Arista Networks Tumbles 8.92% as $2.55 Billion Volume Ranks 27th Amid Institutional Shifts Insider Sales and Analyst Volatility

On September 12, 2025, Arista Networks (ANET) closed with an 8.92% decline, trading with a $2.55 billion volume ranked 27th in the market. Institutional activity and insider transactions emerged as key drivers of short-term volatility, with multiple analysts adjusting price targets amid evolving AI infrastructure expectations.

Plato Investment Management reduced its stake by 16% in Q1, holding 33,243 shares valued at $2.6 million. Meanwhile, DSG Capital Advisors and Ibex Wealth Advisors added or increased positions, while Commonwealth Equity Services and Westfield Capital Management also boosted holdings. Institutional ownership now accounts for 82.47% of the company’s stock.

Analyst activity intensified as Needham & Company raised its target to $155, UBS Group and Piper Sandler followed suit with similar upgrades. However, BNP Paribas Exane downgraded the stock to "neutral." A consensus "Moderate Buy" rating persists, with a $138.79 average target price. The stock’s beta of 1.50 and elevated P/E ratio of 59.38 highlight its volatility relative to the broader market.

CEO Jayshree Ullal sold 1.785 million shares in August for $243 million, marking a 23.86% reduction in her holdings. This followed a broader trend of insider sales, with 6.3 million shares sold in 90 days totaling $795 million. The transactions contrast with recent guidance from Arista’s analyst day, where the company projected 20% revenue growth for fiscal 2026 and 70% AI networking revenue expansion, though margin forecasts of 43–45% drew mixed reactions.

To run a precise back-test of a high-volume trading strategy (e.g., buying the 500 highest-volume stocks daily), practical constraints require simplification. Options include using a liquid ETF proxy (e.g., SPY), testing a fixed basket of 10–20 tickers, or providing pre-defined daily ticker lists for multi-asset simulations. Each approach balances accuracy with data feasibility for multi-year analysis.