The investment mind
your team doesn't have yet.

The Problem

Most investors don't fail from lack of information. They fail from inconsistency.

The best investment firms deploy teams of macro strategists, portfolio managers, fundamental analysts, and quant researchers — working together across every market regime. That coordination costs tens of millions in headcount annually. And it still breaks down under pressure — people leave, biases creep in, and process degrades at exactly the moments it matters most.

Family offices, independent advisors, and most CIOs face the same markets with a fraction of that infrastructure. They don't lack intelligence or conviction. What they lack is a structured, repeatable process that integrates macro, fundamental, portfolio, and risk — every day, without gaps.

MARY was designed to provide exactly that. Not by replacing human judgment, but by ensuring the analytical work is done rigorously, consistently, and at a depth that would otherwise require a team most firms simply cannot afford to build.

The System

MARY reasons like an entire investment committee — at once.

Consider what a CIO actually does every day: synthesize the latest macro data, read the headlines, gauge market sentiment, track how it connects to portfolio positioning, reassess individual companies, weigh sector dynamics, and arrive at a coherent conclusion — all before the market opens. No matter how experienced or talented, no human can do all of this consistently. There are only 24 hours in a day. Information compounds faster than any person can process it.

MARY can. Four specialized agents work in parallel — macro intelligence, portfolio construction, fundamental analysis, and quantitative strategy — integrated into a single, continuous decision process. Not four separate tools. One system where every layer informs the others. She reads the macro regime from 21 leading indicators, constructs portfolios using institutional optimization frameworks, scores companies across quality, value, growth, and momentum, connects the dots across every dimension, and knows when to get in and when to get out. Every day. Without fatigue, without bias, without gaps.

And she learns — in a way most AI systems cannot. MARY's machine learning layer retrains continuously on her own outputs: every regime she classifies, every allocation she recommends, every signal she generates becomes training data for the next cycle. She even produces her own macro assessments independently — classifying stagflation, contraction, or recovery from the data itself, not from consensus. The result is a system that compounds intelligence over time, getting sharper with every market cycle it survives.

How MARY works →

Compounding Intelligence

The longer MARY runs, the sharper she becomes.

Every signal MARY generates feeds back into her learning pipeline. New data becomes training data. New strategies generate new insights that inform better strategies. Her models retrain continuously, her conviction engine adjusts on rolling outcomes, and strategies that degrade are automatically demoted.

As proof of this intelligence at work, MARY's strategy research has produced backtested results across 35 years of market data — spanning the dot-com crash, the 2008 financial crisis, a pandemic, and multiple full regime cycles.

These numbers are not the product — they are evidence that the reasoning architecture works. When the intelligence is sound, the results follow.

How the intelligence flywheel works →

The Founder

Parson Tang is an investment professional with 20+ years of experience across institutional finance — including roles at Goldman Sachs and J.P. Morgan, where he served as Executive Director. He has advised foundations, family offices, and institutional clients across public equities, private equity, and real assets. He holds an MBA from Oxford and a Computer Science degree from USC.

MARY is what he built when he decided the investment process itself needed to be rebuilt.

About ClarityX →

Research

The thinking behind the system

Our research examines how institutional investment decisions are actually made — and where they break down. Market regimes and structural cycles. Portfolio construction as a systematic discipline. The limits of forecasting under uncertainty.

MARY operationalizes these frameworks. The research explains the reasoning behind the system.

Research themes →

Publications

Essays & White Papers

Long-form thinking on AI in institutional finance, portfolio construction frameworks, macro regime methodology, and the future of investment decision-making.

View all publications →