5 Features Prismfolio Will Never Build
Knowing what a tool refuses to do is as important as knowing what it does. Here are five things we've deliberately decided not to build - and why.
Most product announcements tell you what's coming. This one tells you what isn't.
Saying no to features is underrated. Every feature you add is a promise you make to maintain it, a surface you introduce for bugs, and a signal about what the product is for. These five features have come up - in conversations, in feedback, occasionally as suggestions from people we respect. We're not building any of them.
1. AI portfolio recommendations
The most common request, by a wide margin.
"Can you add an AI that tells me what to buy?" "Can the insights suggest rebalancing moves?" "Can you add a GPT-powered advisor?"
No.
Here's the problem: a model that says "consider reducing your tech exposure" is giving investment advice. It doesn't matter that the model is a language model instead of a human - the regulatory analysis is the same, and the liability is real. Telling someone what to do with their portfolio is something that requires a registration, disclosures, and legal structures we don't have and don't plan to build.
But beyond the regulatory issue, there's a product reason too. The "AI recommender" feature turns Prismfolio from an analysis tool into a decision-making tool. And a decision-making tool that's wrong - that confidently suggests a rebalance right before a recovery, or misunderstands someone's tax situation - can do real damage.
We show you what your portfolio looks like. The decisions are yours. We're not going to muddy that by adding a system that pretends to know better.
2. Buy/sell suggestions
Closely related to the above, but worth naming separately because it comes up in a different framing.
"Can you show me what trades would improve my portfolio?" "Can you tell me which positions to trim?"
Specific securities recommendations - "sell QQQM and buy VTI" - are clearly investment advice under both SEC and FINRA definitions. There is no clever technical framing that changes this.
The Health Score and the insights deliberately stop short of this line. They describe what we observe - high tech concentration, above-average expense ratios, low international exposure - without telling you what to do about it. The "not investment advice" disclaimer isn't legal boilerplate we're hiding in the footer. It's a factual statement about what the tool does and doesn't do.
3. Performance prediction
"Can you show me projected returns?" "Can you tell me if my portfolio will beat the market?"
No.
The SEC's guidance on performance predictions is clear, and frankly, the modeling required to make a non-misleading projection is prohibitively complex. Any future return projection would either be meaninglessly hedged ("past performance doesn't guarantee future results") or misleadingly specific ("based on your allocation, expected annual return is 7.2%").
What we plan instead: historical stress scenarios (sensitivity to past market shocks applied to your current allocation) - not retirement Monte Carlo or return forecasts. Those scenarios describe sensitivity, not prediction. The full historical scenario pack is on the roadmap; we will not ship predictive return projections.
4. Auto-rebalancing trades
"Can you automatically rebalance my portfolio?" "Can you connect to my brokerage's trading API and execute the rebalance?"
Absolutely not.
Placing trades in someone's brokerage account - even with permission, even with the best intentions - is a different category of risk. One infrastructure bug, one bad API call, one race condition, and someone has the wrong trades in their account. The consequences are real and irreversible.
We will never execute trades. We will never connect to a brokerage's trading API. The extension reads your positions page. It doesn't touch the order entry flow. Full stop.
If you want automatic rebalancing, there are robo-advisors built specifically for that with the appropriate regulatory structure and operational controls. That's the right tool for the job.
5. Selling your data
This one might seem obvious, but it's worth stating explicitly because the business model of most free financial tools is: your data is the product.
Free credit score apps commonly monetize your data, sharing it with lenders who want to market to people with specific credit profiles. Free portfolio trackers commonly monetize your data, sharing it with financial institutions that want to reach investors with specific asset levels. Free budgeting apps commonly monetize your data, aggregating your transaction data and sharing it with research firms.
This is legal. It's disclosed (somewhere in the privacy policy). It's how the economics work when the product is free.
Prismfolio's economics work differently. The free tier is genuinely free - the extension costs us a small amount to support and we don't monetize it with data. The Plus plan is $99/year list price (founding pricing is currently available). That's the business model. Your portfolio data is used to provide you with portfolio analysis. That's it.
We don't sell it. We don't share it with advertisers or financial institutions. We don't have an "insights" product we license to hedge funds built on aggregate user data. The privacy policy says "we don't sell your data" and that's factually accurate.
Why this list matters
The features we don't build define what kind of tool Prismfolio is as clearly as the features we do build.
We're an analysis tool for self-directed investors who want to understand their portfolios. We're not a robo-advisor. We're not a trading platform. We're not an AI that makes decisions for you. We're not a free product that monetizes your data.
That's the tool. The constraints aren't gaps - they're the design.
For research and context only - not investment advice. The above describes our current product decisions and may be updated as the regulatory and product landscape evolves.