For Recruiters
Should you consume meat? How often will you see a falling star?

Weird and difficult interview questions asked by top hedge funds

Interviewing with a top hedge fund is not dissimilar to interviewing with Goldman Sachs: you'll usually be expected to attend multiple rounds of interviews and the process might be stopped at any point. But while Goldman Sachs has all but given up asking brainteaser style interview questions, hedge funds are still fully committed to putting people on the spot. They may also ask you to sign a non-disclosure agreement to never talk about your interview once it's over. 

Based upon the questions recent interviewees nonetheless say they've been asked on Glassdoor, Wall Street Oasis and other sites, hedge funds' interviews can be challenging. Here's a sample of what you'll need to be ready for.

1. The probability you'll see a falling star in the sky over the course of one hour is 0.44. What's the probability you'll see one over half an hour?  (Asked by Citadel)

2.  You are given two ropes that when lit burn in one hour. Which one of the following time periods CANNOT be measured with your ropes? a) 50 min b) 30 min c) 25 min d) 35 min.  (Asked by Citadel)

3. You have three cards, each labeled n, n+1, n+2, and you don't know n. All cards start facing down so you can see them. You flip one card. If you choose to "stay", you get that card's value. If you don't "stay", then you flip another card. Again, choose to "stay" (and keep the 2nd card's value) or flip the final card and keep the final card's value. Design the optimal strategy to maximize the value of the card that you choose and find the expectation of that value.  (Asked by Citadel)

4.  Calculate 119^2 no pen and no paper, in 1 minute. (Asked by Citadel)

5. How much are you willing to pay to play the following game? You start with $1. You flip a fair coin. If it lands on heads, you double your winnings and flip again. If it lands on tails, the game is over and you collect the money you've won. You continue playing until you land on tails.  (Asked by Point72) 

6.  If two people buy 48 oranges. Pat buys 5 times more than Charles. How many oranges did Pat buy? (Asked by Point72)

7. Is it ethical to consume meat?  (Asked by Bridgewater)    

8. What's your life philosophy? (Asked by Bridgewater) 

9. You are a store manager. You could spend 200 visits for one year and there are total 100 stores. For each store, they have different revenue. For all the stores, the first visit will result in 2% increase in the revenue; 2nd visit -> 3%; 3rd visit -> 0; 4th visit ->0 while 5 th visit -> 10%. Come up with an algorithm to optimize the total increase in… (Asked by Bridgewater) 

10. Is social media good or bad for society? (Asked by Bridgewater)

11. What do you like or hate about Pandas [in Python]? (Asked by Balyasny)

Have a confidential story, tip, or comment you’d like to share? Contact: in the first instance. Whatsapp/Signal/Telegram also available.

Bear with us if you leave a comment at the bottom of this article: all our comments are moderated by human beings. Sometimes these humans might be asleep, or away from their desks, so it may take a while for your comment to appear. Eventually it will – unless it’s offensive or libelous (in which case it won’t.)

AUTHORSarah Butcher Global Editor
  • Ak
    10 November 2017

    The coin flips to dice can be done with an algorithm that terminates with probability 1.
    If the first flip is heads, pick from 1-3, else 4-6.
    Then flip until you get a tails, say it takes n tries (Not counting the flip from above). If n is even pick 1 (this happens with probability (1/4+1/16+1/64+...)=1/3). If n is odd, flip the coin again to decide between 2 and 3.

  • Ja
    5 December 2016

    9. I was thinking too much like a dev. From what I remember of stats you could also construct a bit string of 3-bits with the flips like before and if the value is not a valid die value (so if it is 0 or 7) keep shifting in new bits (dropping the MSBs above 3 bits) from the flips until it is. I _think_ that'll be a uniform distribution then. My only problem with this method is the unbound execution time and all the conditional jumps.

  • Ja
    4 December 2016

    9. Last comment on this - both versions are actually fine although the errors manifest very slightly differently (different die values are more common in either case) - but the error margin is the same in both cases.
    1+floor(6*(decimal_value/(2^n_flips))) should be quite a bit faster over 1+decimal_value%6, depending on compiler and architecture: the 6* and /(2^n_flips) can be converted to two bitwise shifts and one add, the %6 perhaps similarly, but less so.

  • Ja
    4 December 2016

    9!!. Damnit! Serves me right for rushing through them - the value of the die should be 1+floor(6*(decimal_value/(2^n_flips))). Using mod is obviously wrong as it ignores the additional entropy in the MSBs.
    The error is 2^-n_flips. So by ten flips the error is 1/1024.
    I'll get my coat.

  • Ja
    4 December 2016

    8. Sucky question - O(N^2) solution trivial, O(NlogN) less so.
    O(N) I had to look up:
    Manacher's algorithm. Apparently. Took it someone 20 years to realise it. Doubt I'd get it in the interview.

    11. What's a designer baby? Is it a GMO baby that is born with an Apple/Prada logo on their forehead? Or just on the brain? Yeah, that answer will get you past the cultural fit stage of the interview :)
    Or you could start by mentioning the difference between ethics and morals then go on to the effect on society, then maybe religion. But religion is best avoided at interview.

Show more

Apply for jobs

Find thousands of jobs in financial services and technology by signing up to eFinancialCareers today.

Boost your career

Find thousands of job opportunities by signing up to eFinancialCareers today.
Latest Jobs