The market dropped, so which 5 sectors are quietly being bought? A 22-year-old Taiwanese student built a paid Taiwan-stock tool from one bubble chart

Tide only does one thing: after every close, it totals the three major institutional investor buy and sell figures across 108 self-defined thematic sectors and draws them as a bubble chart of "where the money is flowing." It has never appeared on the App Store free chart and has only 250 ratings, yet it sits at No. 33 in the Taiwan finance paid chart. Its users didn't come from the store; they came from a screenshot on Threads. The chart itself is the content, and that's the most worth learning. Copying it as-is in Taiwan has little upside; building it for another market is where it gets interesting.

Jun 5, one Threads post

2026 June 5 was a Friday, and the Taiwan stock index closed down 1.33%. Ryan Chen (陈宏祁), a 22-year-old student who had just returned to Taiwan from the US, posted on his Threads account @ohoh_omg:

The market is down -1.33% on Friday, but 5 sectors have big investors quietly buying. In a selloff like this, the sectors being quietly bought by big players are usually the best watchlist.

Below is a screenshot of the web page he built, tide-tw.app. The post got 3.3K likes, 322 replies and 2.6K shares. Three days later, on Monday (June 8), the Taiwan index at one point fell 2,600 points intraday, the largest intraday point drop in history. That same day the news site abmedia ran a long piece titled "Netizen-built Taiwan stock sector rotation map," saying the tool "spread quickly on Threads," and walking through the bubble chart, the "value ranking" and the "bottom-fishing detector" with screenshots.

The reason he gave for building it is on the "About" page of his site: when he looked at US stocks in America, opening a sector ETF showed where money was moving between industries; back in Taiwan he couldn't find the same thing. Taiwan's institutional data is published daily, but "it's a wall of numbers," and the existing tools that organize it "have interfaces and jargon written for veterans." TVBS reported that he wrote the program with AI.

What it actually is

Opening it gives you a bubble chart: the horizontal axis is how much institutional money flowed net in over the past 5 days, the vertical axis is whether inflows are accelerating or slowing, and bubble size is the scale of money over the past 20 days. The four quadrants are called Rising Tide (earlier called "Main Force"), Rotation, Watch, and Ebbing. Click a sector and you see which stocks in it are being bought.

Two details set it apart from other chip-flow tools:

• The sectors are self-defined. The exchange's official industry categories are only about two dozen, and "semiconductors" lumps together a huge chunk of the market. Tide splits them into 108 thematic groups; for example, memory is split into modules, niche types and high-bandwidth. Each sector is tagged with "how many names are being bought" and "which name is bought the most," so one heavyweight stock can't carry an entire sector.
• Plain-language explanations sit next to the jargon. The longest review (9-05) says: "It presents three major institutional fund flows with bubble charts and tides, the presentation is fresh and intuitive… it doesn't cram in news or ads."

It only covers after-hours data: no candlestick charts, no earnings, no live quotes. Data is updated before 18:30 on every trading day.

Where users come from: screenshots, media, then into the app

The timeline is clear:

The main channel is Threads. Taiwan accounts for over 20% of Threads' global traffic (Ketagalan Media, 2026-03), and that is where Taiwan's retail traders already talk stocks. Chen Hongqi posts a daily after-hours breakdown using his own screenshots, and the product is the post. Same-theme tools like FlowPulse, "What the three big institutions are buying," and Gemr Taiwan Stock Radar: Institutional Moves & Capital Rotation also launched in 2026, each with only 2–5 ratings. They use the same public data. What they lack is a chart people can repost and a person who talks the market on it every day.

Most App users look like they came over from the web version (inferred: the web version had 10k sign-ups first, the App was swamped on day one, and the App supports logging in with web subscriptions). No evidence found that they run paid ads.

Where the money comes from

The free tier is not a trial. Bubble charts, rankings and watchlist alerts are all open, with no ads. Pro sells the next layer down: institutional breakdowns, financing data with a 20-day average price, per-stock history lookups, and unlimited watch slots. One thing that time builds: the official source only publishes same-day institutional breakdowns, not history. Tide saves a copy every day, and it gets more valuable the longer it runs.

The web price is 20% cheaper than in-app, and one subscription works across all platforms. It is steering payments to the web, so the App Store only sees part of the revenue. Total revenue is not public.

A few visible numbers (App Store Taiwan storefront, 11-06): 250 ratings, 4.8 stars, roughly 1 new installs a day over the past 10 days. Not on the Finance Top Free chart's top 200, but #33 on Finance Top Grossing. Small user base, high paying rate. Grossing rank follows the market: it fell to #137 from #11-03, then climbed back to #34 from #11-05, matching the week Taiex dropped 10,000 points in a month (Business Today 11-06). Like June, when the market crashes, everyone wants to know where the money went. The Android version has about 2,300 installs (Google Play, 10-29), with subscriptions handled on the web.

Who it's competing with

In Taiwan, this slot already has two tools gaining traction (Tide and Real-time Capital Radar), and several later entrants didn't take off. Launching another Taiwan stock version means head-to-head competition with a product already covered by the media and a creator with 8,500 followers on Threads.

The parts you can copy

1. The product output is a single post. "The market is down, but these 5 sectors are being quietly bought" is one headline, one chart and one question everyone wants to click. Before building the tool, ask: if a user screenshots this and shares it, will others understand it?
2. The founder talks about the day's events with their own product every day. Not product announcements, but "where the money went today."
3. Web first, App second. Validate on the web, build users, get media coverage; launch the App to bring in users in one go. Subscriptions work across platforms, and the web price is lower.
4. Give generously for free, sell depth and history in Pro. The free version has to be good enough that people share the screenshots.
5. Make "no investment advice" a selling point. No courses, no groups, no sponsored posts, subscriptions only. Where "teachers selling picks" are everywhere, this alone earns people's money.

If it were you

In short: try a different market. The approach, "aggregate public institutional trading data by sub-theme, turn it into a chart retail traders can read, and post it yourself every day," is clear. Taiwan already has two players, and I haven't seen clear equivalents in other markets (inferred, not checked market by market). But it depends on two things followers may not have: real knowledge of local stock themes, and a founder willing to talk the market every day.

Who it's for: indie devs who know a particular stock market well and are already active in its retail communities. 1 person is enough. South Korea (the exchange publishes daily per-stock foreign and institutional net buying) is the closest to Taiwan; Hong Kong (Northbound/Southbound flows) and mainland China (Northbound and Dragon-Tiger list data) also have public data, but there are more tools and tighter regulation. Japan's investor-type data comes out weekly by market, not per stock, so it's a poor fit.

Not for: people who don't trade stocks and can't tell sectors apart. Whether a chart spreads depends on how you split the 108 sectors, and AI can't make that call for you. Also not for anyone unwilling to show their face or post every day, or anyone wanting to build a second one in Taiwan.

How to do it: launch on the web first, after-hours only, one chart: cumulative capital flow on the x-axis, acceleration on the y-axis, with four quadrants that have memorable names. Group sectors yourself and label each one with "how many stocks are buying and which one is buying the most." Free tier with no paywall and no ads. Once you have a few thousand sign-ups, add the App and subscriptions. Pro gets the breakdown data, history lookback and more alerts. Save the same-day official figures every day, since they never get published again.

How to get users: go where local retail traders talk stocks (Threads in Taiwan, online forums and YouTube in South Korea, Xueqiu and Xiaohongshu in mainland China). After each trading day closes, post a chart with one line: "The market is down, but these sectors are being bought." A crash day is the best day to post: everyone is panicking and everyone wants to know where the money went. Pitch local financial media on the story of "young people built a market tool with AI."

When to get in: there's no hard deadline, but ideally get the web version running before the next big sell-off. Tools like this spike on crash days, and they struggle to gain traction in flat markets.

Signs you got it wrong: ① during a flat stretch of several months in the Taiwan market, Tide's Finance Top Grossing rank falls out of the top 100 and doesn't come back. That means it's a business that rises and falls with the market, and that's true in any market. ② A brokerage or a major trading app (Sanzhu, CMoney, Toss in South Korea) builds a "sector capital-flow bubble chart" into its free tier. That takes the room out from under single-purpose tools. ③ The reverse: someone in your chosen market is already posting the same chart daily in local communities and getting traction. That spot is already taken.

---

Data and sources: App Store Taiwan storefront listing, version history, ratings and charts (2026-10-28 to 11-06); Google Play (10-29); tide-tw.app homepage, About page and pricing page; Threads posts by @ohoh_omg and @moneygame17588 (9-06); ABMedia (2026-06-08); TVBS (2026-08-26), Business Today (2026-08-27), LINE Today reprints; Ketagalan Media (2026-03-11). Web revenue and paying-user counts are not public. The date of June's viral post is inferred as 6-05 from "Friday drop of 1.33%" and ABMedia's article 6-08.