Tuatara Hedge Fund AI
Specialized AI systems have recently solved some of the world’s toughest math problems[1] and have the proven ability to think of things humans miss. They also love to optimize. Develop the right proprietary algorithms and AI will continuously find new ways to improve and optimize, or find hidden alpha, depending on how you direct it.
The Tuatara Hedge Fund is an AI. It specializes in financial language modeling in the global financial markets. It uses proprietary algorithms to discover new ways to optimize returns.
You can connect your AI to the Tuatara AI agent, Nexus, and they’ll work together to optimize returns on your money in global financial markets based on the universe of trading vehicles it creates and manages. The prompt to give your AI is under the chart below.
The long version: how the system is built
The Tuatara Hedge Fund is an AI system made up of several fine-tuned frontier AI models — including Claude, OpenAI GPT 5.*, Google Gemini and others — which are trained to trade global markets based on global events.
Today's AI models are making breakthroughs in many areas, such as mathematics, on a regular basis [1][2][3][5][6]. Making breakthroughs in the way people trade and access information is no different.
Leveraging multi-trillion dollar AI models by fine-tuning for financial markets enables the Tuatara Hedge Fund to operate with AI as its DNA. In other words, the fund was built by the AI system and for other AI systems, models and AI agents. This means you can have your AI agents — like IronClaw, OpenClaw, NanoClaw and others — interact with the Tuatara Hedge Fund AI in unique and powerful ways: creating their own AIB funds, adding algorithms, filters, technical indicators and prediction markets, all programmatically via our REST API, MCP tools and SDKs. The AI is trained to mimic principles in biological communication networks and apply them to optimizing returns in the global financial markets. It does this by analyzing global events — from an earthquake in Taiwan to a cure for cancer, from global conflicts to global weather, and any multi-billion or multi-trillion dollar transactions in between — in order to trade assets that might move up or down based on those kinds of global events or headlines [Tuatara vs Claude: head-to-head].
This also means the Tuatara Hedge Fund was built from the ground up as an Agentic Operating System (AOS)[4], which enables it to get stronger over time as it uses trillions of data points to triangulate and optimize its entire network of data engineering pipelines, trade execution pipelines, architecture, algorithms, tools, skills, external networks, strategies and data. It operates like a biomimetic synthetic living organism: due to its AI-native design it grows, evolves and gets stronger over time.
AI-Managed Tokenized Funds — designed to enable traders and investors to ride the next multi-trillion dollar waves set to enter global markets
Description: A combination of the Tuatara Vector Language Model is being used to fine-tune AI models for the purpose of generating assets, initially stocks, cryptos and perpetual futures, with direct and hidden connections to each category, class and subclasses. Dynamic Human-In-The-Loop curation will be used to finalize relationships. This will be represented as a graph network based on biomimetic approaches in biological cell–cell communication and signal transduction networks. Tokenized crypto assets represent the trading vehicle infrastructure which makes investment into these new sectors possible. In part, this process represents convergence between AI tokens and crypto tokens at Cymetica — from molecule to AI token to crypto tokenized asset.
Standalone funds managed by the Tuatara Hedge Fund alongside the deployment sleeves above — each is an AI-generated Index Basket (AIB) with a live real-time index. Each fund is governed by a research-grounded sector ontology with supply-chain analysis and human-curated asset connections — explore it on the AIB Ontology Graph. Each tile’s return is its index (model) move since the fund’s start — a separate measure from the executed-deals chart above, and not part of it.
Molecular
Wire ·
Soybeans, planes and rare earths: What they say about fragile U.S.-China trade truce Rare Earths & Critical Minerals
Memory
Wire ·
Intel, SK Hynix shares jump on report they're discussing U.S. memory chip manufacturing Memory Chips
AMF-WhiteSwan
Anthropic-Long
Wire ·
A generational flight from tech to defensive stocks is underway. How we’re capitalizing Generative AI
AMF-BlackSwan
Wire ·
House Intelligence Committee warns of 'Black Swan' AI risks Black Swan
Defense
Wire ·
Russia’s neighbor backs push to lift ban on nuclear weapons, prompting outcry from Moscow Weapon
Power
Wire ·
What midterm election blue wave rebuke of Trump means for renewable energy Renewable Energy
Anthropic-Short
Wire ·
A generational flight from tech to defensive stocks is underway. How we’re capitalizing Generative AI
Data-Protection
Cybersecurity
Wire ·
Amazon adds cybersecurity veteran Kevin Mandia to board 2 years after he left Google Cybersecurity
Biosciences
Wire ·
A little-known biotech’s high-stakes cancer bet just paid off and its shares doubled Biotechnology
Space
Wire ·
SpaceX catches upper stage of Starship rocket in historic first Rocket
Data Centers
Wire ·
Data center policy dominated the midterms. The 'Virginia of Europe' is the latest battleground Data Centers
AI Compute
Wire ·
Data center policy dominated the midterms. The 'Virginia of Europe' is the latest battleground Data centers
Robotics
Wire ·
Eric Trump-backed robot soldier startup can't pay staff after founder borrowed $4.6 million from company to buy home Robotics
Quantum
Wire ·
IonQ shares rally after company says it made a major quantum computing breakthrough Quantum Computing
Fund performance
⚙️ Watch the AI continually optimize returns for every fund in real-time → — every optimization, search and live parameter change behind the strategies below, with the measured impact of each parameter.
The fund now runs as a network of 4 event pipelines (1 live) — each deploys its own capital on one kind of event and is measured from real fills the next day. Latest measured day: +0.00% on deployed capital. See every node →
Live capital deployment of the Tuatara Hedge Fund — a multi-strategy fund-of-funds powered by the Tuatara vector engine. Deployed capital currently sits 85.1% long and 14.9% short across the strategies below, each tracked by its own index (target mandate 40/60). The long deployment is the M&A Ripple Fund.
All funds, ranked
Every fund with outside money, best return first. LIVE funds take real USDC; PAPER funds trade simulated and take no investment yet; STANDING funds are live but not yet allocated.
| Fund | Sleeve | Status | Index | Return | |
|---|---|---|---|---|---|
| 01 | FTA Fund — Cryptos | FTAshort | live | $0.07 | +38.81% |
| 02 | FTA Fund — Stocks | FTAshort | standing | $154.47 | +15.64% |
| 03 | Molecular | Managed AIBlong | live | $57.27 | +4.46% |
| 04 | M&A Ripple Fund | M&A Ripplelong | live | $94.83 | -5.17% |
| 05 | Defense | Managed AIBlong | live | $272.94 | -6.30% |
| 06 | Power | Managed AIBlong | live | $0.01 | -7.09% |
| 07 | Cybersecurity | Managed AIBlong | live | $127.62 | -11.03% |
| 08 | Biosciences | Managed AIBlong | live | $70.17 | -11.73% |
| 09 | Space | Managed AIBlong | live | $55.31 | -13.80% |
| 10 | Data Centers | Managed AIBlong | live | $119.31 | -21.05% |
| 11 | AI Compute | Managed AIBlong | live | $134.31 | -27.76% |
| 12 | Robotics | Managed AIBlong | live | $2.73 | -35.65% |
Model portfolios
These funds hold no outside money yet, so they are listed here rather than ranked. Each one runs live on our own capital or on paper; the same index and return rules apply.
| Fund | Sleeve | Status | Index | Return |
|---|---|---|---|---|
| Privacy | Community AIBlong | live | $96.54 | +39.05% |
| Memory | Managed AIBlong | live | $198.02 | +0.91% |
| AMF-WhiteSwan | Managed AIBlong | live | $182.05 | -1.46% |
| Anthropic-Long | Managed AIBlong | live | $100.78 | -3.46% |
| AMF-BlackSwan | Managed AIBshort | live | $111.44 | -3.76% |
| Anthropic-Short | Managed AIBlong | live | $19.92 | -7.95% |
| Data-Protection | Managed AIBlong | live | $150.27 | -9.02% |
| Shipping | Community AIBlong | live | $25.45 | -9.77% |
| Gem Screener Top Picks | Community AIBlong | live | $0.12 | -13.78% |
| DeFi Undercurrent Short | Community AIBshort | live | $6.19 | -28.10% |
| Quantum Frontier Index | Community AIBlong | live | $11.65 | -30.75% |
| Quantum | Managed AIBlong | live | $28.94 | -38.53% |
| Moonshot crypto (high risk) | Community AIBlong | live | $0.21 | -42.04% |
| Crypto Bull Long | Community AIBlong | live | $0.01 | -54.65% |
IPO baskets: SpaceX → Anthropic
What our SpaceX IPO baskets returned, beside the Anthropic IPO baskets now running. Short returns are sign-corrected (a falling basket is a gain).
On Anthropic’s first trading day the short basket swaps to the names from both baskets that ran up the most, and shorts them as they retrace.
Deployment is 45.1 points longer than the 40/60 long/short mandate.
FTA Fund — Cryptos
FTA Fund — Stocks
M&A Ripple Fund
Community AIB Funds
Funds created by EventTrader users on the Fund Launcher and managed by the Tuatara Hedge Fund — the owner describes a theme, the Tuatara vector model selects the hidden-relationship basket, and membership hot-swaps are approved by the fund's owner. LIVE funds take real USDC; funds tagged PAPER trade simulated until they go live and take no investment yet.
Privacy
Wire ·
Kalshi takes down controversial Trump 'mention' markets at CFTC's request
DeFi Undercurrent Short
Wire ·
Ether falls 7% following a multimillion dollar hack of a decentralized finance protocol
Shipping
Gem Screener Top Picks
Wire ·
Kalshi takes down controversial Trump 'mention' markets at CFTC's request
Quantum Frontier Index
Wire ·
IonQ shares rally after company says it made a major quantum computing breakthrough
Moonshot crypto (high risk)
Wire ·
Kalshi takes down controversial Trump 'mention' markets at CFTC's request
Crypto Bull Long
Wire ·
Donald Trump Jr. exits advisory role at Kalshi prediction market
Want your own? Build a fund from any theme → · New to AI agents? Start here →
Tuatara Hedge Fund — multi-strategy
A multi-strategy fund-of-funds powered by the Tuatara vector engine.
- Short — market and regime dependent, overweight
- 2-day to 3-month holds
- Global events / headlines
- Long — market and regime dependent, underweight
- 2-day to 3-month holds
- Global events / headlines
- FTA (Failure-To-Adapt to AI)
- Short baskets, multi-week holds — trading the spikes
- Undisclosed specialized data sources, datasets and data triangulation
- View the live FTA Fund index →
- M&A Ripple (long) — new
- Event-driven long baskets around announced acquisitions, 8-trading-day holds
- Ripple names selected by the Tuatara vector engine, weights set by a genetically optimized parameter set; constituents undisclosed while positions are open
- Managed Tokenized Funds (long, conservative) — new
- Tradable AI-generated Index Baskets (AIBs) with components selected by the Tuatara vector model (Molecular, Memory, AMF-WhiteSwan, Anthropic-Long, AMF-BlackSwan, Defense, Power, Anthropic-Short, Data-Protection, Cybersecurity, Biosciences, Space, Data Centers, AI Compute, Robotics, Quantum)
- Dynamic membership: the vector network proposes component hot-swaps as relevance shifts; every swap requires Human-In-The-Loop approval — long-term holds, minimal churn
weighting: equal_fixed, target_weighted or price_weighted;
most are equal-weight with fixed percentages) and are valued in dollars,
sampled hourly with the live level streaming over WebSocket (~15s); every
series begins at the fund's own inception, with no pro-forma stretch. The
FTA indexes apply today's composition across the full period, the same
construction as their since-inception stat. AIB membership is dynamic: the
Tuatara vector network re-measures each component's relevance and proposes
hot-swaps that execute only with Human-In-The-Loop approval. Short-book and ripple constituents are
withheld — see the FTA Fund page
for why. Long-side execution is via our brokerage partner, referred to
here as Brokerage100. Returns on this page are real, measured
returns on the strategy portfolios since inception: real market
prices, timestamped entries, nothing backfilled. Broker fill counts are
additionally disclosed per sleeve; no fund anywhere fills 100% of every
position — the gap between a strategy portfolio and its executed fills
is the universal implementation shortfall (Perold 1988; Almgren
& Chriss 2001; SEC Rule 605 execution statistics), so a fill count
below the portfolio leg count is how live trade execution works
everywhere, not a caveat on the track record.
M&A RIPPLE — DEALS TRADING NOW
Event baskets the fund is executing live at the broker, marked live. Returns are unrealized percentages on each deal's open basket; constituents are withheld while positions are open. The model track's full holdings, open and closed, are on the M&A Ripple Fund page.
| EVENT | LEGS | TRADE ENTRY (UTC) | P&L |
|---|---|---|---|
| Loading live deals… | |||
📊 Research reports → — source-cited comparisons vs funds worldwide, plus alpha-mining & execution analyses.
▶ How it works — what you're actually buying → See how each deal was found, selected, allocated & executed →Talk to NEXUS — or plug your agent in
Humans: ask NEXUS anything about these funds right here — use the NEXUS chat bar at the bottom of this page. Log in and switch NEXUS on (the toggle beside the chat bar; per-use billing) and it will answer on the page and can walk you through investing, allocating capital across the AIB funds, or building your own fund step by step.
AI agents & developers: the same fund actions are available
programmatically — REST API, MCP tools and single-file SDKs let you
create, store and modify your own custom AIB funds.
API, MCP & SDK docs →
· MCP endpoint POST /mcp/v1
· machine-readable instructions at
/llms.txt
The Tuatara swarm, live
Every control, label and stat. Drag to rotate · scroll or pinch to zoom · hover a node for its metrics · open full screen →
Thought vector
A thought vector is a concept popularized by deep-learning researcher Geoffrey Hinton to describe how a neural network represents the abstract meaning of a piece of data—such as a word, sentence, face, or multi-step reasoning process—as a sequence of numbers (a high-dimensional vector) in a continuous “meaning space”.