profits (takes 7–10 years to realize).
Management Fees (2%): A guaranteed annual fee based on the total assets under management (paid immediately and recurring).
"Carry" is hypothetical wealth. "Management Fees" are guaranteed cash flow. If a VC wants to get rich today, they don't need to exit companies; they need to raise larger funds. To raise a larger fund, they need to prove to LPs (pension funds, endowments, family offices) that they have "Access."
This is where the "Bad Bet" becomes a "Rational Marketing Expense."
Let’s say a hot AI infrastructure company is raising at a nonsensical valuation. Every metric suggests it is overpriced. However, Sequoia and Benchmark are rumored to be in the deal. A mid-tier VC fights to get into this round. They write a $5M check. They know, mathematically, this money will likely return <1x.
But they aren't buying equity. They are buying the Logo.
By placing that logo on their website, they signal to LPs: "Look, we have access to the same deals as Sequoia." This signal allows them to raise a new $200M fund. That new fund generates $4M/year in guaranteed fees. The $5M loss on the startup is the customer acquisition cost (CAC) for the $4M/year revenue stream from LPs.
The Markup Game (TVPI vs. DPI)
In the absence of actual liquid returns (DPI - Distributed to Paid-In capital), VCs rely on "Paper Gains" (TVPI - Total Value to Paid-In capital) to judge their success. If a VC invests in a company at $10M, and 18 months later that company raises at $100M from a "dumb money" late-stage fund, the early VC can mark up their book value 10x.
They haven't returned a single dollar to their investors, but they can now show a chart that goes up and to the right. They use this chart to raise their next fund.
This creates a perverse incentive to invest in companies that are "fundable" rather than "sustainable." VCs will back a founder who is great at pitching the next round of investors, even if the underlying business mechanics are rotting, because the markup is what matters for the VC’s fundraising cycle.
The "Defensive" Check and Ego
Beyond the economics, there is the ego. In Silicon Valley, deal flow is status.
If a specific sector is hot (e.g., Generative AI or Crypto), a VC firm cannot afford to have zero exposure to it. If they go to their Annual General Meeting (AGM) and tell their LPs, "We passed on everything in AI because the valuations were irrational," they risk looking out of touch (FOLS). So, they make an "Index Bet." They spray money into the sector just to say they are "active in the space."
They are not betting on the company to win.
They are betting on not being blamed.
If the sector crashes, they can say, "Well, the whole market corrected." (Consensus Failure).If they missed the sector entirely and it booms, they look incompetent.
Therefore, the rational career move is to burn capital on a bad bet to buy insurance against reputational damage.
Deployment Pressure: The "Use It or Lose It" Trap
Finally, there is the mechanics of the fund clock. Most VC funds have a "Deployment Period" (usually 2–4 years). If they do not invest the capital within that window, they often have to return it or lose the ability to charge fees on it. When a fund reaches the end of its deployment period and still has $50M left, the partners are under immense pressure to "get the money out the door." This leads to the "Dumpster Fire" phenomenon: VCs loosening their standards and overpaying for mediocre deals simply because the money must be spent.
They are solving an internal accounting problem, not a market problem.
The Solution: Aligning the Ghost
This theater of logos, ego, and fee-stacking is exactly why the "Self-Service / Quant VC" model is the inevitable disruptor.
By stripping away the "Agency Problem"—the need to look good rather than be good—we return Venture Capital to its original purpose: Arbitrage.
We are buying mispriced optionality. The incumbents are buying overpriced status.
In the short term, the incumbents will look smarter because they have the shiny logos. In the long term, the math always wins. The "Ghost in the Deal" isn't just about analyzing startups; it's about exorcising the ghosts of ego from the investment process itself.
The "Power Law" Trilogy: First chapter: The Map is Not the Territory: Why Venture Capital Misunderstands the Power Law
Second chapter: Why the Smartest Money in VC Looks the Stupidest Third chapter: The Ghost in the Deal: Why Venture Capital Must Become Software
