Rithik Singh

Quantitative research infrastructure · Summer Analyst

Question

Of more than a thousand candidate factors, which ones hold up out of sample, at which holding periods, and in which market regimes?

Built

• A factor-evaluation pipeline with point-in-time, survivorship and look-ahead controls, Newey–West HAC t-tests and Benjamini–Hochberg FDR control, run across eight partitions.
• A vendor's combined six-model equity-ranking framework, pulled apart into its individual models and recombined with non-negative least squares on a walk-forward test.
• A portfolio screening and ranking tool built on the factors that held up, with results conditioned on 11 macro dimensions, used by two portfolio managers.
• An LLM earnings-research agent with deterministic numeric validation.
• The firm's local research databases, about 480 GB of market and fundamental data, maintained and then migrated to the cloud.

Result

• 1,178 specifications screened. The 66 that held a Newey–West t-stat above 3 in at least half their tests were combined into a new model.
• The recombined vendor model showed apparent alpha that did not survive sector neutralization, so it stayed in research, out of the book.
• Earnings-call review took about half as long in Q2 2026 testing.

Stack

Python · SQL · PostgreSQL · Azure SQL · LLM tooling

Algorithmic trading competition · UTEFA

Question

Can a trend-and-momentum strategy stay profitable after a 0.5% fee on every trade?

Built

• An EMA-crossover regime signal that moves between invested and cash, with momentum ranking into a three-stock basket and an ADX trend filter.
• Volatility and ATR tilts on the weights, macro-sensitive exposure scaling, a 5% trailing stop and a 20-trade cap, all built to survive a 0.5% fee on every trade.
• 144 parameter sets compared in a grid search, alongside bootstrap and correlated-GBM simulations.

Result

Third place, with a 24.3% simulated return and a 1.78 Sharpe over 252 trading days, net of fees.Team entry with Ethan Yang, Michael Xu and Sina Sharifi.

Stack

Python

Market making and tender arbitrage · Rotman Interactive Trader

Question

When is a tender or a quote still worth taking after liquidity, inventory and fees?

Built

• Tender-arbitrage and market-making algorithms trading against live order books.
• Tenders accepted on VWAP and liquidity rules, quotes skewed to hold inventory near flat, fees priced into every execution, and positions flattened automatically before the close.

Result

Fourth of 26 teams.Team entry with Michael Xu.

Stack

Python · REST APIs

Monte Carlo decision engine

Question

How likely is a profitable strategy to reach a payout before it breaches a loss limit?

Built

• A Monte Carlo engine that resamples TradingView, NinjaTrader and MetaTrader 5 trade histories into 20,000 paths.
• Each path runs against the account's trailing-drawdown, daily-loss, consistency and withdrawal rules, with optional weighting toward recent trading days.
• Correlation analysis and portfolio optimization across several accounts at once.

Result

In use, and checked against the realized outcome of a live strategy, which went on to reach a payout.

Stack

Python · FastAPI · Next.js