The Cognitive Biases That Quietly Run Your Cap-Table Decisions

Cognitive biases are not personality flaws; they are architectural features of human judgment. In cap-table decisions — valuations, dilution thresholds, follow-on commitments, co-investor selection — those features produce predictable, measurable errors. The good news is that predictable errors can be anticipated and partially corrected. The bad news is that awareness alone rarely does it.

Why the cap table is a bias amplifier

Most founders treat cap-table decisions as financial problems: model the dilution, stress-test the waterfall, negotiate the liquidation preference. That framing is correct but incomplete. Every number on a term sheet is also a psychological event. It arrives under time pressure, in conditions of genuine uncertainty, and with enormous personal stakes attached — precisely the conditions under which cognitive shortcuts dominate deliberate reasoning.

In their landmark 1974 paper in Science, Amos Tversky and Daniel Kahneman identified three cognitive heuristics — representativeness, availability, and anchoring — that “lead to systematic and predictable errors” in judgment under uncertainty.¹ Kahneman later formalised the underlying mechanism in Thinking, Fast and Slow (2011): System 1, the fast, automatic, intuitive mode, handles the overwhelming majority of decisions; System 2, the slow, deliberate, analytical mode, is called in only when System 1 cannot produce a quick answer.² Cap-table negotiations feel like System 2 work. They are, in practice, largely System 1 work dressed in spreadsheets.

The five biases examined below are not a complete taxonomy. They are the ones with the clearest evidence base and the most direct line to equity decisions. Each section names the bias, explains the mechanism, maps it to a specific cap-table moment, and offers a structural — not merely attitudinal — countermeasure.

Movement I: Anchoring — the first number owns the room

Anchoring is the tendency to rely disproportionately on the first piece of information encountered when making a judgment. All subsequent estimates are adjusted from that anchor, but the adjustments are systematically insufficient — even when the anchor is known to be arbitrary.³ Tversky and Kahneman demonstrated this in their original roulette-wheel experiment: participants who saw the wheel stop on 65 guessed, on average, that 45% of UN member countries were from Africa; those who saw it stop on 10 guessed 25% — a 20-percentage-point gap produced by a number everyone knew was random.⁴

In startup valuation, the mechanism is identical. Investors, including venture capitalists and private equity firms, exhibit anchoring bias when negotiating startup valuations: initial price discussions and previous funding rounds often serve as reference points, leading to valuation stickiness where investors fail to fully incorporate new information into their assessments.⁵ This is particularly consequential in rapidly changing industries, where an early-round anchor may be wholly obsolete by the time a Series B is being priced.

Research consistently shows that negotiators who make the first offer often achieve better economic outcomes — provided the anchor is credible.⁶ The implication for founders is direct: if you have done the work, go first. If you have not, the investor’s opening number will govern the conversation whether you intend it to or not.

Countermeasure: Form an independent valuation estimate — using comparable transactions, revenue multiples, and a discounted cash-flow sanity check — before any number is introduced into the room. Seeking multiple reference points before making financial decisions is one of the most reliable anchoring antidotes identified in the literature.⁷

Movement II: Loss aversion — the equity you hold feels worth more than it is

Loss aversion, a central component of Kahneman and Tversky’s Prospect Theory (1979), describes the asymmetric weighting of losses and gains: empirically, losses tend to be treated as if they were roughly twice as large as an equivalent gain.⁸ The theory describes losses and gains measured from a reference point — not from absolute wealth — which means that the pain of giving up 5% of your company is evaluated against your current ownership, not against the value of what you receive in exchange.⁹

This is the psychological engine behind founder resistance to dilution. A founder who owns 60% of a company worth $5 million evaluates a Series A that would reduce her stake to 45% of a company worth $20 million as a loss of 15 percentage points — not as a $4 million gain in absolute equity value. The percentage feels like the asset. The absolute value is the actual asset. Loss aversion makes it systematically difficult to hold both truths simultaneously.

The same mechanism operates on the investor side. Loss aversion plays a significant role in how investors perceive and react to potential losses; the pain of losing often outweighs the joy of an equivalent gain, which can skew decision-making — including decisions about whether to lead a bridge round in a struggling portfolio company rather than accept a write-down.¹⁰

The endowment effect — closely related to loss aversion — compounds this. In psychology and behavioral economics, the endowment effect is the finding that people are more likely to retain an object they own than acquire that same object when they do not own it.¹¹ Founders who built their cap table from a blank page are among the most vulnerable: the people most vulnerable to over-persistence are often the people who built the thing from scratch.¹²

Countermeasure: Reframe dilution decisions in absolute value terms before evaluating them in percentage terms. Run the post-money ownership calculation first, then the percentage change. The order of operations matters because it determines which number becomes the psychological reference point.

Movement III: Overconfidence — the forecast that never fails (until it does)

Overconfidence is the tendency to overestimate one’s own skills, knowledge, or the likelihood of a favourable outcome relative to objective base rates.¹³ In entrepreneurship, empirical studies have established overconfidence as “a defining trait of entrepreneurs — practically a caricature.”¹⁴ That is not a criticism of founders; it is a description of a selection effect. The willingness to start a company in the face of high failure rates requires a degree of optimism that, by definition, departs from the actuarial evidence.

The problem is that the same optimism that initiates a venture also distorts the terms on which capital is raised. Overconfidence and optimism inflate targets and hinder adaptability, while planning fallacy and anchoring cause misaligned strategies.¹⁵ A founder who genuinely believes her three-year revenue projection is conservative will accept a valuation ratchet or a milestone-based tranche structure that she would reject if she held a more calibrated view of execution risk.

The VC side is not immune. A study examining 53 venture capitalists found that their investment decisions are biased by overconfidence, and that this overconfidence increases with more information — meaning that a more thorough information-gathering process can paradoxically produce greater, not lesser, overconfidence.¹⁶ Separately, research found that 96% of VC managers in one sample exhibited considerable overconfidence, defined as an overestimation of the likelihood that a funded company will succeed.¹⁷

Countermeasure: Apply the outside view before the inside view. Before building a bottom-up forecast, look at the base rate for companies at your stage, sector, and geography. Counterfactual thinking — explicitly imagining scenarios where current assumptions do not hold — is one of the most effective tools for reducing overconfidence in investment decisions.¹⁸

Movement IV: Confirmation bias — due diligence as a closing argument

Confirmation bias is the tendency to seek out, interpret, and recall information in a way that confirms pre-existing beliefs, while discounting evidence that contradicts them.¹⁹ In the financial sector, investors rely more strongly on and spend more time focusing on information that confirms their preceding knowledge or opinion about a given investment opportunity.²⁰

The structural problem in venture due diligence is that confirmation bias is baked into the process. Exploratory due diligence — the phase conducted early in the investment process — is designed to produce a go/no-go decision. Once a term sheet is signed, confirmatory due diligence verifies information already obtained.²¹ The sequence means that by the time the most rigorous scrutiny occurs, the investor has already committed emotionally and reputationally to the deal. Due diligence, at that point, is less an investigation than a closing argument.

The Theranos case is the canonical illustration. When prominent investors were presented with fact-based evidence that the company was fraudulent, they refused to believe it — a textbook confirmation bias failure in which social proximity and prior conviction overrode contradictory data.²² The dynamic is not unique to high-profile frauds; it operates in ordinary due diligence on ordinary deals, every day.

For founders, confirmation bias manifests differently but with equal cost: it produces pitch decks that present only the evidence supporting the bull case, and it shapes which investor questions get thorough answers versus which ones get redirected. A founder who has convinced herself that her market-size estimate is correct will unconsciously filter the data she presents to support that conviction.

Countermeasure: Assign a formal devil’s advocate role in any investment committee or founder board discussion. The pre-mortem technique — developed by cognitive psychologist Gary Klein and formalised in a 2007 Harvard Business Review article — asks a team to imagine that a plan has already failed and work backward to identify the reasons.²³ Gary Klein found it improves the ability to identify reasons for future failure by about 30 percent, because it gives people permission to voice the doubts that optimism and group dynamics usually suppress.²⁴

Movement V: The sunk-cost fallacy — past investment is not a reason to continue

The sunk-cost fallacy is the tendency to continue investing in a decision based on the cumulative prior investment — time, money, resources — rather than the present and future potential of that decision.²⁵ Sunk costs are, by definition, unrecoverable. They should be irrelevant to forward-looking decisions. They are not.

In venture capital staging, the evidence is direct. A study based on 30,602 investment decisions about US-based portfolio companies from 2009 to 2019 found that both the amount of capital previously invested and the intensity of monitoring significantly increase the probability of continued investment — underscoring the sunk-cost fallacy’s role in VC follow-on decisions.²⁶ VCs who have invested more, and monitored more closely, are more likely to fund the next round — independent of whether the underlying business merits it.

For founders, the sunk-cost fallacy manifests as an inability to kill features, abandon go-to-market strategies, or accept that a co-founder relationship has structurally failed. Grit is sticking with a vision when the going gets tough, but the fundamental hypothesis remains valid. Sunk-cost fallacy is sticking with a specific tactic or product when the hypothesis has been disproven, simply because you paid for it.²⁷ The distinction is not philosophical; it is operational. A founder who cannot make it will burn runway defending a position that the market has already rejected.

The identity dimension compounds the financial one. A founder who has built a startup has an identity-level stake in the project succeeding; public commitments make this worse, because visible reversal looks like failure while quiet persistence looks like resilience.²⁸ This is why the most dangerous sunk-cost decisions are made by the most committed founders — not the least committed ones.

Countermeasure: Establish kill criteria before a project begins, not after it is in trouble. A pre-agreed threshold — a specific metric, a specific date, a specific market signal — removes the decision from the domain of identity and places it in the domain of process. Formally recording how past decisions were made at the time of the decision, rather than reconstructing them from memory, is a further structural safeguard against sunk-cost rationalisation.²⁹

The interaction effect: when biases compound

These five biases rarely operate in isolation. They interact, and the interactions are multiplicative rather than additive. A founder who is anchored to a high valuation from a previous round (anchoring), is reluctant to accept dilution (loss aversion), believes her projections are conservative (overconfidence), has filtered her data room to support the bull case (confirmation bias), and has already spent eighteen months on the current product (sunk-cost fallacy) is not facing five separate problems. She is facing one compounded problem: a decision-making environment in which every cognitive force is pushing in the same direction — toward a deal that feels right and may be wrong.

The same compound operates on the investor side. A VC who anchored to a high entry valuation, is loss-averse about a write-down, is overconfident in her sector expertise, has confirmed her thesis through selective due diligence, and has already deployed significant capital into the company will find it nearly impossible to make a clear-eyed follow-on decision. The bias blind spot — the finding that people who know more about biases are no less susceptible to them — means that awareness is necessary but not sufficient.³⁰ The fix is structural.

What this means

The cap table is, ultimately, a record of decisions made under uncertainty. Some of those decisions will have been made well; others will have been quietly governed by anchors, loss aversion, overconfidence, confirmation, and sunk costs that the decision-maker never consciously registered. The goal is not to become a bias-free decision-maker — no such person exists. The goal is to build a decision environment in which the most consequential choices are made slowly enough, and with enough structural friction, that System 2 gets a seat at the table before the term sheet is signed.

Founders who do this are not more cautious. They are more precise. And in equity decisions, precision compounds.

Business Growth Accelerator (a FounderWise brand) works with founders and operators on the decision frameworks that govern growth — including the equity decisions that shape everything downstream. If this piece raised questions about your own cap-table process, start a conversation.