J.P. Morgan runs a HireVue video interview for its 2027 Commercial and Investment Bank Markets Summer Analyst Program in Hong Kong, and it cuts a large share of applicants before a trader watches a single clip. The people who get through have usually rehearsed the real questions. This one is a single question interview, Question 1 of 1, with close to three minutes of prep and two minutes to answer on camera. Do not let the friendly wording fool you. It is a probability test in a trading costume, and it is built to catch people who trust their gut instead of the numbers.
Here is the question, word for word:
Scenario: You are an intern on a macro trading desk. The team is studying an indicator that tries to predict whether the stock market will go up or down tomorrow.
Historically, the indicator works as follows:
• When the market goes up, the indicator correctly signals "UP" 70% of the time.
• When the market goes down, the indicator incorrectly signals "UP" 20% of the time.
Historically, the market goes:
• Up 60% of days
• Down 40% of days
Today the indicator flashes "UP."
Your trader asks you to think about what that signal really means.
Question: Given the statistics above, is the probability that the market will go up tomorrow equal to 70% of the time? What is the probability if not? Explain your reasoning
The trap is the number 70. It is sitting right there, it matches the signal, and it feels like the answer. It is not. The 70 percent is how often the indicator fires "UP" when the market actually goes up. The question asks the reverse: given that it just fired "UP," how likely is the market to rise. Those are two different quantities, and the gap between them is the whole test.
How you lose points
- You answer "yes, 70 percent." You read the signal accuracy straight off the page and hand it back as the probability. This is the exact mistake the question is fishing for.
- You flip the conditional. You treat the probability the indicator says "UP" given a rising market as if it were the probability of a rising market given an "UP" signal. That is base rate neglect, and it is the classic error.
- You ignore the 60 and 40. The market rises 60 percent of days on its own. Drop that prior and your answer is guaranteed wrong, because Bayes needs it.
- You forget the false alarm path. The indicator also says "UP" on 20 percent of down days. If you never count those, you never build the denominator.
- You say "higher than 70" and stop. Sensing the direction is not enough. On camera, with no number and no working, it reads as a guess.
- You bury the answer in words. A trader wants the number first, then the reasoning. Three minutes of narration with no clean 84 percent at the end loses the room.
How you pass
A strong answer says up front that no, it is not 70 percent, and then reaches for Bayes. Set the prior: the market rises 60 percent of days, falls 40 percent. Set the two signal rates: given a rising market the indicator says "UP" 70 percent of the time, given a falling market it says "UP" 20 percent of the time. Now weigh the two ways an "UP" signal can appear. On rising days it happens 0.60 times 0.70, which is 0.42. On falling days it happens 0.40 times 0.20, which is 0.08. Together, an "UP" signal shows up on 0.42 plus 0.08, which is 0.50 of all days.
The probability the market rises given an "UP" signal is the rising slice over the whole, 0.42 divided by 0.50, which is 0.84. So the real answer is 84 percent, not 70. Close on the intuition a trader actually cares about: the signal is informative, it lifts your estimate from the 60 percent base rate up to 84 percent, but the 70 percent figure was never the answer to the question asked. It was the indicator's hit rate on up days, a different conditional. Naming that distinction out loud is what separates a strong answer from a lucky one.
Get the ones for your role
This is the only question in the interview, and it rewards people who have drilled conditional probability before the timer starts, not in three minutes of prep. The full J.P. Morgan Markets question pack is on OpenInterview if you want to see the real prompts across the Sales and Trading roles before you record.
