STEP 5 OF 53 min

Broaden the sample. Spread single-stock errors.

Use a multi-stock portfolio to spread differences in algorithm performance while reviewing shared risks and actual results.

Three things to remember

Not live pricesWe show daily closes. Check the Prices as of date.

Ranges have limitsBased on historical data, for range-bound phases only. Historical ranges are not current signals.

Decide independentlyAI and scores do not guarantee returns. Verify quotes with your broker before trading.

HANDS-ON PRACTICEFictional sample · For learning

Individual gains and losses. See the portfolio total.

Each sample stock receives $1,000 and has a fixed gain or loss. Start with one, then add others to see gains offset losses. Selecting all gives a positive total in this example. These fictional outcomes exclude costs and do not guarantee real investment profits.

Select sample stocksDEMO-T1TechnologySample gain/loss: −$80 · -8%DEMO-T2TechnologySample gain/loss: +$100 · 10%DEMO-T3TechnologySample gain/loss: −$60 · -6%DEMO-H1HealthcareSample gain/loss: +$100 · 10%DEMO-U1UtilitiesSample gain/loss: +$100 · 10%

Portfolio total gain/loss−$80

Selected stocks1

Total investment$1,000

Portfolio return-8%

With one stock, the result depends on that stock: it may gain or lose.

Why does selecting all profit in this example?More stocks always guarantee a profitGains from winning stocks cover losses, leaving a positive total

Open your watchlist

APPLY IT IN THE WORKSPACE

Start with a free starter stock

Compare sectors, markets and strategy suitability. Spread individual errors and record all outcomes and costs; a watchlist is not automatic allocation.

Open workspace

Understand the principles

Expand a topic to explore it

Do not depend on a single stock

Algorithms and AI can perform differently across stocks. One success or failure does not establish overall performance. Where affordable, a portfolio of independently researched stocks can spread the impact of a single-stock mistake. Check range-strategy suitability and plan stops and profit-taking for each stock. Broadening the sample is not a reason to trade blindly.

Several tickers are not independent opportunities

Review concentration across sectors, markets and economic drivers. Stocks in the same sector can fall together, and algorithms can make shared errors. Spreading stocks and position sizes seeks to balance single-stock differences and concentration risk; it does not automatically increase the win probability of each trade or the portfolio.

Review the complete portfolio sample

Record all signals used, quote dates, actual fills, position sizes, costs, successes and failures instead of selecting only correct predictions. Compare net portfolio returns, losses and concentration, not just the number of rising stocks. More samples help reveal algorithm differences but do not establish an improved win rate by count alone. Diversification cannot guarantee profits or prevent all losses. Saving stocks to this website's watchlist organizes research, not portfolio allocation.