Simulated Portfolio Live Beta

Compare the system AI portfolio against your own custom portfolio. Use the toggle to switch instantly between them.

Experiment Logic

Capital Capped: $100,000 (Initial)
Allocation: Tiered on AI Conviction

Strategy family

Total Equity

Total Return

Invested Capital

Active Positions

Strategy Comparison

Portfolio Stats

Unrealized Profit --

Realized Profit --

Win Rate

--%

Utilization

--%

Top Holdings

Current Holdings

Company	Entry Date	Shares	Entry Price	Current Price	AI Score	Return
Loading portfolio data...

Recent Transactions

How this trading bot works 1 / 4

What is this bot doing?

Every morning at 04:00 UTC, our AI engine reads the previous day's biotech news and assigns each covered company a sentiment score from 0 to 10 across three pillars: financial health, technology / pipeline progress, and regulatory news. A few minutes later, this trading bot wakes up, looks at every score, and decides what to buy, sell, or hold.

There is no human intervention, no insider information, and no chart-pattern guesswork. The bot is a transparent proof-of-concept for sentiment-driven trading: if the sentiment engine has real predictive value, this bot's $100k simulated portfolio should beat XBI over time. If it doesn't, you'll see that too — every trade and every miss is on this page.

No real money is at risk. This is a paper-trading experiment. Nothing on this page is investment advice — see the disclaimer footer.

The three risk strategies

The bot ships with three risk presets. The showcase you see above runs the Balanced preset. When you spin up your own paper portfolio (see below) you pick whichever profile matches your tolerance.

Conservative

Slow & steady. Capital preservation first.

Min score to buy: 7.5 (high conviction only)
Risk per trade: 0.25% NAV (~$250 on $100k)
Hard stop: -5%
Trailing stop: -6% from peak
Cash buffer: 20% always held back
Max positions: 12
Regime halt: XBI -5%

Best for: lower volatility, fewer trades, smaller drawdowns.

Balanced

Showcase

Middle-of-the-road. The default the public showcase runs.

Min score to buy: 7.0
Position sizes: $1k / $2k / $4k by conviction tier
Hard stop: -7%
Trailing stop: -8% from peak
Take profit: +20%
Max positions: ~20
Cash buffer: ~10%

Best for: getting a sense of what the strategy does under normal market conditions.

Aggressive

Chase momentum. Wider stops to ride volatility.

Min score to buy: 6.75 (lower bar)
Risk per trade: 0.85% NAV (~$850 on $100k)
Hard stop: -10%
Trailing stop: -10% from peak
Take profit: +30%
Max positions: 25
Cash buffer: only 5%

Best for: maximising upside in a friendly tape; accepts deeper individual drawdowns.

All three share the same signal (the AI sentiment score) — they differ only in how strict the entry filter is, how big each position is, and how quickly they cut losers. The same news flow can produce wildly different outcomes depending on which preset you pick.

Reading the dashboard

Here's what every number on this page means.

KPI tiles (top row)

Total Equity — what the portfolio is worth right now: open positions + cash + realised profits.
Total Return — % gain or loss since the $100k starting capital.
Invested Capital — dollars currently tied up in open positions; the rest is cash waiting for the next buy signal.
Active Positions — how many distinct companies the bot currently holds.

Holdings table

Entry Date — when the bot bought.
Shares — quantity held.
Entry / Current Price — what was paid vs the latest quote.
AI Score — today's sentiment score for that company. If it drops below the strategy's broken-thesis floor, the bot exits.
Return — unrealised gain/loss on that single position.

Performance chart

Total Equity tracked daily since inception. Compare against the dotted line (XBI biotech ETF) to see whether the bot is beating the benchmark or just riding the sector tape.

Transactions

Every buy and sell is logged with the reason — "Take Profit", "Trailing Stop", "Thesis Broken", etc. Click into one to see the score and price snapshot the bot used to make the decision. This is the audit trail.

Use the System / My Portfolio toggle at the top to switch between the public showcase and your own paper portfolio (once you've enabled it — see the next section).

Run this strategy in your own paper portfolio

You can have the bot mirror every showcase decision into a private $100k paper portfolio scoped to your account. Your equity curve, holdings, and KPIs sit alongside the public ones — and you control the risk preset.

Sign up or log in — free, takes a minute. Create an account or sign in.
Hit the My Portfolio tab at the top of this page, then click Enable Paper Bot Mode. We seed your portfolio with a fresh $100k — clean equity curve, no inherited positions, so your KPIs are comparable to the showcase.
Pick a risk preset: Conservative, Balanced (default), or Aggressive. You can change it later but every change starts a new "regime" for KPI reporting so the comparison stays honest.
Sit back. Every morning when the showcase bot trades, the same decisions are replayed into your portfolio (resized to your preset's risk budget). You'll see your own holdings, transactions, and equity curve update automatically.

Bonus: mirror to a real broker paper account

Want to see the trades land in a real broker's interface? Connect a Trading 212 demo account (paper money, no real funds) via the Broker Integrations panel below the KPIs. Generate an API key in T212 → Settings → API (Beta) → Generate, paste it once, and every bot decision will also fire as a market order in your demo account.

Your API key is encrypted server-side via Google Cloud KMS — we never see the plaintext.
Sync is opt-in and per-trade idempotent; you can toggle it off at any time.
Trading 212 live trading and Interactive Brokers integration are coming later — for now both have "Open in T212 / IBKR" deep-link icons next to every ticker so you can act manually.

Reality check: this is a public experiment, not a recommended trading system. The bot has a short live track record. Even when it works, past performance is not a guarantee of anything. Treat it as a window into how AI-driven sentiment analysis could drive execution — not as financial advice.