Quantitative Researcher

A typical week often involves data work, model development, backtesting, and the steady cadence of research review — pulling and cleaning data, writing in Python or R, building model variants, evaluating performance against historical benchmarks, presenting research to senior researchers or PMs. You're often deep in code and statistics for hours at a stretch. Model performance and research output tend to be the operating measures.

The friction tends to come from the gap between backtested and live performance — a strategy that looks brilliant in research often underperforms in production, and the researcher has to be honest about overfitting and regime change. Variance across employers is sharp: at major hedge funds and asset managers the discipline runs in mature research infrastructure; at smaller shops you may have less data infrastructure and more responsibility per researcher.

Folks who do well here often bring graduate-level math or statistics training, strong programming, and intellectual honesty about model limits. PhD or master's backgrounds in quantitative fields anchor entry. The trade-off is the steady risk of dry research periods — months of effort can produce nothing useful, and that's part of the work.