The Heuristics Behind Fund Choices, and Why Performance Reviews Are Where They Do the Most Damage
South African listed property returned roughly 46% in 2025 and around 84% from May 2024. Numbers like these do something predictable to human decision-making, and the effects show up most clearly not at the point of investment, but in the annual review meeting.
Performance reviews are behavioural pressure points. The client arrives with a story about the past year already formed, and the adviser's task is to work with that story without being captured by it. Understanding which heuristics generated the story is the difference between a review that improves decision quality and one that quietly degrades it.
The heuristics that drive fund selection
Fund choices are rarely made through expected-return arithmetic. They are made through mental shortcuts that are efficient most of the time and expensive at cyclical turning points.
Representativeness and extrapolation. Clients treat a short run of strong returns as representative of the asset's underlying character, projecting the recent past forward as if it were a stable property of the fund rather than a phase of a cycle (Tversky & Kahneman, 1974). A sector that re-rated sharply off depressed valuations looks, through this lens, like a sector that "delivers 40% years". The mechanism ignores the base-rate question of how often such years occur and what conditions produced this one.
Availability and salience. The funds that come to mind are the funds in the headlines and at the top of the performance tables. Availability is a memory phenomenon, not an evidence phenomenon (Tversky & Kahneman, 1973). Performance tables are, in effect, availability machines: they make the trailing winner cognitively cheap to retrieve and the disciplined middle of the distribution invisible.
Anchoring. Clients anchor on reference points that feel meaningful but carry no forward information. Some anchor on the price at which they nearly bought. Others anchor on an old drawdown, and for South African listed property the 2018 to 2020 period remains a powerful negative anchor that kept many investors out of the entire recovery. Category-level aversion built on one salient episode is anchoring dressed up as prudence.
Regret aversion and the action-omission asymmetry. People weight the anticipated pain of a decision that goes wrong more heavily than the pain of inaction that goes wrong, which biases them toward the status quo (Samuelson & Zeckhauser, 1988). Until, that is, the status quo itself becomes the salient error. A large missed run inverts the asymmetry: suddenly inaction is the regretted act, and the pressure to correct it becomes urgent precisely when the correction is least attractive.
Social proof. When colleagues, media and adviser peers are all discussing the same sector, participation starts to feel like the default and abstention like a position that requires justification. Herding compresses the perceived risk of a crowded trade at exactly the moment the actual risk is rising.
Client one: the investor who missed the run
The first conversation in review season is with the client who watched the rally from the sidelines. The behavioural state here is counterfactual regret, and it is genuinely uncomfortable. The danger is not the emotion itself but its most common resolution: late-cycle entry as regret repair.
The mechanism is straightforward. The client is not buying the asset's forward prospects; they are buying relief from the counterfactual. This is why missed-run conversations so often produce the classic returns-chasing pattern, where money arrives after the repriceable component of the return has already been paid to earlier holders.
The adviser's caveat is to decompose the return before discussing any allocation. A large portion of a re-rating rally is a one-off normalisation: discounts to net asset value closing, discount rates falling, sentiment recovering. That component is unrepeatable by construction. What remains is the earnings engine, which in the current property case is high-single to low-double-digit distribution growth at fair rather than distressed valuations. The honest framing is that the client is being offered a different asset from the one that produced the headline number, and the decision should be made on the asset that exists now.
Two conversational moves help. First, shift the evaluation from outcome to process: the decision not to hold a sector in 2024 may have been entirely reasonable given the information and the client's risk profile at the time, and outcome bias should not be allowed to retro-convict a sound process. Second, if exposure is genuinely appropriate for the client's strategy, structure the entry to neutralise regret dynamics, through phased implementation and pre-agreed sizing, so that the allocation reflects the plan rather than the emotion.
Client two: the investor who caught the run and now resists trimming
The second conversation is harder, because it involves asking a client to act against a position that has just rewarded them handsomely. Rebalancing after a strong run collides with at least three mechanisms simultaneously.
The house money effect describes the tendency to take greater risk with recent gains, which are mentally coded as the market's money rather than one's own (Thaler & Johnson, 1990). A client sitting on an outsized property weighting after a 46% year is psychologically playing with winnings, and the felt cost of leaving them exposed is low even when the portfolio-level risk is now materially concentrated.
Status quo bias and extrapolation reinforce each other. Selling a winner feels like an active bet against continuation, whereas holding feels like no decision at all, even though an unrebalanced portfolio is itself an active and growing overweight. The client's implicit model says the asset that just performed is the asset most likely to perform, when the return decomposition often says the opposite: the more of the move that came from re-rating, the less of it is available to repeat.
There is also an identity component. The position has become evidence of the client's (or the adviser's) good judgement, and trimming it can feel like retracting the claim. This is where the conversation quietly becomes about self-worth rather than portfolio construction, and advisers who miss that shift end up arguing valuations against emotions.
The caveats for the adviser are, first, to reframe rebalancing as harvesting rather than exiting. The client is not being asked to abandon a view; they are being asked to convert an unplanned overweight back into the deliberate weighting the strategy specified, banking part of the unrepeatable component while retaining exposure to the repeatable one. Second, lean on pre-commitment. A rebalancing rule agreed in the investment policy before the run makes the trim an execution of the client's own prior instruction rather than a fresh judgement call made under the influence of house money. Decision architecture set in a cold state is precisely what protects the client in the hot one.
The common thread
Both clients are responding to the same trailing number through different reference points, and both are at risk of converting a backward-looking emotion into a forward-looking allocation error. The one wants to buy the past; the other wants to keep holding it.
The adviser's structural task in review season is therefore consistent across both conversations: decompose returns into repeatable and unrepeatable components, evaluate decisions on process rather than outcome, and let pre-agreed rules carry the weight that willpower cannot. Performance reviews are not reporting exercises. They are the moments where the next cycle's behavioural errors are either installed or prevented.
References
Samuelson, W., & Zeckhauser, R. (1988). Status quo bias in decision making. Journal of Risk and Uncertainty, 1(1), 7–59.
Thaler, R. H., & Johnson, E. J. (1990). Gambling with the house money and trying to break even: The effects of prior outcomes on risky choice. Management Science, 36(6), 643–660.
Tversky, A., & Kahneman, D. (1973). Availability: A heuristic for judging frequency and probability. Cognitive Psychology, 5(2), 207–232.
Tversky, A., & Kahneman, D. (1974). Judgment under uncertainty: Heuristics and biases. Science, 185(4157), 1124–1131.