Prospect Theory describes how people choose between probabilistic alternatives involving risk. But the original model was designed for simple gambles, like a coin flip. Real-world decisions are far messier, involving many possible outcomes with varying probabilities. This is where Cumulative Prospect Theory (CPT) comes in.

CPT models decision-making in two stages. First is the "editing" phase, where a person simplifies the available options into a more manageable form. They might round probabilities, discard extremely unlikely outcomes, or frame outcomes relative to a reference point.

Second is the "evaluation" phase. Here, the edited prospects are evaluated using two core functions: a value function (like the one in original Prospect Theory) and a probability weighting function, which is where things get really interesting.

Expected Utility Theory assumes we treat probability linearly. A 20% chance of something happening is perceived as exactly twice as likely as a 10% chance. CPT shows this isn't true. We systematically distort probabilities in our minds. The probability weighting function captures this distortion.

Generally, we overweight very small probabilities and underweight moderate and high probabilities. This gives the function an inverse S-shape. The jump from a 0% chance to a 1% chance feels huge—it’s the difference between impossible and possible. But the jump from a 50% chance to a 51% chance feels negligible.

This function has a different shape for gains than for losses, reflecting our different attitudes toward risk in each domain. Research shows these weighting functions can vary across demographic groups and cultures, influencing everything from insurance markets to public policy decisions.

A key challenge for both Expected Utility Theory and basic Prospect Theory is that they assume a stable reference point. But our reference point for what counts as a gain or a loss is fluid; it's shaped by recent experiences. This is called reference point endogeneity—our frame of reference is determined from within the situation, not outside of it.

One powerful example of this is "Gain Overhang." An investor who has recently experienced significant gains on a stock may update their reference point. The new, higher price becomes the baseline. Instead of feeling wealthy, they feel protective of their paper profits and become more risk-averse, fearing a drop back to the original price.

Conversely, after a series of losses, an investor's reference point may shift downwards. To get back to their original breakeven point, they might take on excessive risk, a behaviour known as "disposition effect"—holding onto losing investments for too long in the hope they will recover. This is driven by diminishing sensitivity in the loss domain: each additional dollar lost hurts less than the one before it, encouraging risk-seeking behaviour to erase the loss.

So, when should we use CPT over the more traditional Expected Utility Theory (EUT)?

EUT describes how a perfectly rational agent should behave to maximise wealth. It’s a normative model. CPT is a descriptive model—it describes how people actually behave, with all our biases and mental shortcuts. While EUT fails to explain many real-world market behaviours, CPT provides a powerful framework for understanding the psychology that drives asset prices away from their fundamental values.

Time to test your understanding of these advanced concepts.

Understanding these nuances helps bridge the gap between idealized economic models and the often-unpredictable reality of human decision-making in financial markets.