Mangrove CEO Tim Darrah sat down with Marty Kendall of MNAV on the Bitcoin for Corporations live stream, hosted by Jake Fenton at Strategy World in Las Vegas. The conversation covered how Mangrove uses AI in strategy development, why the company keeps AI out of live execution, and what institutions with $100 million or more in held digital assets are looking for when they want to put those assets to work.
Why you should watch
Bitcoin treasury companies are stacking digital assets on their balance sheets. The question they’re all asking now: how do we make those holdings productive? Tim and Marty represent two sides of that answer. MNAV is building what Kendall calls “the Bloomberg for the Bitcoin treasury ecosystem,” tracking metrics like Bitcoin yield per share across 130+ companies. Mangrove is building the strategy infrastructure that lets institutions find, validate, and deploy algorithmic strategies against their held assets, with hard separation between AI-assisted design and deterministic execution.
What you’ll learn
Tim walks through Mangrove’s approach to AI in trading, and why it’s built differently from the two dominant patterns he’s seeing in the market right now: AI research tools that are mostly prompt wrappers on base models, and AI trading bots that take control of your wallet and execute autonomously. Mangrove’s architecture sits in a third category. AI handles strategy design, backtesting, and validation. Once a strategy clears the platform’s minimum thresholds (Sharpe ratio, Sortino ratio, max drawdown, internal rate of return), it moves to a deterministic execution engine where AI plays no role. The logic: when a strategy’s rules reduce to true-or-false conditions, you don’t need a large language model to evaluate them.
Here’s what Tim and Marty covered:
- 2 million+ strategies and counting — Mangrove’s strategy database has more than doubled from just over 1 million. The AI Strategy Engine generates over 50,000 new strategies per day, each tested against the platform’s performance thresholds before it can be deployed live.
- Market regime awareness — A strategy that worked three months ago might not work now. Tim explained why Mangrove evaluates strategies in the context of specific market conditions, so institutions can match strategies to current regimes and rotate as conditions shift.
- Open source signal library — Over 200 technical indicators and chart patterns make up the foundation of Mangrove’s strategy framework. The signal library, SDK, and a trading agent are all open source and available for developers.
- API-first, like Stripe — Most of Mangrove’s services are available today through its Developer API. Tim compared the approach to Stripe’s early days: API-first for developers who want better infrastructure, with the consumer-facing platform building out behind it.
- Institutional demand at the $100M+ level — Tim shared that institutions managing $100 million or more in held assets are the ones driving the most interest in Mangrove’s strategy generation engine. They want to put idle assets to work generating yield instead of watching them sit while currency debasement chips away at purchasing power.
- Bitcoin volatility, reframed — Tim broke down why the volatility conversation usually misses the point. Bitcoin’s overall volatility has compressed since 2017, and the majority of it skews to the upside. His take: market drawdowns of 50% from cycle highs are historically normal, not surprising.
- Upcoming research paper — Mangrove has a paper coming out on a predictive model that can assess a strategy’s probability of success without running a backtest or looking at market data.
Kendall rounded out the conversation with MNAV’s power law forecasting tools, a block-height-versus-price model that smooths out the noise from Bitcoin’s early mining irregularities. Both agreed that building in a down market is what sets the foundation for the next cycle.
About Tim Darra
Dr. Timothy Darrah is the co-founder and CEO of Mangrove. He’s an Army veteran, former NASA Fellow, and assistant professor at Vanderbilt University, where he sits on the faculty committee developing Vanderbilt’s generative AI master’s program. Before Mangrove, he spent more than a decade building AI technologies in defense and aerospace, including work at Deloitte and on autonomous systems.
Mangrove provides trading tools and infrastructure. Nothing in this video or post constitutes financial advice. Digital asset trading involves significant risk, including potential loss of principal. Past strategy performance, whether backtested or live, does not guarantee future results.
