Do automated strategies need fractional shares?

No, but sizing precision suffers without them. A strategy allocating '98% of a $5,000 sleeve' to a $700 ETF can buy exactly 7 whole shares ($4,900) — close enough. The friction appears with expensive tickers and small sleeves: a $1,000 sleeve targeting a $700 ETF holds one share and 30% idle cash, distorting results versus the backtest's assumption. Whole-share rounding also matters for rebalancing strategies, where each adjustment quantizes. Where the broker supports fractional orders the platform can size exactly; where it doesn't, prefer cheaper share classes of the same index (QQQM over QQQ-priced alternatives, VOO vs higher-priced peers) or size sleeves at least ~10x the share price.