I am an M.A. candidate in Quantitative Methods in the Social Sciences at Columbia University,
working on market microstructure, stochastic finance, financial econometrics, and market design.
My current research studies Automated Market Makers, CEX–DEX arbitrage, liquidity provision,
inventory risk, and execution quality using stochastic modeling, simulation, and empirical market
analysis. I am particularly interested in what happens to the classical microstructure objects
(routing, price discovery, latency, and adverse selection) when liquidity for the same asset is
fragmented across venues with very different settlement and information properties.
Prior to Columbia, I completed a B.Sc. in Mathematics (Honors) with a minor in Economics at Lady
Shri Ram College for Women, University of Delhi. Along the way I have worked on fiscal ranking
methodologies and trade analysis at NITI Aayog, India's federal policy think tank; on transport
economics and commuting behaviour at SPIESR, work that grew into two Springer chapters; and on
causal program evaluation at Columbia's Institute for Social and Economic Research and Policy.
My research interests are Market Microstructure, Quantitative Finance, Financial Econometrics,
Causal Inference, Stochastic Optimization, and Algorithmic Trading. I am particularly interested in
how market design, latency, liquidity concentration, and routing incentives shape outcomes in
decentralized and centralized markets: where price discovery actually happens when the same asset
trades on many venues at once, how arbitrageurs and liquidity providers respond to the frictions
each venue imposes, and what this implies for the design of exchanges and lending markets.
Methodologically, I work with stochastic processes and simulation, econometric and causal-inference
designs (difference-in-differences, panel methods, quasi-experimental estimators), and
optimization, from mixed-integer programming to reinforcement learning. I care about empirical work
that is reproducible end to end: my projects ship with their data-collection pipelines and code.
Research
Market Design and Liquidity Provision in Automated Market Makers
Work in Progress · Mentored research under Prof. Steven Campbell, Columbia University
Developing stochastic models of liquidity provision, inventory risk, and cross-exchange arbitrage
across centralized and decentralized exchanges. The project asks how order routing, price
discovery, and latency behave when liquidity for the same asset is fragmented across multiple
venues; how arbitrage links centralized and decentralized markets and how quickly mispricings
close; and how adverse selection, meaning informed flow picking off passive liquidity providers,
manifests differently across venue designs and feeds back into routing incentives. The approach
combines stochastic modeling and simulation with empirical analysis of cross-venue trade and
quote data.
Planning for a Fast Metropolizing Developing Economy
Springer Nature. Accepted in 2024 [forthcoming].
Naga Siva Gayatri Dittakavi and Nachiket Gosavi
A book chapter on the planning challenges of a rapidly metropolizing developing economy, drawing
on our earlier empirical work on commuting behaviour and residential location in India.
Optimizing Urban Economic Potential Using MOORA and FUCA Techniques
Taylor & Francis. Revise & Resubmit.
Naga Siva Gayatri Dittakavi and Yograj Singh
Applies two multi-criteria decision-making methods, MOORA (Multi-Objective Optimization on the
basis of Ratio Analysis) and FUCA (Faire Un Choix Adéquat), to construct a composite
ranking of city districts by economic potential, and compares how the choice of aggregation
method affects the resulting ordering.
Commute Distance and Residential Characteristics: Weathering the Storm
Lecture Notes in Civil Engineering, Springer Nature Singapore. Published. Contributor.
Tests whether the developed-country finding that residential self-selection plays little role in
commute distance holds in India, using the nationally representative 76th round of the NSSO. Over
two-thirds of job-related trips cover under 5 km, relocation for employment is rare, and
household location and physical characteristics show negligible-to-weak association with commute
distance, implying that the four-step transport model cannot be the sole basis for transport
planning in Indian cities. Presented at IMRC 2024, IIM Ahmedabad, and TPMDC 2024, IIT Bombay.
Link
Examines whether the gender gap in commuting documented in developed economies appears in India,
drawing on household-level data from the 76th NSSO round. Stratified regressions compare women's
commute patterns across urban and rural settings and identify their determinants, with
implications for women-centric transport policy.
Link
Three Person–Two Cuts Problem
Éclat, Annual Journal of the Department of Mathematics, LSR, University of Delhi, 2023. Published.
Naga Siva Gayatri Dittakavi, D. Bharadwaj, and N. Kumar
A fair-division paper on allocating a continuously divisible good among three agents using only
two cuts, written under the department's faculty mentorship program.
Journal PDF
Personal Projects
Deep Reinforcement Learning & Systematic Trading Strategies
Extended Sood et al. (2023) into a PPO portfolio-allocation framework integrating
earnings-transcript sentiment, macro-regime indicators, and options-implied volatility features
across equities, fixed income, commodities, and cash, with turnover penalties and ESG exclusions
built into the action space. Applied SHAP for explainability of the learned policy and
stress-tested the agent over the 2020 COVID shock and the 2022 tightening cycle, where it
outperformed constrained mean-variance optimization on Sharpe, Sortino, Calmar, and
turnover-adjusted net returns.
Portfolio Optimization via Mixed-Integer Quadratic Programming
Formulated portfolio construction as a Mixed-Integer Quadratic Program in Gurobi, encoding the
constraints institutional portfolios actually face (cardinality limits, minimum-lot thresholds,
and sector-concentration caps) within a mean-variance objective. Benchmarked against
unconstrained mean-variance and equal-weight (1/N) baselines across equity, fixed-income, and
commodity universes, quantifying the efficiency cost of each constraint class along the
risk-return frontier.
Built an IMC Prosperity trading dashboard and Python backtester supporting SMA, EMA, RSI,
Bollinger Bands, and volatility filters, with signal visualization, parameter sweeps,
transaction-cost assumptions, and performance attribution, so that a strategy idea can go from
hypothesis to cost-adjusted tearsheet in one workflow. Ranked 85th nationally in the United
States in the 2026 Prosperity Challenge.
MPATH-India
Founded a public-health research initiative and recruited over 70 volunteers while developing a
geospatial disease-surveillance framework. The framework combines k-means clustering with
Principal Component Analysis to identify high-risk regions from multivariate health and
demographic indicators, improving prevalence mapping and helping target limited screening
resources where they matter most.
Selected Talks
2024
Commute Distance and Residential Characteristics: Weathering the Storm @ IMRC, IIM Ahmedabad
Commute Distance and Residential Characteristics: Weathering the Storm @ TPMDC, IIT Bombay
2023
Understanding Crime Using Mathematics: An Analysis Through Partial Differential Equations @ Anupama Dua Paper Presentation
2022
Limits to Commute: The Case of Indian Women @ TPMDC, IIT Bombay
How Far Do Women Commute? @ National Seminar on Changing Characteristics of the Indian Labour Market, SPIESR
Professional Experiences
Mentored Researcher under Prof. Steven Campbell @ Columbia University
May 2026 – Current
Research on market design and liquidity provision in Automated Market Makers: stochastic models
of liquidity provision, inventory risk, and CEX–DEX arbitrage, with attention to order
routing, price discovery, latency, and adverse selection in multi-venue markets.
Departmental Research Assistant for Prof. Sandhya Kajeepeta @ Columbia ISERP
Oct 2025 – Dec 2025
Applied quasi-Poisson regression and difference-in-differences designs to estimate causal policy
impacts; diagnosed model robustness across competing specifications and produced
publication-ready visualizations for downstream research deliverables.
Research Intern, Economics and Finance Vertical @ NITI Aayog
Jul 2024 – Aug 2024
Redesigned fiscal performance rankings for all 28 Indian states via Wroclaw taxonomy, and
authored a 15-page logistics benchmarking report identifying five global best practices that
directly informed policy directives cutting transport costs by 9%. Performed granular trend
decomposition on the Economic Survey 2024 External Sector chapter, isolating the structural
drivers of 12% year-on-year export growth.
Research Intern, Transport Research @ SPIESR
Jul 2022 – Aug 2023
Worked on transport economics and commuting behaviour in India using nationally representative
NSSO survey data; this research developed into the Springer chapters Limits to Commute
and Commute Distance and Residential Characteristics.
Education
Columbia University in the City of New York
Sep 2025 – Dec 2026
M.A. in Quantitative Methods in the Social Sciences
Coursework: Stochastic Processes & Applications, Advanced Econometrics, Statistical Methods
for Finance, Time Series Analysis & Panel Data, Algorithmic Trading, Asset Allocation,
Natural Language Processing.
Lady Shri Ram College for Women, University of Delhi
Aug 2021 – May 2024
B.Sc. Mathematics (Honors), First Class · Minor in Economics
Awards & Academic Honours
AMTI National Mathematics Olympiad: All India Rank 8, across a four-stage national examination.
NASA Ames National Space Settlement Design Contest: 3rd place internationally, judged at NASA Ames Research Center.
Grade XII CBSE: 3rd rank at state level.
Beyond Research
I play tennis and I follow it closely. Playing supplies regular first-person evidence of the gap
between knowing what a shot should do and executing it; watching the professional game means
seeing people for whom that gap has nearly closed. I like the sport for some of the same reasons I
like my research field: it is strategic, measurable, and honest about variance.
The other standing commitment is reading, usually several books at once. The rotation alternates
between mathematics and economics read for pleasure and fiction, which supplies the kind of
problems that do not have closed-form solutions.
Contact Me
The fastest way to reach me is email. I am always glad to talk about market microstructure,
crypto-economic research, or doctoral study.
Or write to me directly: nd2921 (at) columbia (dot) edu
