Zero-Risk Bias

The zero-risk bias is the preference for reducing a small risk to zero over reducing a large risk by the same amount—even when expected value favours the latter. An investor will pay heavily to eliminate a 1 per cent chance of ruin, then fail to reduce a 40 per cent chance of significant loss by the same unit. This bias explains excessive insurance, hedging, and portfolio fragmentation, and it costs investors significantly in foregone returns.

The discontinuity of zero

Zero has a special place in the human mind. It is categorically different from one-tenth or one-hundredth. A risk of zero is “safe”; a risk of 0.1 per cent is still “unsafe”. This mental boundary is not rational. Objectively, the difference between 0 and 0.1 per cent is the same as the difference between 0.1 and 0.2 per cent. But psychologically, it feels entirely different. The discontinuity drives zero-risk bias.

In a classic 1978 experiment, Fischhoff and colleagues asked people to allocate money between reducing risk in two scenarios. In the first scenario, Scenario A had a 1 per cent risk of injury, and Scenario B had a 10 per cent risk. Subjects were asked to divide their budget between reducing A and reducing B. Most allocated disproportionately to A, often enough to reduce it to zero, even when the same budget could reduce B more significantly. Why? Because eliminating A—going from 1 per cent to 0 per cent—felt like a categorical win. Reducing B from 10 per cent to 8 per cent felt like tinkering. The math said allocate to B; psychology said eliminate A.

In insurance and hedging

Zero-risk bias explains a large share of insurance and financial hedging behaviour. Consider auto-insurance. The true statistical expected loss for an average driver in a given year is modest, often hundreds of dollars. Yet people pay premiums far exceeding expected loss to reduce the risk of catastrophic loss to (nearly) zero. This is not irrational if catastrophe creates real hardship. But it is often taken to extremes: people over-insure against tiny risks (losing a phone, a checked bag) and under-insure against likely ones (extended income loss).

In financial hedging, zero-risk bias drives excessive spending on tail risk protection. An investor might pay $20,000 annually for put options that protect against a 2–5 per cent portfolio decline—a decline with low historical probability and a cost far exceeding expected benefit. Meanwhile, the same investor accepts unhedged exposure to a 20–30 per cent drawdown risk that is far more probable. The hedging is not about expected value; it is about the psychological comfort of touching zero risk in the tail.

Portfolio concentration and fragmentation

Zero-risk bias can produce paradoxical portfolio errors in both directions. On one side, an investor might insist on a “safe” core—all Treasuries, all cash—to achieve zero stock market risk, sacrificing real return to eliminate a probabilistic loss. On the other side, the same investor might take wild concentration risk in a single stock to chase high return, tolerating asymmetric ruin scenarios that rational diversification would avoid.

Both errors stem from the same bias: the investor achieves zero risk in one dimension (stock exposure, or concentration), so they feel safe, even though the other dimension contains larger risks than intelligent asset allocation would tolerate. A rational portfolio trades off risks efficiently, accepting non-zero risk in multiple dimensions if the tradeoff improves expected return per unit of total risk. Zero-risk bias prevents this tradeoff.

The certainty effect

The bias extends beyond zero risk to any “certainty.” Economists call this the certainty effect. An outcome that is certain (100 per cent probability) is overweighted relative to a probabilistic outcome, even if the probabilistic outcome has higher expected value. In an investment context, a guarantee of 3 per cent annual return might feel more attractive than a 50/50 bet on 0 or 6 per cent, even though the latter has the same expected value (3 per cent) but higher variance. The certainty of the 3 per cent eliminates uncertainty, and that elimination is valued beyond its economic contribution.

This explains why annuities are often unattractive to buyers despite insurance companies’ operational edge: the certainty of a fixed payout appeals more than a variable portfolio of higher expected value. It also explains reluctance to shift from bonds to equities even when historical returns favour equities—the certainty of bond coupons has psychological weight beyond the numbers.

Measuring the cost

The direct cost of zero-risk bias is easy to estimate in insurance and hedging: the premium paid above expected loss. If you pay $100 to insure against a $30 expected loss, the cost is $70. In broader portfolio allocation, the cost is opportunity: foregone returns from avoiding stock exposure to achieve cash “safety,” or foregone diversification by concentrating in one “safe” holding.

A less visible cost is fragmentation and complexity. To achieve zero risk in one dimension while accepting risk in another, investors layer on hedges, separate “buckets,” and rules. A portfolio with a “safe” core, a “growth” allocation, a “hedge,” and an “opportunistic” sleeve becomes complex and expensive to manage. A simpler, fully diversified portfolio with efficient risk tradeoffs would likely perform better and be easier to monitor.

When zero-risk bias is rational

There are contexts in which near-zero risk is genuinely justified. If a specific risk creates systemic risk or ruin—the possibility of losing essential capital or becoming insolvent—then paying to eliminate that risk can be rational even at high cost. A business paying insurance to avoid catastrophic liability is hedging genuine ruin; an individual homeowners insurance is protecting against loss of irreplaceable shelter. These are not pure psychology; they are rational risk management.

The bias appears when the same logic is applied to small, probabilistic, non-systemic risks. Paying heavily to reduce equity volatility from 15 per cent to 14 per cent is not rational. Accepting high concentration risk to avoid broad market risk is not rational.

Checking the bias

A practical discipline: insist on quantifying the cost per unit of risk reduction. If you are considering a hedge or insurance product, calculate the cost divided by the statistical expected loss reduction. If it is above, say, 1.5 to 2 times the actuarial expected benefit, the price is inflated by zero-risk bias. Similarly, if you are considering a portfolio allocation to achieve “safety,” quantify the return foregone per unit of risk reduced. A portfolio that cuts equity exposure from 60 per cent to 20 per cent to reduce volatility might sacrifice 2–3 per cent of expected return annually to cut volatility by 20–30 per cent. Is that tradeoff rational given your goals? Often it is not—the certainty was simply more attractive.

Another approach: make risk-reduction decisions independent. Rather than deciding “I want zero equity risk” or “I want zero drawdown risk,” decide the target level of portfolio volatility or drawdown you can tolerate, then allocate asset classes and hedges to hit that target most efficiently. This removes the psychological appeal of zero and forces comparison of cost per unit of risk reduction across options.

See also