Independent Normals Conditional Probability

This probability question is built around the geometry of two independent standard normal variables viewed together in the plane. In quant prep, it's a classic way to connect one-dimensional distributions (like a single normal) with their joint behavior in higher dimensions. Rather than focusing on messy formulas, it spotlights the symmetry and structure that emerge when you consider both variables at once.

It trains geometric intuition for joint distributions, conditional probability, and the effect of linear constraints on symmetric random vectors. You practice seeing events as regions, understanding how conditioning reshapes the sample space, and recognizing when symmetry lets you avoid heavy computation. It also reinforces fluency with standard normal variables, which are everywhere in quant interviews.

This matters for quant interviews because front-office roles demand fast, conceptual thinking about probability under constraints. Being able to turn algebra into clean geometric reasoning is a core skill for trading, derivatives pricing, and risk. On quant interviews, such questions separate candidates who just memorize formulas from those who truly understand joint distributions, conditioning, and symmetry-key for tackling more complex models and high-dimensional problems under time pressure.

What it tests

Whenever you have independent, identically distributed continuous random variables, their joint distribution in two dimensions is rotationally symmetric about the origin. Conditioning on a linear combination (like $X + Y > 0$) carves out a half-plane or wedge in this space, and probabilities of further events (like $X > 0$) correspond to the fraction of this region's angle or area that also satisfies the second condition. The symmetry means that geometric reasoning—measuring angles or areas—can often replace algebraic integration. This works because the density is constant along circles centered at the origin, so the probability of being in a sector is proportional to its angle. The principle holds for any pair of independent, identically distributed normal variables and any linear constraints, not just the specific numbers here.