← Back to Blog

Somewhere in your brokerage account sits a document called your risk profile. You filled it out in five minutes, answering questions like "how would you react if your portfolio dropped twenty percent?" You picked "I'd hold my position, it's temporary" — and you were being honest. It's just that the person answering had nothing falling at that moment.

Since then, that document has taken on a life of its own. It gets referenced when they match you with instruments and leverage. It decides what's allowed for you and what "doesn't fit your profile." And all of it is built on a calm person's five-minute opinion about how a panicking person will behave.

This article is about why you can't measure risk that way, how science measures it instead, and what to do if you want to know not your opinion of yourself, but your real number.

Your risk profile is your opinion of yourself

Look at how any risk-tolerance questionnaire works. It gives you hypothetical situations and asks you to pick a reaction. A twenty percent drawdown. A choice between a guaranteed small gain and a probable large one. Your investment horizon. You read the question, imagine yourself in the situation, and pick the answer that resembles you.

Three words in that description do all the work: "imagine," "resembles," and "pick." You're not in the situation — you're imagining it. And the one imagining it is a calm person who is risking nothing at that moment: no position on fire, no money ticking, nothing falling. He imagines the drawdown the way a person imagines a toothache from a description — approximately, without the sensation.

And then the drawdown actually happens. And the one at the keyboard in that moment is not at all the one who filled out the questionnaire. The same calm mind that wrote "I'd hold my position" won't even be at the wheel when the real risk shows up.

What science says about self-report

This isn't our observation — it's a long-standing problem in risk research. The reliability of self-reports remains a matter of debate, and studies show that the link between what a person says about their riskiness and how they actually take risks turns out to be weak.

The reason isn't that people lie. The reason is deeper and more interesting: the part of you that answers questions and the part that makes decisions under pressure are different systems. The first knows how you'd like to see yourself and can arrange words into a coherent portrait. The second tells you nothing — it just acts, fast, bodily, without explanations. Asking the first about the second is roughly like questioning a press secretary about what's happening behind a closed door: the answer will be polite, well-structured, and not necessarily true.

That's exactly why, at some point, researchers stopped asking — and started watching.

That's why they invented the balloon

In 2002, Carl Lejuez and colleagues published a method that came out of precisely this problem: there was no good behavioral measure of risk propensity — as opposed to questionnaires. And so the Balloon Analogue Risk Task was born.

It's deceptively simple. There's a balloon on the screen, and you can inflate it. Every pump of the pump adds money — a small amount that accumulates on that balloon. You can stop at any time and bank what you've built up. Or you can pump once more, because each next pump is worth more than the last. But the balloon has a limit, and it's unknown to you: pump one time more than it can take, and the balloon pops — and everything accumulated on it disappears.

Notice what this method doesn't have. You're never once asked whether you're a risk-taker. You're not invited to rate yourself on a scale. You just pump — and your behavior draws your profile on its own: where you stop, how close you get to the limit, what you do after one balloon has popped on you. The answer isn't told — it reveals itself.

What the balloon sees that the questionnaire can't

Lejuez's research delivered the detail this article is worth reading for. Riskiness on the balloon turned out to be linked to people's real-world risk-taking — and it reached a portion of that behavior that demographics and risk questionnaires alone left unexplained.

Let's put it carefully, because it's easy to overstate: the behavioral test doesn't "beat" the questionnaire. It sees what the questionnaire doesn't see at all. Your words about risk and your behavior in risk are two different sets of data, and the second can't be derived from the first.

And there's something the questionnaire can't measure in principle — what happens to you after a loss. When the balloon pops, a reaction appears: some people turn more cautious and pump the next balloons less; others start pumping harder, chasing it back. No question on paper can reach that, because reaching it requires that you actually lose something — game money, but real to you in that moment.

What this looks like in the terminal

Your "enough" point isn't an abstraction. It makes decisions for you every day, just under different names.

Position size: how much you're willing to stake when the setup looks very good. Leverage: where your line runs between "working for me" and "working against me." The moment of taking profit: that very "I'll hold a little longer" — pump after pump, while the balloon is intact — we covered this separately. And the most expensive place: the prop challenge, where a target-inflated size breaches the daily limit in a single trade, which we also wrote about.

In all these situations, the same value is at work. You don't choose it anew each time — you have just one, and it's stable. Only the balloons change.

The Bubble — learn your point

At NST, this method is called The Bubble. You pump — we calculate: where your stopping point is, how close to the limit you get, how your behavior changes after a popped balloon. The result can't be faked, because there is no "correct" number of pumps: too early and you lose what you earned, too late and you lose everything. There's only your way of finding that edge.

What comes out isn't a label of "risky" or "cautious," but a number. And, as with all our tests, that number has no bad values — only conditions that don't suit it. A high risk appetite with leverage and no limiters is one bad day erasing a month. The same appetite in a style where risk is structurally capped is simply your way of working. How a trait meets a style is a conversation we have separately.

What this changes

All risk advice is built the same way: "risk no more than one percent per trade." A good rule, I won't argue. But notice that it knows nothing about you — it's identical for the person who stops on the fifth pump and the one who goes to the fifteenth.

Knowing your number moves the conversation from someone else's rule to your own calculation. If you know your stopping point is late, and that after a loss you start pumping harder, you don't need reminders about discipline — you need limiters that don't ask your opinion in the moment: a pre-set size, an automatic stop, a daily limit. If your point is early, your problem probably isn't blowing up the account but leaving profit on the table, and the work to be done is somewhere else entirely.

The same rule can't be right for both. A map can.

Start with the free test

The Four Doors is free, no registration, a couple of minutes — it shows how you decide when the rules are hidden. The full NST map includes The Bubble: your stopping point, and what a popped balloon does to you.

The questionnaire asks who you think you are. The balloon shows who's pumping.

Take The Four Doors — free →

FAQ

What is risk tolerance? A person's willingness to accept uncertainty and possible losses in exchange for potential gain. It's usually assessed with a questionnaire, but it's important to separate two things: stated tolerance — what you say about yourself — and behavioral tolerance, how you act when the risk is real. They don't always match.

How accurate are brokers' risk-profile questionnaires? As a tool for regulatory compliance and rough client sorting, they work. As a measure of your real behavior under pressure, poorly: studies show only a weak link between self-reported riskiness and actual risk-taking behavior. The questionnaire captures your idea of yourself in a calm moment.

What is the Balloon Analogue Risk Task? A behavioral method published by Lejuez and colleagues in 2002: a participant inflates a virtual balloon, each pump adds money, but the balloon can pop and take the accumulated amount. It was created specifically as an alternative to self-report questionnaires and is widely used in risk research. NST's adaptation is called The Bubble.

How do I find out my real risk tolerance? Not through questions about yourself. Two things work: look at your own data — how your position size and behavior changed after a run of losses — or take a behavioral test where the risk is real in the moment. The second is faster and doesn't depend on how honestly you kept your records.