House Money Effect

The house money effect is the increased willingness to take risk after a recent gain. The term borrows the gambler's sense that winnings are "the house's money" rather than one's own, and it is the name Richard Thaler and Eric Johnson used for the finding in a 1990 paper in Management Science. They began from the point that "normative theory implores decision makers to only consider incremental outcomes," while "real decision makers are influenced by prior outcomes," and they reported data from real money experiments supporting "a 'house money effect' (increased risk seeking in the presence of a prior gain)." The same paper described the opposite situation as the break-even effect: "in the presence of prior losses, outcomes which offer a chance to break even are especially attractive."

Why prior outcomes should not matter, and why they do. In standard decision theory, a choice between two gambles depends on the outcomes those gambles could produce from here. Whether the decision maker won or lost an hour ago is already in their wealth and should play no further part. Thaler and Johnson's question was how real decision makers combine a prior outcome with a new choice, and they proposed an "editing rule" describing how decision makers frame that combination. The house money label captures the intuition behind the result: money recently won does not yet feel like one's own, so risking some of it feels less like risking a loss.

The break-even effect. After a loss, ordinary risks look unappealing, but an option that offers a chance of getting all the way back to even draws people in, even at poor odds. In investing this is the pull to add to a losing position, or to switch into something riskier, to "make it back". It overlaps with the disposition effect, the tendency to hold losers too long, and with the sunk-cost fallacy, but the distinguishing feature is the attraction to the specific outcome of breaking even.

How it differs from mental accounting. Mental accounting is the broader habit of sorting money into categories by source or purpose and treating each category differently, so that a tax refund or a gift can be treated differently from salary. The house money effect is a specific, dynamic consequence: a recent gain changes how much risk a person is willing to take with the next decision. Both rest on the same idea that a dollar's history changes how it is treated, which is the idea the economist's reminder that money is fungible pushes against.

What the evidence does and does not show. The original findings come from controlled experiments with real money, not from studies of retirement accounts or real estate. They establish that prior outcomes shift risk-taking in a predictable direction; they do not provide a size for how much a particular investor will change their allocation after a good year.

The effect works through how a gain or loss is framed. A recent gain becomes a cushion in the person's mind, so a new risk is measured against the cushion rather than against total wealth. A recent loss becomes a hole, so an option that could fill it looks better than its odds deserve.

Consider an example with invented figures. Silas bought $50,000 of his employer's stock, and it is now worth $100,000. He mentally files the $50,000 gain as "the house's money" and plans to put it into short-dated call options that he estimates could triple or expire worthless. Measured against total wealth, his position is simple: he owns $100,000 of stock, and the plan would move half of it into a bet where he expects to lose everything more often than not. If he would not take $50,000 from his savings account to make the same bet, the gain's history is the only thing making it look acceptable. The same reasoning applies in reverse. If the stock had fallen to $30,000, a plan to move the remainder into a riskier position "to get back to $50,000" would be the break-even effect, and the right test would again be whether the bet makes sense for $30,000 of wealth, not whether it could repair the loss.

Pros

• Naming the effect gives a simple test for any decision after a big gain or loss: would the same choice make sense with money that arrived some other way?
• It identifies the two moments when the pull is strongest, right after a large gain and right after a large loss, so a person can plan for them.
• Recognizing the break-even pull can stop a loss from turning into a larger one.

Cons

• The original evidence is experimental and does not say how large the effect is in real portfolios.
• Not every change in risk-taking after a gain is a bias: a larger portfolio can rationally support more risk in dollar terms.
• The effect is easy to rationalize as "playing with profits", which is exactly what makes it hard to notice.