AI-native quant research
Describe a strategy you already have in plain words. Vaquero's AI agent turns it into executable logic and runs it through a realistic backtesting engine, so you can see what survives history.
Research and backtesting infrastructure. No strategy recommendations, no trade signals, no order execution.
Join the waitlist and extra credits are waiting on your account the day the beta opens.
01The reason
Vaquero began with two problems. Neither is one a person can get past alone.
Timelines fill with how much an automated strategy made, and with images captioned backtest. Neither the data behind them nor the number of re-runs it took to get that one picture is anywhere on the post. Nobody reading alone can tell a real result from a good story, and what is left is the fear of missing out.
People hand a strategy to their own LLM agent now. The history within reach is a few years of closing prices, and the code that comes back reads values from after the decision date or quietly drops every company that got delisted. The curve climbs, and what it measured nobody knows.
Vaquero is a backtesting environment where every strategy runs on the same data, the same costs and the same rules, whoever wrote it. You write the strategy as a sentence and an AI agent built for backtesting turns it into a specification, so there is no code to write. What it takes to start is an idea.
02The process
Vaquero is an AI-assisted quantitative research and backtesting platform for strategies you define yourself. Four steps, and none of them asks you to write code.
You already have the idea. Write it the way you would say it out loud: what it trades, what puts you in, what takes you out.
The AI agent turns that sentence into a structured, executable specification. It does not pick a strategy for you, and where a missing choice would move the result, it asks.
The specification runs over years of crypto and US equity history. What the market charges comes off, and the US equity series is price-only, which the result states alongside the figures.
The equity curve, where the money went, and what the book held on the last day. For a run whose instruments you named, the fills come back with their price and time, and pressing a point on the curve brings back the holdings at that decision with the timestamp of the decision itself.
03The engine
One place to build a strategy, test it against years of crypto and US equity history, and read the markets it trades.
Every idea goes through the backtesting engine before it goes anywhere else, with fees, slippage and funding carried through.
Crypto and US equities, with macro and FX read on the same screen.
Research logic remains visible, editable, and ready to challenge.
04The method
A backtest is a claim about the past. These are the three places that claim usually breaks, and what the engine does about each.
tThe causal boundaryEach trade fades by what execution costs
History runs through the engine from the left. A decision reads only what had closed by this line, and the field past it holds nothing the engine is allowed to use. Every entry and exit the run made lands as a mark on the rail below, in the order it happened, so the whole run can be read back trade by trade.
A decision sees closed bars and the values that were public on that date, and nothing else. Nothing revised later reaches it, so the same run returns the same result.
Fees are a rate on the order, the spread is charged on what it crosses, and funding accrues on the open position at the settlement times the data shows. The push your own order puts into the book is priced separately, against what that day actually traded.
Each result keeps its execution assumptions with the sample length, skewness, and kurtosis measured from its return history. Survivorship comes off the grade as a deduction, and its size is printed with the result.
05 / The desk
An example of a member's own wording. Not a recommendation to buy or sell.
Crypto, US equities and macro on one board.
06The community
No follower counts. No screenshots of curves.
07Questions
Vaquero is an AI-assisted quantitative research and backtesting platform that helps you implement and historically test trading strategies you define yourself.
Vaquero is
Vaquero is not
Vaquero is an AI-assisted quantitative research and backtesting platform. You describe a strategy you already have in plain language, an AI agent implements it as an executable specification, and the engine runs that specification over historical crypto and US equity data.
No. You provide the strategy idea. Vaquero's AI agent helps translate and implement that user-defined strategy for historical testing. Vaquero does not recommend what strategy you should trade.
It reads your description and writes it out as a structured specification: what it trades, the conditions that open and close a position, how often it rebalances, and the execution assumptions. Where a missing choice would move the result, it asks you rather than filling it in.
A decision sees closed bars and the values that were public on that date, and nothing revised later reaches it. Fees are a rate on the order, the spread is charged on what it crosses, and funding accrues on the open position at the settlement times the data shows. The push your own order puts into the book is priced separately, against what that day actually traded.
Each result keeps its execution assumptions with the sample length, skewness, and kurtosis measured from its return history. Survivorship comes off the grade as a deduction and the result prints how large that deduction was, so the bias is a number you can argue with rather than an assumption.
Crypto and US equities, including ETFs. The engine also reads macro and FX series, perpetual funding, Form 4 and congressional filings, US fundamentals and seasonality.
Neither. Vaquero publishes no signals, no picks and no model portfolios, and it is not a broker: nothing you build here places an order or holds your money.
Each run carries its own execution settings: starting capital, the round-trip fee, how much of a session's volume one order may take, the impact model, the benchmark and the cash rate. They are stored with the result, so a run can be re-read on the terms it was made under.
Join the waitlist and extra credits are waiting on your account the day the beta opens.
Vaquero is a research and education service, not investment advice. We recommend no buying or selling, publish no trading signals, and guarantee no return. Credits are the monthly AI usage allowance that comes with a plan. It resets when a new billing period starts and does not roll over.
