Evaluating Canadian Online Brokerages: Beyond the Ranking Headlines

Beyond the headline

The captured metadata confirms publication date and outlet only. Source text was not retrievable at ingestion. Specific brokers, rank positions, fee tables, and scoring methodology remain unverified against this feed. Treat the roundup's existence as an indexing event — not a recommendation — until the methodology and the full text are inspected directly.

For Canadian traders, any "best of" list is a shortlist filter. The mechanical layer beneath the ranking decides whether the shortlist is worth reading at all.

What rankings tend to skim

Editorial comparisons default to surface-level attributes: commission headline, mobile rating, support hours. The execution-layer numbers that actually move P&L rarely appear in editorial scoring:

• Order routing. PFOF disclosure, smart routing across Canadian venues, and whether the router prefers maker rebates or price improvement.
• DOM depth and charting stack. Adequate Level 2 exposure, charting that does not stutter on tick refresh.
• API endpoints. REST and WebSocket availability, documented latency under load.
• CAD/USD spread. Hidden cost on US-listed positions for CAD-funded accounts.
• Inactivity and admin fees. Dead money that erodes small accounts quickly.

Filter before switching

Until the Forbes full text is accessible:

1. Pull the commission schedule for each shortlisted broker. Decompose total cost on a projected trade volume, not the advertised headline rate.

2. Open a paper or small live account. Measure slippage on market orders at the 9:30 ET open and during the midday liquidity gap.

3. Verify account protection limits and the regulated entity operating the account.

4. Pull the broker's outage incident log over the trailing 12 months.

5. Confirm API documentation, rate limits, and historical uptime SLA before connecting any systematic layer.

A ranked list tells you which brokers got screened. It does not tell you which broker your execution path will survive on. The mechanical test is yours to run.