Systematic Equity Research

*Researching equity signals under the constraints that decide whether an idea survives: turnover, correlation, exposure, and out-of-sample behavior.*

**Summary:** At WorldQuant BRAIN, I research and submit systematic equity signals, supported by a continuous backtesting workflow and a corpus of 7,700+ tested signals.

**Status:** active · **Year:** July 2026 – present **Tags:** Quant research, Python, Backtesting, Time series, Validation

---

A signal is only the beginning. *RESEARCH CONTEXT*

As a Research Consultant at WorldQuant BRAIN, I research systematic equity signals across price and volume, options-implied volatility, analyst revisions, short interest, and news sentiment. The work focuses on a US large-cap universe, with sector neutralization and explicit risk constraints.

• **Research corpus:** 7,700+ (tested signals in the supplied résumé snapshot)
• **Research throughput:** 2k–4k / day (simulations in the cloud workflow)
• **Platform milestone:** Gold level (WorldQuant BRAIN)

Make the research repeatable. *THE WORKFLOW*

**Pipeline:**

1. **Form a hypothesis** (input) — Identify a candidate mechanism or data relationship. 2. **Build the signal** (process) — Translate the hypothesis into a testable expression. 3. **Run the experiment** (process) — Backtest and log the result. 4. **Check the constraints** (rule) — Sharpe, turnover, correlation, and exposure. 5. **Evaluate and iterate** (output) — Compare candidates and out-of-sample behavior.

A public overview of the research process. Proprietary signal definitions and underlying research data are not included.

I built a continuous backtesting workflow on a cloud VM, logging results into a shared research corpus. Keeping experiments comparable makes it possible to investigate why a candidate changes, instead of selecting a result on a headline score alone.

Performance in context. *WHAT I EVALUATE*

Turnover A promising backtest must be considered alongside how frequently the signal trades.

Correlation A candidate needs to be evaluated against existing signals, not only in isolation.

Exposure Sector neutralization and risk constraints shape how a signal is interpreted.

Generalization Out-of-sample checks help distinguish a repeatable relationship from a result that only fits the research sample.

The figures above describe a research snapshot and simulation throughput. They are not realized returns or a claim of future performance. The résumé records a selected 2.45 Sharpe backtest at 0.53 turnover; that is a backtest statistic, not a live portfolio return.

**Tools:** Python · Signal research · Backtesting · Out-of-sample validation · Time-series analysis · Cloud experiment automation

• [Current résumé](/cv/rajya-goel-cv.pdf)