Explain Daily Signal Construction and a Prediction Workflow

Read the full interview experience this question came from →

Quick Overview

Explain daily quantitative signal construction and prediction at an appropriate level of disclosure, with point-in-time inputs, workflow validation, ownership, and confidentiality boundaries.

Explain Daily Signal Construction and a Prediction Workflow

Company: J.P. Morgan

Role: Quantitative Researcher

Category: Machine Learning

Difficulty: medium

Interview Round: Technical Screen

Walk through a quantitative project in which you constructed a daily signal and used it to produce predictions. Explain the workflow, your role, and how you checked that the signal and prediction process behaved as intended. ### Constraints & Assumptions The source reports a deep dive into prior strategy work, specifically daily signal construction and prediction. It does not disclose a particular model, dataset, target, formula, or trading rule. Use your own permitted experience, or explain the method at a high level without revealing confidential details. Do not invent a proprietary strategy to fill the gaps. ### Clarifying Questions What can you share about the prediction target and daily cadence? When were inputs actually available relative to prediction time? Which part did you own? How was the signal transformed into a prediction and evaluated? ### What a Strong Answer Covers A clear daily data-to-signal-to-prediction sequence, timing and availability assumptions, validation, personal contribution, and an explicit boundary around restricted details. ### Follow-up Questions How would you rule out look-ahead leakage? What would happen if an input arrived late or was revised? How would you distinguish a useful prediction from a promising but unreliable historical result?

Overview: Explain daily quantitative signal construction and prediction at an appropriate level of disclosure, with point-in-time inputs, workflow validation, ownership, and confidentiality boundaries.

Read the full J.P. Morgan Quantitative Researcher interview experience this question came from

|Home/Machine Learning/J.P. Morgan
J.P. Morgan logo
J.P. Morgan
Mar 18, 2026
mediumQuantitative ResearcherTechnical ScreenMachine Learning
0
0

Walk through a quantitative project in which you constructed a daily signal and used it to produce predictions. Explain the workflow, your role, and how you checked that the signal and prediction process behaved as intended.

Constraints & Assumptions

The source reports a deep dive into prior strategy work, specifically daily signal construction and prediction. It does not disclose a particular model, dataset, target, formula, or trading rule. Use your own permitted experience, or explain the method at a high level without revealing confidential details. Do not invent a proprietary strategy to fill the gaps.

Clarifying Questions Guidance

What can you share about the prediction target and daily cadence? When were inputs actually available relative to prediction time? Which part did you own? How was the signal transformed into a prediction and evaluated?

What a Strong Answer Covers Guidance

A clear daily data-to-signal-to-prediction sequence, timing and availability assumptions, validation, personal contribution, and an explicit boundary around restricted details.

Follow-up Questions Guidance

How would you rule out look-ahead leakage? What would happen if an input arrived late or was revised? How would you distinguish a useful prediction from a promising but unreliable historical result?

Loading comments...