Dashboards and Market Scanners: What They Are For

The short answer

A scanner takes a list of instruments, computes something from each one's price data, and hands back the ones that currently satisfy a condition you specified. That is the whole function.

It finds candidates. It does not judge them, and it cannot act on them. The confusion that costs people money is treating a shortlist as a set of conclusions, when all the scan has established is that these instruments satisfy a condition — not that the condition is worth acting on.

The line the platform actually draws

This is not a stylistic preference of ours. It is built into how the software is structured, and MetaQuotes documents it in unusually blunt terms:

"Custom Indicators are custom developed technical indicators intended for analyzing price dynamics. Custom indicators are only used for analyzing symbol price dynamics. Indicators cannot trade."

Three words doing a lot of work: indicators cannot trade. The component that computes a value from prices and the component that places an order are separate things, and the first is not permitted to do the second's job.

A scan is that first component applied across many instruments at once. However the results are presented — a grid of tiles, a heat map, a ranked table — the thing underneath is arithmetic on price data, and it has no mechanism for deciding anything.

The instrument list it runs over is similarly plain:

"The Market Watch window provides an overview of price data of financial instruments: quotes, price statistics and tick chart."

"The Market Watch features real-time quotes of financial instruments and other price data: spread, volume, etc."

Quotes, statistics, spread, volume. No opinions in there.

Why the find-versus-judge distinction matters

A scan inherits whatever you asked for. If the condition has no edge, the shortlist has no edge — it is just a shorter list of instruments with an arbitrary property in common. The scan cannot warn you about this, because it has no notion of whether the condition is any good.

More filters feel like more rigour and are not. Each additional condition narrows the list and increases the feeling of having found something special. It is the same machinery as trying many strategy variations until one looks good, which is the subject of why good backtests fail live. A shortlist of four from a universe of four hundred is not evidence; it is a selection you made.

Presentation implies confidence that is not there. A dashboard ranking results top to bottom invites you to read the top one as best. It is first on whatever you sorted by, which is a different statement.

A note on vocabulary

Worth being straight about: we looked for MetaQuotes documentation describing a feature called a "scanner" or a "dashboard", and there isn't one. Those are vendor words for products built on top of the documented pieces — indicators, the symbol list, custom programs.

That is not a criticism of such products. It does mean that when a vendor describes what their scanner does, you are reading a marketing claim and not a platform specification, and the underlying behaviour is still only ever "compute something from price data and filter on it."

Using one well

Know the condition exactly. If you cannot state in one sentence what the scan tests, you cannot interpret its output. A scan whose rules are hidden produces a list you have no basis for trusting.

Treat the output as the start of work. The shortlist tells you where to look, which is genuinely valuable — it is a time-saving tool, and time is the scarce resource. The analysis still has to happen afterwards, by you.

Keep the count in view. How many instruments were scanned, and how many passed? Four out of four hundred means something different from four out of six.

Do not confuse it with prediction. An indicator is computed from prices that already exist. A model that forecasts is a different machine with different failure modes, and indicators, models and language tools separates the three. If a product does both, judge the halves separately.

Nothing here is a recommendation to use any particular tool; see the risk disclosure.