FAQ
01Is this real money?
No. The system trades a paper trading account: orders, prices and fills are real, the money is simulated. The point is to check, over several weeks, that live behaviour matches the tests.
02Can I copy the trades?
The site is not built for that and gives no personalised guidance. Trades are published after the close, without quantities or prices, for entertainment and study. Any investment decision is yours and carries the risk of losing capital.
03What is the difference between Normal and Hard?
A · Normal is diversified (stocks, sector ETFs, bonds, gold), with weights that shrink names moving together. B · Hard holds only stocks with strong momentum and high beta: it can rise more and swings more. Details in How it works.
04Why do trades only show up in the evening?
By design: the public page shows the day's activity after 10 pm Italian time, once the market is closed. It is a story of what the system did, not a signals service.
05What do XP, levels and HP mean?
XP tracks gains and losses on sales, the level rises when the portfolio passes value milestones, HP drops as the portfolio falls from its peak. At zero HP for B the emergency brake fires: game over until it is re-enabled.
06Why is the trader asleep?
When the New York Stock Exchange is closed (evening and night in Europe, weekends, US holidays) nothing happens: the trader rests and the monitors switch to a screensaver.
07Can the backtest results be trusted?
They are out of sample: the model was trained only on data before each test period, and costs are included. They are still simulations with known limits, the main one being that B's stock list was compiled today, so it favours names that already rose. See Track record.
08Where do the news come from?
The system reads public news feeds (Yahoo Finance, CNBC, MarketWatch, Benzinga, Seeking Alpha), SEC filings and central-bank and government press releases. A language model writes a one-line summary in its own words; we show that line and the name of the outlet, never the article itself. The tone label (positive, negative, macro) describes the news, not what to do about it.
09Is there AI involved?
The system uses statistical and machine-learning models together with classic rules. In out-of-sample tests the forecasting model added no measurable edge: the result comes mostly from the rules. We say so openly.
10Is real-time access available?
It exists technically, with an access key, but it is not active. Any activation requires a legal review of the rules that apply to this kind of service.
11Do you collect personal data?
No: no accounts, no profiling cookies, no tracking tools. Your browser may only remember local preferences such as sound. Details in Legal.