The 'No-Math' Quant Interview Puzzle
A quant interview puzzle is making the rounds for its focus on probabilistic thinking over calculation. The challenge: Prove the probability of bulb X outlasting bulb Y is x/(x+y), given their expected remaining lifespans are x and y. It’s a test of logic, not just formulas.
The puzzle's elegance lies in its reliance on a core concept in probability theory: the memoryless property of the exponential distribution. This property implies that the remaining lifespan of a bulb is independent of how long it has already been operating; an old bulb is as good as a new one in terms of future life expectancy. This is a common, if counter-intuitive, assumption in many probability puzzles. Given the memoryless property, at any given moment, the probability of one bulb failing before the other is proportional to their failure rates. If bulb X has an expected lifespan of x and bulb Y has a lifespan of y, their failure rates are inversely proportional to their lifespans, namely 1/x and 1/y, respectively. This setup can be viewed as a "race" to failure. The probability that bulb X "wins" this race (i.e., fails first) is its failure rate divided by the sum of both failure rates. This can be expressed as (1/y) / (1/x + 1/y). A little algebraic simplification of this complex fraction yields the elegant solution: x/(x+y). Firms like Jane Street, Citadel, and Jump Trading are known for using such puzzles in their quantitative interviews. They are less interested in a candidate's ability to perform complex calculations under pressure and more in their ability to demonstrate deep, intuitive understanding of fundamental concepts like expected value and probability. The thought process and the ability to articulate a logical framework for the solution are more valued than arriving at the correct answer through rote memorization of formulas.
