Observatory

The problem

Retail-investor dashboards are designed by financial media and trading platforms, both of which monetize attention. When markets get scary, the visualizations get more dramatic: bigger red, more blinking, more exclamation points. The UI is calibrated to keep you watching, not to help you think.

That’s the opposite of what a long-horizon retail investor needs. Reactive investors trade more. Trading more costs money in fees, taxes, and the gap between what you bought and what you should have held. The standard design pattern for financial dashboards is actively harmful to the people using them.

The specific problem: nobody builds for Howard Marks readers. For people who understand that volatility is an information-free event. For investors who want a regime-first read on the macro environment and enough signal to know when their thesis is being tested, not when to panic.

Who it serves

Long-horizon retail investors who want to understand macro regimes and make informed allocation decisions, not get told what to buy. People who follow Dalio’s economic-machine framework, read the Bridgewater memos, and want one honest view of where the world is before they decide anything.

Not traders. Not people who want alerts. People who check in once a week, want to understand the regime, and leave without feeling like they missed something.

How it helps

The design insight is behavioral, not technical. Stress in financial markets increases the cognitive load of the people watching them. More information, more noise, more visual urgency creates more reactive behavior. So the counter-intuitive move is to reduce information as volatility rises.

As macro stress rises, the dashboard gets calmer: fewer flashes, more white space, longer baselines, signal-aware context that filters noise before it reaches the UI. The data is the same. The framing is honest. Legal-by-design with no BUY/SELL/RECOMMEND language anywhere. Factual statements only.

The behavioral bet: if the visual environment doesn’t amplify the emotional signal of a volatile market, the investor makes better decisions. The design is trying to be the Howard Marks memo, not CNBC.

The stress-calibration logic lives at the data layer, not the display layer. The dashboard doesn’t decide what’s stressful and then change colors. It reads regime signals (yield curve, credit spreads, volatility indices) and adjusts information density accordingly. The decision happens upstream. The display is just honest about what the data says.

What it does

A regime-first dashboard pulling FRED, ECB, Eurostat, CoinGecko, and Twelve Data into a single view. Sparklines, percentile bars, signal-aware context panels. Dark theme only. JetBrains Mono for financial data throughout: the same typeface across all my tools, the same signal that this is a precision instrument.

Cloudflare Worker proxies all data sources so API credentials stay server-side and the client stays simple. Vite, React 19, Tailwind 4. Custom design system with no component library. Everything is purpose-built for the data display problem, because generic components make assumptions about information hierarchy that financial data doesn’t fit.

What I learned

The hardest part wasn’t the data layer. It was deciding what not to show during volatile windows.

Most dashboards add information when stress rises. The instinct is: if something is happening, give the user more data. The design point here was the opposite: when something is happening, the user needs less noise, not more signal. The threshold for what earns a display slot gets higher, not lower. Removing things from view is a design decision. It’s also a systems-thinking decision: you have to trust that the regime signal is sufficient, and that an investor who sees “elevated stress: reduce noise” will act better than one who sees eight blinking charts.

The second lesson was about legal design. Building a personal-finance tool without triggering securities regulations required designing around the problem rather than through it. “Here is the regime signal” is categorically different from “here is what you should do.” The difference isn’t just legal. It’s actually better product design. The user who gets a factual regime read and draws their own conclusion is more engaged than the user who gets a recommendation they didn’t make. Agency in the analysis produces better decisions than confidence in someone else’s.

The design philosophy here is the same one that runs through Floodplain and through Valet’s trust tier system: show the information, surface the trade-off, leave the decision with the human. Don’t be an authority. Be a precision instrument.

Status

Active thesis project. The behavioral-finance framing holds up the more I look at it. The design ethos is the public deliverable for now; the implementation is ongoing.