Directional & Automated Trading with AI: The New Community

Directional and automated trading is the combination at the heart of our renewed community: after many years of hands-on directional trading, our founder Stefano Mastria has finally added an automated, quantitative component — made possible by artificial intelligence — and now shares both inside one research project. In short: we trade a few markets we know well by hand, we let software work many markets and accounts on our behalf, and we let the data, not intuition, tell us what to do.

This is not a beginner’s primer. It is an overview of what the community actually is today and how it works: the two AI engines behind the strategies, the DB Strategie database, research accounts versus production accounts, and the collaboration-through-research philosophy that holds it all together.

What does directional and automated trading mean in our community?

Directional trading is the discretionary craft the founder started with in 2009: you study a handful of markets deeply and manage positions by hand. Automated trading is the newer layer: fully automatic or semi-automatic software that scans, finds and manages opportunities for you across many markets, many accounts and even several prop trading firms. The two are not rivals — they are two faces of one method. As the founder puts it in the video, with automated strategies “we do a job a bit like a money manager”, because we no longer have to hunt for every trade; the system does that, and our work becomes supervising risk, capital and allocation.

The long-standing rule for the directional side has not changed: work on few markets you know very well. What has changed is that the automated side lets us extend that same disciplined logic across a breadth that manual trading could never reach. That is the practical meaning of directional and automated trading here — not a promise of easy gains, but a wider toolset governed by the same principles.

Why did AI unlock the automated side after past failures?

Automation was not a new ambition. Years ago the founder ran a project called “Forex Robot” and later hired traders and developers — in his own words, a lot of money spent and a lot of work, with disappointing results. What finally made the leap possible was artificial intelligence: it let him replicate his directional approach inside automated models that, he says, do a good job. Two AI agents were built for this, both on Claude Code (the video calls it “cloud code”; the correct name is Claude Code).

The first engine is Realistic Engine AI: it takes a base strategy and works on its parameters — risk per trade, monthly drawdown limits, break-even, and so on — as a starting point for research. The second, more advanced, is Natural Law iQuant: it uses pre-opening, morphological-analysis models built on algorithms that the AI itself suggested after studying the data. The first three iQuant models are live and more are on the way. A separate YouTube channel, also called Natural Law iQuant, is where the founder publishes the data and AI work in detail, while the community channel covers the project and what we can do together. Whichever engine is used, the underlying philosophy is the same: the market gives a finite set of events that tend to repeat, and the data — read quickly and accurately with AI — tell us the realistic expectation for a given market and strategy. Nothing more, nothing less.

“It is a data-driven business: whether it is directional or automated, it is always based only on the data.”

Few markets or many markets: the money-manager role

One of the clearest ideas in the video is the shift in the trader’s role. In pure directional work you cannot realistically follow dozens of instruments — depth beats breadth, so you pick a few and know them cold. Automated strategies invert that constraint: because the software is fully or semi-automatic, the opportunities are found and managed for you, so you can run many markets, many accounts and several firms in parallel. Your job moves up a level, toward money management: choosing which strategies run where, sizing risk, and watching the numbers rather than chasing entries. Prop trading and leverage widen what is reachable, but they cut both ways — they amplify losses as much as gains — so the daily discipline of managing small, controlled costs stays exactly where it was.

What is the DB Strategie, and how do research and production accounts differ?

The DB Strategie is the strategy database at the centre of the automated side. It currently holds around a hundred automatic and semi-automatic strategies, each built with StrategyQuant and then optimized with AI — described by the founder as the “best of the best” selected out of several million elaborated first- and second-generation strategies, with a third generation in the works. They are available to base and premium members alike so they can experiment, test and learn to use them. Each strategy is assigned to a limited number of users, a deliberate choice to avoid many people running the identical system on the same broker. The backtests are computed on roughly 9.5 years of Darwinex data — a long window chosen precisely because currency markets tend to repeat past behaviour, so a wide history exposes the natural problems a strategy must survive before any profit appears.

Around those strategies, every member can open two kinds of accounts. Research accounts are demo or small live accounts used purely to gather data and analyse behaviour. Production accounts are real accounts, or accounts held with Darwinex or a prop trading firm. Every account gets its own analysis page — maximum drawdown, maximum profit, best and worst days — connected in real time to MetaTrader. Remarkably, the founder built the whole platform with AI: “I understand nothing about software development, and I built it all with AI.” The more data the research accounts gather, the more raw material there is to improve the trading — which is exactly why the community pools it.

A worked backtest example (and why it is only a starting point)

In the video the founder walks through one strategy on GBP/JPY (numbered 2.1.165). He shows that raising risk from 0.2% to 0.5% per trade, with a 1% monthly drawdown limit and a 1% break-even, over the full 9.5-year window changes the equity curve markedly on paper. He is emphatic about the caveat, and so are we: this is an accurate backtest simulation produced on hedge-fund-grade computing power, not a homemade number — but it is still only a starting point. A backtest describes how a strategy behaved on past data; it does not guarantee real or future results. His own workflow makes the point: he only launches live research on the backtests that already look promising on paper, and discards the rest. The 2026 backdrop he mentions — the war between the US and Iran, the war in Ukraine — is a reminder that real markets throw up problems that have to be managed, which is educational in itself.

“No one holds the truth in their pocket that just produces on its own — results are earned with the work of research.”

The research-collaboration philosophy — and how you take part

The founder is blunt about the nature of the project: it is a collaboration built on research, not a signal-selling shortcut. He shares the steps forward he makes with his own trading and the tools behind it; members can research alongside him, choose strategies, take care of them, optimize them and feed the resulting data back into the community. Results, he repeats, are earned through study, research and practice — the only route he knows. That frankness is deliberate, and it is the healthiest way to read everything else on this page.

Concretely, the reserved area is organised around a dashboard with daily to-do indications that update as you complete them, live sessions on Monday, Wednesday and Friday inside work groups, a welcome video, and a base-versus-premium structure. Progress is tracked through formative credits: the more you study, the more resources you unlock — importantly, credits change your access level, not the price. The founder also shows the real-time balance of his own directional and automated accounts, framed as his own results and not a promise, alongside what he calls the two balances every trader keeps: “an economic balance and an experience balance.” Premium adds the advanced Natural Law iQuant strategies, the Realistic Engine editor (monthly credits for private optimizations) and one-to-one coaching. Around all of this sit the Power Tools — a “Quick Orders” tool for MT5 and Pine Script indicators — personal API codes to activate quant strategies on MT5, and an “Ideas / Algo Funds” section for combining strategies in synergy, such as a “Fondo Nettuno”. None of it is sold as a guarantee; it is a workshop for doing the research together.

A final note on scale and honesty. The founder invests about $5,000 a month in high-quality computing power — the kind of spend a speculative fund makes — and is weighing a larger budget of roughly $60,000 to $120,000 a year for third-generation strategies, even seeking support from providers like Google and Amazon AWS, because, as he says, “I am not a millionaire hedge fund.” He puts that on the record precisely because directional and automated trading, done properly, is expensive, slow research — not a machine that prints money.

Frequently asked questions about directional and automated trading

What is directional and automated trading?

It is the combination our community now works with: directional (discretionary, manual) trading on a few well-known markets, plus automated (quantitative) strategies that a software runs across many markets and accounts. The two share the same risk-first, data-driven logic; automation simply extends the reach and turns the trader’s role into something closer to money management.

How did AI change the community?

Artificial intelligence made it possible to replicate the founder’s directional approach inside automated models after earlier attempts had failed. It powers two engines built on Claude Code — Realistic Engine AI and the more advanced Natural Law iQuant — and it lets the team read large amounts of market data far faster and more accurately, for both the directional and the automated side.

What is the DB Strategie?

The DB Strategie is the community’s strategy database: around a hundred automatic and semi-automatic strategies built with StrategyQuant and optimized with AI, backtested on about 9.5 years of Darwinex data. Members can experiment with and test them, and each strategy is assigned to a limited number of users to avoid duplication on the same broker.

Is this a get-rich-quick scheme? Does it promise profits?

No. Nothing here is a promise of profit or return. Every figure mentioned — including the GBP/JPY backtest — is historical or simulated backtest data, and backtests do not guarantee real or future results. The project is explicitly a research collaboration: results, in the founder’s own words, are earned through study, research and practice, and trading carries a concrete risk of losing capital.

Is this article investment advice?

No. It is for informational and educational purposes only and reports the method and personal opinions of our founder as presented in a video. It is not financial advice nor a solicitation to trade, and every decision remains the reader’s own responsibility.

The contents of this article are for informational and educational purposes only and report the method and personal opinions of the founder of Trend Following Traders as presented in a video on the channel; they do not constitute financial advice nor an invitation to trade. All figures cited — backtest results, percentages, risk and drawdown levels, the GBP/JPY example, computing budgets and account balances — are historical, simulated or contextual data: backtest results do not in any way guarantee real or future results, and the founder’s own account balances are shown as his personal results, not as a promise to anyone else.

The references to the community DB Strategie, to the quant strategies, to the Realistic Engine AI and Natural Law iQuant engines, to optimization with artificial intelligence and to prop trading describe research and production tools, not promises of return. No profit is guaranteed: trading — directional or automated — and the use of financial leverage and prop trading carry a concrete risk of losing capital.