We Really Wanted the Stars to Pick Stocks
We built a financial astrology app, and then we did the thing you are not supposed to do: we checked whether it works. Here is everything we found, including the parts that flopped. Especially those.
The question we could not leave alone
If you build software that turns birth charts into portfolios, at some point, late at night, you ask the obvious question. Does the chart part actually do anything?
Most apps in this genre never ask. We asked repeatedly, over more than a thousand pre-registered experiments, with the kind of statistical controls normally reserved for people who manage other people's money. Not because we are joyless. Because we wanted to know, and we figured you might too.
The stranger's chart test
The experiment that settles it is almost rude in its simplicity. Take a real person's birth chart and run it through the full engine: derive their sectors, pick their assets, choose their buy days, invest a fixed amount monthly for years of market history. Then do the exact same thing with shuffled charts, random birthdays in random cities, strangers who do not exist. Same money, same dates, same rules.
If the chart carries information, the real chart should reliably rank near the top of that lineup. Ours did not. Across hundreds of shuffled charts and every method we offer, the real chart landed around the middle. A coin flip.
We ran the full product path too, month by month, three real people including Warren Buffett, the one famous investor with a reliably recorded birth time. His own chart ranked dead last among sixteen strangers' charts. The man himself is fine, we suspect.
The 9.9x result we did not write home about
Here is the part that taught us the most. In one grid of experiments, one cell lit up: Buffett's chart, read through the Kabbalah lens, over 2022 to 2024, returned 9.9 times the money put in. If we were a different kind of company, that number would be the billboard.
But we had run the same grid on the shuffled charts, and random nobody-charts were posting 7x results in the same space, because concentrated five-stock portfolios in a bull market produce fireworks for anyone. One dazzling cell out of seventy-two is what that kind of grid hands out for free. So we wrote it down, filed it under noise, and moved on, a little wistfully.
What actually survived
Something did, and it surprised us by being useful.
The lenses genuinely change the shape of the ride. In our replays, the chart-derived portfolios often leaned defensive: one subject's account dipped about 13 percent at its worst while every benchmark dipped around 23. And the effect has structure. Across every sector tradition we offer, the full composite scoring method produced shallower drawdowns than the synastry-only method, six out of six, no exceptions. Lens choice moved worst-case dips across a range from about 14 to 28 percent.
That is not a return advantage, and we are careful to never call it one. It is a risk texture, and it holds regardless of whose chart is driving. Different lenses build calmer or spicier portfolios, consistently, and you get to pick your temperament.
The other survivors are quieter still: the discipline of a schedule, the self-knowledge of seeing your habits described, and the small ritual of checking in with a plan instead of a feed.
Why tell you any of this
Because you should know what a tool is before you lean on it. This one is a mirror with excellent craftsmanship: real ephemeris math, real classical sources, every derivation traceable to a named rule. The mirror does not know what stocks are about to do. Nobody's mirror does.
We wanted the stars to pick stocks. They declined. What is left is something we like better than what we were looking for: an honest, strangely beautiful way to look at your own relationship with money, with all the test results stapled to the back.
Try it in the app: every claim on this page links to a pre-registered experiment in our public research log, and the receipts view inside Astro-Edge shows what is validated and what is not, always one tap away.
FAQ
Does astrology actually work for investing? Not for predicting returns, as far as careful testing can tell. Our own experiments compared real birth charts to shuffled ones through the entire app and found the real charts performed like random ones. What survived testing was a consistent effect on portfolio risk shape, plus the ordinary benefits of investing on a schedule.
What is a placebo test in backtesting? It is a control where you rerun a strategy with deliberately meaningless inputs, like a random stranger's birth chart, to see whether the meaningful input was doing anything. If the fake inputs perform as well as the real one, the strategy's results came from something else, like the era or the asset mix.
Why would an astrology app publish negative results? Because the people using it deserve to know what the tool can and cannot do. We enjoy the traditions, we built them carefully, and we tested them honestly, and the testing is part of the product rather than a secret.