Salesforce’s 1.35% Rally Can’t Offset 22.9% Volume Drop Falls to 45th in U.S. Trading Activity

On October 2, 2025, Salesforce (CRM) closed with a 1.35% gain, while its trading volume of $1.82 billion marked a 22.9% decline from the previous day, placing the stock at 45th in trading activity across U.S. markets. This performance reflects mixed investor sentiment ahead of key earnings releases from cloud software peers later this month.

Recent developments highlight strategic shifts within Salesforce’s enterprise solutions division. The company announced a partnership with a leading AI infrastructure provider to enhance its Einstein AI platform, targeting mid-market clients. Analysts noted this move could accelerate adoption of low-code tools among smaller businesses, though implementation timelines remain unspecified. Separately, Salesforce’s board approved a revised share buyback program, signaling confidence in long-term cash flow stability despite near-term margin pressures.

To run a robust back-test for “buy the 500 most actively traded stocks each day and hold for one day” we need to pin down several practical details. Please confirm (or let me auto-fill) the following so I can proceed: 1. Universe • Default suggestion: all common stocks listed on NYSE + NASDAQ + NYSE Arca (ex-ETFs, ex-OTC). Is that acceptable? 2. Ranking logic & trade timing • Rank by previous day’s dollar trading volume (shares × close price). • Buy at the next session’s open, exit at the same day’s close (1-day holding period). • Equal-weight each position. • If this timing is fine, I’ll apply it; otherwise please specify. 3. Transaction assumptions • Default commission/slippage: 10 bps per round-trip (5 bps in, 5 bps out). • Cash drag ignored (returns reported gross of interest). • OK to use these? 4. Performance metrics • I’ll deliver cumulative return, annualised return, volatility, Sharpe, max drawdown and turnover. 5. Tool limitation note • Our current tool-chain treats each ticker individually, so to replicate a 500-stock basket I’ll aggregate all signals into a synthetic “portfolio index” on my side, then feed that series to the back-test engine. No accuracy is lost, but it’s important you’re aware of the workaround. If the defaults are acceptable just reply “proceed”, otherwise let me know which items you’d like adjusted and I’ll tailor the test accordingly.