I'm quite new to the site, but I am concerned that a portfolio optimized with past data may have no bearing at all on its future performance. Has anyone been around long enough to speak to this concern. Have you outperformed a relevant benchmark with actual invested money?

Also, if you've been here awhile, what tools on the site do you find most useful?

Thanks for reading!

Based on my personal experience, it's important not to take the numbers from portfolio optimization algorithms too literally. They should be seen more as rough estimates, helping you to identify overallocation in your portfolio and guiding you towards better balance. If you're optimizing for maximum return, the resulting portfolio will likely underperform out-of-sample. However, if your goal is to achieve a smooth performance line with minimal drawdowns, and your portfolio is already well-diversified, then optimization can yield solid results – numbers that can be used as a reliable guide.

As a side note, I think it would be fantastic if PortfoliosLab could add a feature to the portfolio optimizer that allows for calculating allocation as of a specific date in the past to see how that allocation would have performed after that. Or an option to run the optimizer periodically, like every quarter, would be extremely interesting. I'd certainly experiment with that to see how dynamic rebalancing could impact performance over time.

Thanks BullRun. Yes, I'm really trying to get a decent Sharpe ratio going forward and decent returns. I particularly like this site's feature to do risk parity position sizing. Intuitively, you're getting better diversification with risk parity sizing, and studies HAVE shown better absolute and risk-adjusted returns with risk parity sizing vs equal sizing or market cap sizing.

And yes, BullRun, that's a great idea. You really want out-of-sample testing to validate your modeled portfolio.