Euclidean Fundamental Value ETF

Euclidean Fundamental Value ETF (ECML) is an actively managed fund from ETF Architect. It holds about 60 to 70 US value stocks chosen by machine-learning models built by its sub-adviser, Euclidean Technologies Management. Unlike many value funds, it leaves out financial companies, gives each stock a roughly equal weight and screens out likely "value traps." The trade-off is cost: a 0.95% annual expense ratio, or about $97 in the first year on $10,000.

• Over the year to Oct 27, 2026, ECML returned +22.5%.
• Its worst full calendar year was 2024 (+2.4%); its best was 2025 (+6.8%).
• The deepest fall in its price history was 24.7% (Nov 25, 2024 to Apr 8, 2025), recovered by Jan 14, 2026.

As of Jun 30, 2026, the fund held 62 securities; the ten largest made up 22.12% of assets.

• Over the three months to Jun 30, 2026, investors withdrew a net $12.6M (creations minus redemptions, per the N-PORT report).
• Monthly total returns reported for that quarter: +4.92%, -1.65%, +3.12%.

ECML does not track an index. Its sub-adviser, Euclidean Technologies Management, starts with every stock listed on the New York Stock Exchange and Nasdaq. It then removes non-US companies, financial companies and any company worth less than $1 billion, which leaves about 1,600 names. A neural network forecasts each company's earnings for the next year. A neural network is a machine-learning model trained on past data, here including financial statements, short interest and past price moves. The forecast is then cut back to reflect how uncertain it is.

Each company is ranked by "earnings yield," which is the reduced forecast divided by enterprise value. Enterprise value is a company's stock market value plus its debt, minus its cash. A second model estimates how likely each cheap stock is to be a "value trap": a stock that looks inexpensive but is likely to land among the worst performers the next year. Likely traps move down the list. The managers then review the top 60 to 70 names and can drop a company over problems the model can't see, such as a bankruptcy filing or restated financial results.

Stocks start at equal weights of about 1.5% to 2% each and can grow to a cap of about 10%. The portfolio is rebalanced every quarter, and more often after major shocks. At least 80% of assets go to companies that look cheap on earnings-to-price and book-to-market ratios.

• ECML charges 0.95% a year.
• Using the prospectus's standard example, $10,000 invested would cost about $97 in the first year and $1,166 over 10 years (assuming a 5% annual return).

ECML is a concentrated, actively managed value strategy with a 0.95% fee and about three years of history, so it fits better as a smaller side position than as the base of a portfolio.

ECML launched in May 2023, so it has no record through the 2008 financial crisis, the 2020 COVID crash or the 2022 bear market. Its deepest drop was 24.7%, from November 25, 2024, to April 8, 2025. It took about nine months to recover and was back at its old high on January 14, 2026. A retiree who sold shares for spending money during that stretch would have locked in losses. Its weakest calendar year, 2024, still gained 2.45%. Income is small: the trailing yield was 1.15% as of October 27, 2026, paid once a year. Costs are high. Under the SEC's standard example, $10,000 would pay $1,166 in costs over 10 years. Past results don't predict future returns, and a retirement plan should look at the whole portfolio, not one fund.

Matched by holdings overlap from the latest N-PORT reports.

No close alternatives found.