Stock Board

About this tool

Most stock trackers want an account before they will show you a price. The ones that do not are usually heavy: a full terminal, dozens of columns, a news feed, an app to install. This one is deliberately small. You type a code, it shows the last price, the change and a candlestick chart. The list is yours and it is saved in your own browser.

That last part decides the whole design. There is no server holding your watchlist, because there is no server at all — this is a static page. Your list lives in this browser’s local storage, under a key that belongs to this tool. Clearing site data removes it, and it does not follow you to another device. In exchange, nobody can see which stocks you follow, and there is no sign-up to abandon.

Quotes and candles come from Eastmoney’s public market endpoints, the same ones their own quote pages call. The request is made by your browser directly to them, not routed through us — which is why a page with no backend can show live prices without cost. The endpoints support cross-origin requests, so this works without any proxy in the middle.

It covers Shanghai and Shenzhen listings by their six-digit code, Hong Kong by its five-digit code, and US tickers by their symbol. If you are not sure which market a name belongs to, the code you already know is usually enough: the tool decides from the shape of what you type. Red-up/green-down and green-up/red-down are both available, because which one looks correct depends on where you learned to read a quote screen.

A word on the data itself. These are public quotes from a third party, and there is no guarantee of accuracy, completeness or timeliness. Delayed prices are normal outside a paid feed. Treat this as a glance at the market, not as the basis for a trade. Nothing here is investment advice.

Market Board puts the eight things people actually check on a trading day onto one screen: index levels with the advance/decline split inside each one, the count of A-shares rising, flat and falling across the whole market, the most heavily traded stocks, the limit-up ladder, where main-force money went, every industry and concept board ranked by that flow, and the ETFs that can be traded the same day. It is a single page, it needs no account, and it stores nothing of its own.

Counting how many A-shares went up sounds like arithmetic and is really a definition. The advance/decline figures attached to an index describe that index’s own members, and the Shenzhen Component Index holds only 500 of them; adding the Shanghai and Shenzhen tallies therefore stops short of the whole market. This page adds up the 31 regional boards instead — between them they cover every listed A-share exactly once — which is what produces the figure of roughly 5,600. The page states which definition it uses, because the alternatives are all defensible and never agree with each other.

Two tabs cover things a general market page rarely shows. The limit-up ladder groups every stock that closed at its daily price limit by how many consecutive days it has done so, strongest rung first, and orders each rung by how early in the session the seal went up. The consecutive-board count exists only in Eastmoney’s limit-up pool, which is why this tab cannot be built from an ordinary quote list. The T+0 tab lists the ETFs you can buy and sell within the same session — money-market, cross-border and commodity funds — and deliberately leaves out the equity ETFs, which on the mainland can only be sold the following day.

Everything here comes from Eastmoney, a third party, and is reshaped in your browser. Quotes are delayed, and outside trading hours they are simply the last close, so a board opened at the weekend shows Friday. None of it is a recommendation: the rankings describe what happened today, not what to buy, and a stock sitting at the top of the turnover list is no more attractive for being there. Where a number depends on a choice of definition — the advance/decline count above all — the page states the choice instead of leaving you to guess.

How the intraday average line is rebuilt when the feed omits it

The average price is turnover divided by volume, but the divisor differs by instrument — shares for one market, lots for another. Rather than keep a table of scales, the tool tries both and keeps whichever result falls inside the day’s own low and high.

A cumulative volume-weighted average price has a property that makes the check possible: it must lie between the lowest and the highest price of the session. That is not an approximation, it follows from what an average is.

So the tool computes turnover divided by volume, and turnover divided by a hundred times the volume, and accepts whichever lands inside the day’s range. On a mainland stock the first result is a hundred times too high, so the second wins; on a Hong Kong listing the first is already right.

When neither falls inside the range it draws no average line at all. An index is the usual case: its volume and its turnover are not measured on the same basis, so no divisor works, and a line built from the wrong one would look plausible and be wrong.

Why the two outer buckets of the distribution share one label

The upstream feed labels the deepest fall and the largest rise both as “>10%”, so the tool reads the eleven buckets by their position rather than by their text. Read by label, the two ends would merge into a single bar with nothing looking wrong.

Eleven buckets run from the deepest decline through flat to the largest gain, and the first and last of them carry the same label. That is not a quirk of one feed: the label is a bound, and the two bounds point in opposite directions.

Position is what distinguishes them. The tool marks each bucket as falling, flat or rising according to where it sits — the middle one is flat, everything before it is a decline, everything after it a gain — and the chart is drawn from that, never from the text.

Because the buckets and the headline counts arrive as two independent groups of numbers, the tool also checks that they agree: the five decline buckets must add up to the declining count, the five gain buckets to the rising count, and the middle bucket to the flat count. If the feed ever inserts or drops a bucket, that check fails before a silently shifted chart is published.

Why a figure sometimes appears as a dash instead of a number

Because the feed did not send a number. Only genuine numeric values are accepted; anything else, a digit string included, becomes a dash. A change in the upstream format therefore shows up on the page instead of being papered over.

Data feeds signal a missing value in their own way, and the dash on screen is that same signal rendered. The rule is one line long: a value is used only when it is a real number.

Numbers arriving as text are rejected too, which looks pedantic until you consider why. Every quote endpoint here is requested with a flag that makes it return true numbers, so a string where a number belongs means the shape of the response has changed. Coercing it would hide that; showing a dash exposes it.

The trade-off is deliberate. Being generous with input would let a changed feed keep working by accident, and the figures would look right until one day they did not. A visible gap is easier to trust than a value nobody can account for.

Two colour rules live on the same page, on purpose

The ranking side is fixed to red-up, green-down because that is the de facto standard in Chinese markets. The watchlist side can be switched to green-up, red-down for people who also follow US stocks. The rules differ deliberately, not to suit taste.

In Chinese markets red-up green-down is a convention rather than a preference: market terminals, broker apps and financial media all follow it. The ranking is read by people looking at the Chinese market, so it follows the convention and needs no translation in the reader head.

A watchlist is a personal tool and may hold both markets at once. When comparing across them, mixing two colour systems in one table makes a green US stock read as a fall, so a single switch flips the whole set instead of deciding market by market, row by row.

There is one more protection beyond colour: the structural stylesheet hard-codes no colour values and references theme variables only. Directional colour is the single exception and it carries a fallback default, so if the sync script ever fails to apply its class the page still renders red-up rather than in no colour at all.