Quantara

Inspiration

Over the past year, escalating tariff announcements from the United States government have triggered immediate and often severe volatility across global equity markets. Major indices like the S&P 500, Nasdaq Composite, and Dow Jones Industrial Average have reacted not only to policy decisions themselves but to the uncertainty surrounding them.

Markets struggle with discontinuities. Traditional financial models assume smooth price evolution, while geopolitical policy moves create sudden structural shocks. At the same time, prediction markets often reflect sentiment shifts before institutions fully price them in. Currently, no applications focus on how tariff data affects general trends in markets, both for individual countries and economic sectors.

This led us to ask a core question:

What if we could quantify tariff uncertainty before it hits and simulate its ripple effects across global markets in real time?

Quantara was built to transform policy uncertainty into measurable, forward-looking financial insight.

What it does

Quantara is a full-stack predictive analytics platform that stress-tests global markets against shifting tariff landscapes.

By bridging prediction market sentiment with stochastic financial modeling, we provide a real-time macroeconomic impact simulator.

Our platform:

Calculates a “Surprise Gap” between market expectations and tariff probability
Visualizes geopolitical risk via an interactive global choropleth map
Simulates 90-day price trajectories for major indices and sectors
Models discontinuous policy shocks using jump-diffusion mathematics
Translates complex quantitative outputs into plain-English explanations through an AI analyst

Instead of reacting after the market moves, Quantara enables proactive macro risk assessment.

How we built it

Quantara operates through a dual-engine architecture supported by an intelligence layer.

1. The Probability Engine (Sentiment Analysis)

We ingest data from Kalshi prediction markets to compute a Surprise Gap — the delta between implied probability and expected tariff outcomes.

This gap acts as a forward-looking policy stress indicator. We surface this data through an interactive global choropleth map, allowing users to instantly identify geopolitical hotspots.

2. The Impact Engine (Stochastic Forecasting)

We convert the Surprise Gap into projected financial impact using a Merton Jump-Diffusion model.

Unlike traditional geometric Brownian motion models, this framework captures:

Jump Component: The instantaneous market shock following a tariff announcement
Diffusion Component: The longer-term cost absorption and repricing process

We run Monte Carlo simulations to generate 90-day probabilistic trajectories for:

Major indices (S&P 500, Nasdaq, Dow)
Sector-specific exposure
Country-sensitive industries (e.g., automotive, semiconductors)

The result is a dynamic forecast distribution rather than a static estimate.

3. The Intelligence Layer (RAG Chatbot)

To make complex quantitative outputs accessible, we integrated a Retrieval-Augmented Generation (RAG) chatbot powered by Snowflake Cortex.

This AI analyst:

Interprets jagged simulation graphs
Correlates projections with policy documents
Connects outcomes to supply chain metadata
Explains sector-specific valuation cliffs in clear, plain English

Users can query the system to understand why a specific sector is projected to move, not just how much.

Challenges we ran into

Translating qualitative geopolitical uncertainty into a measurable numerical shock variable
Calibrating jump intensity and diffusion parameters without overfitting
Aligning prediction market probabilities with financial volatility data
Maintaining interpretability while preserving quantitative rigor
Integrating real-time data pipelines within a hackathon timeline
Configuring the chatbot with API authentification issues

Accomplishments that we're proud of

Successfully integrating prediction market sentiment into stochastic financial modeling
Implementing a working Merton Jump-Diffusion simulation engine
Building an interactive geopolitical risk visualization interface
Deploying a RAG-powered AI analyst for contextual interpretation
Delivering a cohesive full-stack macro stress-testing platform within a hackathon timeframe

What we learned

Markets price uncertainty faster than finalized policy decisions.
Prediction markets can serve as leading macro indicators when contextualized properly.
Advanced financial models gain exponential value when paired with interpretability layers.
The intersection of macroeconomics, stochastic modeling, and LLM systems represents a powerful new class of decision-support infrastructure.