Approach-avoidance framing replaces risk-vs-attraction in investment decisions
Approach-avoidance framing replaces risk-vs-attraction in investment decisions
One-line summary: Josh Steiner's reframe — most decisions involve both a desire to approach and a desire to avoid the same object, so the operative question for an investor isn't "what's the risk vs the upside" but "what do you have to believe to be true in order to be attracted to this?"
What it is
Steiner argues that investing analysts who frame an opportunity as "risks vs. attractions" miss the underlying psychological structure of the decision. The pattern in most real decisions is approach-avoidance — the same thing draws you in and makes you anxious, not because there are separate good-and-bad attributes but because the desirable feature is the source of the anxiety. The classic example: a snake at a kid's birthday party feels different from a pit of snakes that isn't in front of you, even though both are conceptually risky, because only the first activates the approach side of the dynamic.
The framework's actionable substitution: instead of cataloging risks alongside attractions on parallel lists, ask "what do you have to believe to be true in order to be attracted to this?" This converts the decision from a tradeoff format into an explicit hypothesis format — the analyst now owns a falsifiable belief, not a vague balance.
Why it matters to psychology
Connects to behavioral economics' approach-avoidance literature (a classic conflict type in motivation research, from Lewin's field theory and Miller's conflict-resolution work). Steiner's contribution is the operational substitution for practitioners — replacing the risk-vs-attractions cognitive frame (which preserves the conflict in unresolved form) with a single-statement hypothesis format (which forces a falsifiable commitment).
Also intersects with judgment-under-uncertainty literature: the substitution shifts the decision-maker from System 1 affect-balancing toward System 2 hypothesis-articulation, which has different error modes (overconfidence in hypothesis specificity vs. anxiety-driven under-action).
Evidence
- josh-steiner in 2026-05-18-podcast-capital-allocators-making-mistakes-josh-steiner-ep-502: "Most decisions that we make involve both a desire to approach something and a desire to avoid that very thing itself. We're not anxious about things that we have no desire to do. No one feels particularly anxious about falling into a pit full of vipers, because you don't really face that as a possibility. You might well feel anxious if you're at your child's birthday party and they're going off and they're passing pets around and one of which is a giant snake."
- josh-steiner in same: "In investing, there's a recognition that almost every compelling investment has an approach avoidance capacity. It's that willingness to understand what is compelling about it and what is making us fearful... that's why when we're looking at investment I don't like risks and attractions. These things aren't diametrically opposed. A more effective tool is to say what do you have to believe to be true in order to be attracted to this?"
- josh-steiner in same: "We are trying to be attracted to be investors. We have to put money to work. Just thinking about the risk is ineffective."
Implications
- For investment committees (where Steiner has served on Yale's IC nearly a decade): an explicit "what do you have to believe" statement on every decision converts post-hoc rationalization into pre-commitment, making the eventual review more honest.
- For research-process design: this is a specific operational substitution, not a high-level frame — it can be enforced as a one-line addition to investment-memo templates.
- For mistake post-mortems (the broader topic of the source episode): the approach-avoidance frame predicts that mistakes cluster in cases where the analyst suppressed one side of the conflict rather than articulating the underlying belief.
Open questions
- Does this framework produce different outcomes than the standard pre-mortem / red-team exercise, or is it isomorphic with a single-prompt format?
- Is the falsifiable-hypothesis framing more durable against motivated reasoning, or does it just relocate the bias into the hypothesis articulation?