Paper Trades

Day — of — — left until final scoring on Nov 26 New York — UTC — —

This is a simulation, not investment advice. No real money is involved. It is a 30-day experiment in which an AI model makes trading and betting decisions using real-world prices, to see how its judgment holds up. Nothing here is a recommendation to buy, sell or bet on anything. This is an independent project, not affiliated with any brokerage, sportsbook or AI company. Final results for all five will be posted on Nov 26, win or lose.

Five separate paper-trading experiments, each run and scored on its own on Nov 26, 2026. Each one started with its own — (US$100,000 for Chinese stocks) and is judged only against that starting amount; the results are never added together. Last updated —.

The big figure is total net profit since the start. "Last 24 hours" compares with the check-in closest to 24 hours earlier, or the start if a simulation is newer. "Worst drop" is the biggest fall from a high point, measured at check-ins. Stocks are compared with putting the same — into the S&P 500, crypto and metals with putting it into Bitcoin, and Chinese stocks with the CSI 300 index. Positive means the portfolio is ahead. The Chinese stocks simulation is counted in US dollars.

Coming up

The next events that could move these simulations. Times are Eastern.

WhenEventAffectsWhy it matters

How each simulation is doing

Pick a simulation. Move your cursor along the chart (or drag a finger on a phone) to see the exact value at each check-in.

Distance to the next trade

Every rule tied to a price, closest first. When a price reaches its level, the AI makes that trade at the next update.

What made or lost money

Profit or loss on each position and settled bet since it was opened, in Canadian dollars. Cash and USDC are left out.

Recent decisions

Every trade, bet and hold decision across all four, newest first. Times are Eastern.

WhenPortfolioActionDetail

About this experiment

This page tracks a 30-day paper-trading experiment, Oct 26 to Nov 26, 2026. An AI model was given — of pretend money in each of four simulations, plus US$100,000 for a fifth in Chinese stocks added on Oct 28, and asked to manage it as it saw fit. The page owner checks in through the month, and the AI looks up current prices and news, decides whether to trade, and records each decision and its reasoning here.

  • Independent. This is an independent experiment, not affiliated with, sponsored by or endorsed by any brokerage, sportsbook, exchange or AI company.
  • Not advice. Nothing on this page is financial, investment, tax or betting advice, and the AI is not a licensed advisor. The picks are an experiment, not recommendations. Do your own research and talk to a qualified professional before risking real money.
  • No real money. No trades or bets were actually placed. Every position and result is hypothetical.
  • Real prices, simplified. Prices come from public sources (exchange quotes, market news sites and sportsbook odds) at the time of each check-in. They may be delayed or differ from prices you could actually get. Fees, taxes, bid-ask spreads and slippage are not included, so real results would be lower.
  • High risk on purpose. The stocks portfolio holds 3x leveraged funds, and crypto, metals and sports bets can swing sharply. These choices were made to test aggressive decision-making, not because they suit most people. Leveraged funds can lose a large share of their value in a single day.
  • Short window, lots of luck. Thirty days is too short to prove skill. A good or bad result here says little about future performance.
  • Sports betting is for adults only. Odds are shown for the experiment, not as an endorsement of any sportsbook. If you bet for real, set a limit you can afford to lose, and seek help if it stops being fun.
  • Results, win or lose. The final results for all five simulations will be posted here on Nov 26, whether they are good or bad. Losing trades and bets stay in the log.
  • How it's scored. Values are in Canadian dollars at the latest exchange rate, except the Chinese stocks simulation, which is counted in US dollars. Each simulation is compared with a simple benchmark (the S&P 500 and TSX for stocks, holding Bitcoin for crypto and metals, the closing line for sports, the CSI 300 for Chinese stocks). US markets are closed on Nov 26 for Thanksgiving, so the stocks score uses the Nov 25 close.

Data last refreshed —. The sources used for prices, odds and news are listed below.

Sources