Is there a way to compare two private portfolios (using one as a benchmark?)

Is there a way to use my portfolios as the benchmark field for optimizing another portfolio?

Comparing two portfolios?

Hello people!  I just subscribed to the Plus level because it offered "Bring Your Own Data," which I understood to mean Capital Market Expectations such as user-defined expected returns, volatility and correlations, but I don't see where we can enter this data.  Am I correct that this tool only utilizes historic data and we cannot enter institutional forecast data?

What is "Bring Your Own Data"

We are excited to announce the latest addition to our suite of portfolio optimization tools — the [Hierarchical Risk Parity (HRP) Portfolio Optimizer](https://portfolioslab.com/tools/portfolio-optimization/hierarchical-risk-parity). This new feature complements our existing range of portfolio optimizers, offering you a more diverse set of options to fine-tune your investment strategy.

Based on advanced graph theory and machine learning algorithms, the HRP approach provides a sophisticated method for asset allocation. It leverages the hierarchical clustering of assets, based on their risk characteristics, to create a portfolio that maximizes diversification. In many cases, the HRP method has been shown to produce more robust outcomes compared to traditional risk parity portfolios.

Explore this advanced tool today!

Introducing Hierarchical Risk Parity (HRP) Portfolio Optimizer

Do you count dividend payments in portfolio optimizations?

Your broker shows your portfolio return, but few show where it came from. A 20% return could come from a single lucky bet or from growth spread evenly across all your holdings and that's a big difference. The same return can rest on very different foundations: broad participation across holdings, or one position explaining most of the result. We added **Portfolio Growth Contribution** to [Portfolio Performance](/tools/portfolio-performance) to show which one you're actually holding.

The new waterfall chart breaks total portfolio growth into the contribution from each holding. Positive bars add to growth, negative bars detract from it, and together they sum to the portfolio total.

<img height="417" width="633" src="https://portfolioslab.com/static/user-content/1/wp8yg0bzl829lvqo47nvwxiq.png" />

*The portfolio returned +15.39% over 5Y. VTI contributed +19.04% while TLT dragged −14.77%. Without contribution, you'd never see that a single bond position nearly erased all your equity gains.*

Contribution is different from a holding's standalone return. It combines a position's performance with its weight in the portfolio and accounts for drift and rebalancing. Hover over a bar to see its average allocation for the selected period. Switch between YTD, 1Y, 5Y, 10Y, and Max to see whether the drivers changed over time.

It also separates a weak holding from a real portfolio drag. A single position can plunge and barely move the result because its allocation was small; another can fall less but matter far more because it was a core holding. You see what actually moved your portfolio before making the next allocation decision.

**Availability**

Portfolio Growth Contribution is live now for standard portfolios in [Portfolio Performance.](/tools/portfolio-performance) YTD and 1Y views are available to Free users. Plus, Pro, and Max users can also compare 5Y, 10Y, and full-history results. The contribution chart supports only static portfolios at release.

PortfoliosLab's new waterfall chart shows how much each holding contributed to your total portfolio return, so you can see what actually drove or dragged performance.

See What's Driving Your Portfolio Returns

PortfoliosLab pages can be dense. A symbol or portfolio analysis may include returns, drawdowns, risk-adjusted ratios, benchmark metrics, dividends, correlations, and much more. That is useful data, but it is not always easy to interpret or turn into a clear next question.

We added **Copy for AI** to make that workflow simpler.

### What we added

Symbol pages, portfolio pages, and stock comparison pages now include a **Copy for AI** button. It turns the current page into clean markdown context that you can paste into ChatGPT, Claude, or another AI chat.

<img height="207" width="605" src="https://portfolioslab.com/static/user-content/1/haewimuuv5azrwaswl9yd4uy.png" />

The copied output is structured for analysis rather than screenshots. It includes labeled sections, exact metric names, date ranges, benchmark settings, and tables where the data needs to be compared side by side.

That matters because AI tools are much more useful when they are grounded in the actual numbers. Without providing exact context to AI, you can never tell if it's guessing or not. If the model does not know the benchmark, time period, risk-free rate, drawdown history, or correlation data, it may fill in the gaps. So instead of asking a model to infer from a screenshot or a few manually typed metrics, you can give it the full analytical context and ask a sharper question.

It also makes PortfoliosLab analysis more interactive. You can start with the platform's calculations, then use AI to explore the interpretation: what changed, what looks fragile, what is redundant, and what scenario would hurt the portfolio most.

### What you can ask

Copy for AI is useful when you want a second pass on the data. You can ask it to explain a metric, challenge a thesis, compare trade-offs, or stress-test a portfolio against a specific scenario.

For example, after copying a semiconductor ETF page, you might ask why a one-year Sharpe ratio looks exceptional while the all-time Sharpe ratio is much lower. After copying a portfolio, you might ask whether ten holdings are genuinely diversified or mostly one sector bet. After copying a stock comparison, you might ask whether the extra return is worth the extra drawdown and volatility.

<img height="326" width="637" src="https://portfolioslab.com/static/user-content/1/o9fjuge6mof87ntiu34g5gal.png" />

You can also check [documentation](/docs/getting-started/analyze-with-ai) page with prompt examples and sample copied output.

**Availability**

Copy for AI is available on Plus and Pro plans.

Symbol, portfolio, and comparison pages can now be copied as structured markdown context for ChatGPT or Claude

New: Copy PortfoliosLab Pages for ChatGPT or Claude

Portfolio diversification is easy to talk about and hard to measure.

Most portfolios look diversified on the surface. A few ETFs, maybe some stocks, maybe a bond fund. But once you look at how the positions actually move together, the picture often changes. Five holdings can behave like two bets. A "defensive" position can stop hedging when rates move. Two ETFs with different names can be almost the same exposure.

We built [Diversification Analysis](/tools/diversification-analysis) to make that visible.

### What we added

The new tool analyzes your whole portfolio, not just one symbol at a time. It looks at correlations between positions, effective asset count, diversification ratio, portfolio clusters, and low-correlation candidates you can add.

<img height="449" width="655" src="https://portfolioslab.com/static/user-content/1/begncq8why94trqj7u8xrafy.png" />

The first thing you see is a plain-English summary of the portfolio. It names the specific positions that are helping or hurting diversification, and explains the trade-off in normal language. Metrics are useful, but sometimes you need the tool to say: these three holdings are mostly doing the same job.

Below that, the tool groups positions into correlation clusters. This is where redundant holdings become obvious. If a portfolio has several large-cap growth ETFs, they may sit in the same cluster even if their issuer, name, and category labels are different.

<img height="282" width="670" src="https://portfolioslab.com/static/user-content/1/tod87gv27es8zsxsbiwr2s46.png" />

We also added diversification metrics across multiple periods. Effective asset count shows how many independent bets your portfolio really has. Diversification ratio shows how much volatility reduction you get from combining the assets. Seeing those numbers over 1Y, 3Y, 5Y, 10Y, and all-time helps catch drift. A portfolio can become less diversified without any transaction, simply because correlations changed.

### New candidates, not generic lists

The tool now surfaces ETF and stock candidates that historically moved differently from your current holdings.

This is not a generic "top ETFs" list. Candidates are filtered by correlation to the actual portfolio, then by quality. For ETFs we avoid inverse and leveraged products and require meaningful assets under management. For stocks we focus on larger companies and include risk / return rank, so the list is not just "lowest correlation at any cost."

<img height="299" width="672" src="https://portfolioslab.com/static/user-content/1/tm4a40n73cqrxbqjxqxdfzc7.png" />

You can add a candidate directly into the portfolio input, rerun the analysis, and see what changed. That makes the workflow much more practical. Instead of guessing whether a commodity ETF, short-duration bond ETF, or defensive stock helps, you can test it against the current portfolio structure.

### Why this matters for optimization

Portfolio Optimization works better when the input universe is sane.

If you feed an optimizer five highly correlated ETFs, it can still produce nice-looking weights. But the result may be mostly a reshuffle of the same exposure. The optimizer is not magic. It can only choose from the assets you give it.

Diversification Analysis helps before optimization. Use it to remove duplicates, spot clusters, and find candidates with genuinely different return drivers. Then run optimization on a cleaner universe.

This creates a better workflow:

1. Start with [Diversification Analysis](/tools/diversification-analysis).
2. Remove obvious overlap.
3. Add a few low-correlation candidates.
4. Run [Portfolio Optimization](/tools/portfolio-optimization).
5. Validate the optimized result with walk-forward backtesting.

That sequence is more useful than optimizing first and trying to explain the result afterward.

### A few ways to use it

You have a 3-5 ETF portfolio and want to know what is missing. Run Diversification Analysis, look at the clusters, then check ETF candidates. If all holdings behave like equity beta, the tool will usually make that obvious.

You are preparing a portfolio for optimization. Before running [HRP](https://portfolioslab.com/tools/portfolio-optimization/hierarchical-risk-parity), [Risk Parity](https://portfolioslab.com/tools/portfolio-optimization/risk-parity), or [Mean-Variance Optimization](https://portfolioslab.com/tools/portfolio-optimization), use the clusters to remove duplicates and add assets with different behavior. The optimization result should be more stable because the input set is better.

You are reviewing an existing allocation after market conditions changed. Correlations are not fixed. The 1Y view may show that a hedge is no longer doing what you expected, while the longer windows show whether this is temporary or structural.

[Try Diversification Analysis](/tools/diversification-analysis) with your portfolio

New Diversification Analysis shows how your portfolio actually behaves by finding correlated clusters, redundant holdings, diversification metrics, and low-correlation ETF or stock candidates

Platform Updates

Discussions by tag optimization