House Money Effect vs. Break-Even Effect: The Two Biases Behind Every Blown-Up Recovery

Two of the most useful findings in behavioral finance came out of the same 1990 study, and most people only ever hear about one of them. The house money effect is the tendency to take bigger risks with money you’ve just won, because part of you doesn’t fully count it as “real” yet. Its mirror image, the break-even effect, is the tendency to take bigger risks specifically when a bet offers a shot at getting back to exactly where you started after a loss. Different triggers, same outcome: both push you toward a riskier bet than you’d otherwise take, at exactly the moments — right after a win, right before recovering a loss — when you’re least equipped to evaluate it clearly.

The study behind both

Both effects come from a single paper: economists Richard Thaler and Eric Johnson, “Gambling with the House Money and Trying to Break Even: The Effects of Prior Outcomes on Risky Choice,” published in Management Science in 1990. Thaler and Johnson ran a series of experiments giving people a prior gain or loss, then offering them a follow-up bet, and measured how the prior outcome changed the choice.

Two clean patterns showed up. After a prior gain, people got more willing to take a risky follow-up bet than they’d been with no prior outcome at all — the house money effect. After a prior loss, people also got more willing to take risk, but only for bets that offered a real chance of getting back to exactly break-even; they stayed risk-averse for bets that didn’t offer that specific shape of outcome. Thaler and Johnson called that second pattern the break-even effect, and explained both through what they called quasi-hedonic editing: people mentally segregate a gain from their original stake, and integrate a new loss with a still-open prior loss instead of treating it as its own separate event. In plain terms — your brain keeps a running story about whether you’re “up” or “down” overall, and it will take on real risk trying to keep that story’s ending happy.

House money effect: what it looks like

The direct version: once you’re sitting on a gain, the money above your original stake starts to feel like it isn’t fully yours — like you’re playing with the casino’s money, not your own. That feeling is what lets you take a position size or a level of risk on that gain that you’d never have taken with your original capital.

This shows up constantly in crypto specifically, and there’s now real data on why. A 2023 NBER working paper by Darren Aiello, Scott Baker, Tetyana Balyuk, Marco Di Maggio, Mark Johnson, and Jason Kotter, using transaction-level data across millions of accounts, found that the marginal propensity to spend out of unrealized crypto gains is more than double the marginal propensity to spend out of unrealized equity gains of the same size. People treat a paper gain in a volatile, newer asset as noticeably more “spendable” than the same-sized paper gain in a stock they’ve held for years — which is exactly the mental accounting the house money effect predicts, just measured in real transaction data instead of a lab.

The trading version of the same instinct: you turn $2,000 into $9,000 on a single position, and instead of the position now being “up $7,000 that I should treat carefully,” it becomes “$2,000 that’s mine and $7,000 that’s free money I can afford to lose.” You size the next trade off the full $9,000, at a risk level you’d never have accepted with the original $2,000 — and when it goes wrong, you don’t just give back the gain, you’re frequently down more than your original stake, because the position size scaled with a number you weren’t actually willing to lose.

Break-even effect: what it looks like

The mirror pattern shows up after a loss, and it’s the mechanism behind most of what gets called “revenge trading” — the specific, narrower version where the goal isn’t just a good trade, it’s a trade sized and shaped to get back to exactly zero. (If the emotional tells of that pattern are more useful to you than the mechanism behind it, 5 Signs You’re Revenge Trading covers how to catch it in the moment.)

The break-even effect is narrower than general loss aversion: Thaler and Johnson found people don’t just get more risk-tolerant after any loss — they get more risk-tolerant specifically for bets shaped like “a real shot at getting back to even,” and stay cautious otherwise. That’s why doubling down on the exact position that hurt you feels so much more compelling than starting fresh with a smaller, unrelated trade of similar risk. Objectively they’re similar bets. Subjectively, only one of them promises to make the whole story go away.

Side by side

	House money effect	Break-even effect
Triggered by	A recent gain	A recent loss, specifically near recoverable
What changes	Risk-tolerance on the gain, not your original stake	Risk-tolerance on bets that could restore your starting point
How it feels	“This is free money, I can afford to lose it”	“One more move and this never happened”
Common form	Oversized position on a winner; not taking profit	Doubling down on the losing position; bigger size to “make it back”
Who it hits	Anyone recently up	Anyone recently down, closest to even
The actual risk	Giving back the gain and the original stake	Turning a recoverable loss into a much bigger one

Both rows describe the same mechanism from opposite sides: your brain isn’t pricing the next bet on its own merits, it’s pricing it against a story about where you “should” be.

Why recovery is the exact moment both effects converge

If you’ve clawed back from a loss and are approaching your original balance, you’re standing in the one spot where both effects can fire at once. The break-even effect has been pulling you toward bigger risk the whole way back — and the moment you cross the line and start posting a paper gain again, the house money effect picks up the exact same pull for a completely different reason. There’s no natural pause between “trying to get back to even” and “playing with the gain,” which is a large part of why so many recoveries in this niche don’t actually stick: the account gets back to zero and keeps going, on the same elevated risk appetite, straight through zero and into a new loss.

A worked example

Say you started with $5,000, drew down to $2,800 over a bad stretch, and have since clawed back to $4,900 — almost even. Two separate pulls are active right now, even though it doesn’t feel like two: the break-even effect is pushing you to take one more slightly oversized position to close the last $100 gap and “finish” the recovery, and if that position wins and pushes you into a genuine gain above $5,000, the house money effect will immediately start arguing that the amount above $5,000 is safe to risk more aggressively than the $5,000 base ever was.

The fix in both cases is the same question, asked at a fixed dollar figure instead of a feeling: would I take this exact position, at this exact size, if my account currently showed a totally unrelated number — say $4,900 with no memory of ever being at $5,000 or $2,800? If the honest answer is no, the size is being set by the story, not by the trade.

Three rules that work against both

Fix your position sizing to your account value, not your narrative. A rule like “never risk more than X% of current balance on one position” doesn’t care whether that balance includes a recent gain, a partial recovery, or neither. Strip the story out of the input.

Treat “almost even” as a signal to slow down, not speed up. The data says risk appetite rises as you approach break-even — so build in a deliberate pause exactly there: a smaller size, a day off, a second look at the setup, precisely when the pull to finish strong is at its highest.

Realize gains into something outside the story. Move a fixed portion of any real gain out of the account it was made in — cash, stablecoin off-exchange, a separate brokerage account — on a schedule, not a feeling. Money that’s been physically moved is much harder to mentally recategorize as still “in play.”

If the more general pattern behind all of this — losses registering harder than equivalent gains — is what you want to understand first, Loss Aversion Is Making You Worse at This covers the underlying wiring both of these effects run on.

FAQ

Is the house money effect just another name for FOMO? No — they’re triggered by different things. FOMO is driven by watching other people win and fearing you’ll be left out; it can hit you even with no gains of your own. The house money effect is driven specifically by your own recent gains — it only activates once you’re already up, and it changes how you treat that specific pool of money, not your appetite for a new trade in general. You can have one without the other, though in a hot market they often show up together.

I finally got back to break-even after a bad stretch. Am I safe now? Break-even is actually one of the more dangerous points to be at, not a finish line. Thaler and Johnson’s research found that people take on more risk specifically when a bet offers a shot at getting back to exactly even — the closer you get, the stronger the pull to force the rest of the way with a bigger, faster move instead of just stopping. Treat crossing back to zero as a moment to reassess your position sizing, not a green light to keep pushing with the same intensity that got you there.

How do I stop treating gains as “not real money” I can afford to risk? Move it. The house money effect survives almost entirely on money staying inside the account where it was won, still showing up as a number on a screen next to your original stake. The moment a portion of a gain is withdrawn, converted to stablecoin and sent off-exchange, or transferred to a separate account you don’t day-to-day trade from, it stops feeling like house money and starts feeling like your money — because at that point, functionally, it is. A standing rule (e.g., move 25% of any position once it’s up 100%) does this automatically, before the feeling has a chance to argue you out of it.

Does this apply the same way to stocks as it does to crypto? The underlying mechanism is identical — it comes from how people mentally categorize money, not from the asset class. But crypto’s volatility and 24/7 markets compress the cycle: you can go from a painful loss to house-money overconfidence to a bigger loss in the same week, where the equivalent swing in a stock portfolio might take months. That’s not a different bias, just a faster clock on the same one.