Backtesting Arena

Backtesting Arena

Back to blog

Market Data Licensing: How Price Data Is Licensed — and Why We're Switching Off Four Asset Classes

Market data licensing is tiered, and the decisive line runs between derived values and raw prices. What that costs — and why we're switching off four asset classes for now.

Backtesting Arena·July 14, 2026·4 min read·1 views
Market Data Licensing: How Price Data Is Licensed — and Why We're Switching Off Four Asset Classes

Market Data Licensing: How Price Data Is Licensed — and Why We're Switching Off Four Asset Classes

Stocks, ETFs, commodities and forex are being removed from this platform. What stays: crypto.

The occasion is a market data licence. How those licences work is the part that's useful even if our asset classes don't interest you.

The tiers

Market data licences are tiered, and the steps are steeper than they look.

TierWhat it permits
PersonalYou look at the data. Yourself. That's it.
Internal commercialYou work with it inside your company and show customers derived quantities — a state, a metric, a statistic. A multiple of the personal tier.
Raw values to end usersYou show prices. Charts. Trade lists with entry and exit prices. Separate contract class, own negotiation.

The decisive line runs between derived and raw

The line between the top two tiers is the most expensive one in this business — and it doesn't run where you'd expect.

It doesn't separate "private" from "commercial." It separates derived from raw.

A regime state is derived. A metric is derived. A win rate is derived.

A price chart is raw. A trade list with entry and exit prices is raw. Which puts a full backtester in tier three.

The licence doesn't come from the vendor

The obvious escape is switching providers. It doesn't work.

The same tiering appears across every major market data provider. It isn't the business policy of individual companies.

The licence comes from the exchange, and it's passed through.

If you deliver stock prices, you eventually pay the exchange — no matter how many hands the data passed through first. The vendor is the till, not the rights holder.

And there is no free path

The second obvious escape is free public databases. It runs into the same wall.

A public economic database is an aggregator, not a rights holder. The interesting price series there carry copyright restrictions — or have been removed entirely at the rights holders' request and no longer exist.

That's a useful thing to know on its own: a free source is not a licence-free source. Someone still owns the numbers, and they don't stop owning them because a public institution republishes them.

There's the licensed path and the derived path. There is no third.

The arithmetic

The correct tier for a backtester costs a multiple of what a personal plan costs.

On our platform, stocks, ETFs, commodities and forex together drew a fraction of the crypto traffic — at a fraction of the engagement. Much of it came from paid advertising, not organic demand.

The licence makes the cost real. The usage makes the answer easy.

This isn't a stance against equities. It's arithmetic with today's numbers.

What stays — and what can come back

Not a single strategy disappears.

The backtest engine, the filters, the sample gates, the regime logic, the cycle analytics — none of it is tied to an asset class. The code is asset-agnostic, and it stays that way. What disappears is four data sources and the dashboards that stood on them.

Which is why the shutdown is built to be reversible. The asset types stay in the data model.

Two routes back:

Demand carries the licence. If equity backtests attract enough users to carry the correct tier, we switch them back on. Not a principle — a threshold.

Or: tokenised equities. Equity tokens trade on public blockchains. The price sits in the ledger. A blockchain doesn't pass through an exchange licence. The history is short because the segment is young — but it grows from today, and it's free. We think this is the likelier route.

Our derived analytics aren't affected by any of this. A state with a justification is worth more than a number without one — that's how our knowledge layer for agents has been built from the start.

What we're doing instead is laid out in our overview of how the platform works.

Frequently asked questions

What does a market data licence for stock prices cost? Personal plans are cheap but don't cover commercial use. The internal commercial tier at the major vendors runs several times higher. Delivering raw prices to end users is a separate contract class with individual negotiation.

Can I display market data in my app? Only under a licence that explicitly covers it. What matters isn't whether you make money from it, but whether the data reaches an external user — including in derived form.

What counts as "derived" data? Quantities computed from prices without showing the prices themselves: a regime state, a metric, a statistic. A price chart or a trade list with prices is not derived. It's raw.

Why doesn't switching vendors solve it? Because the tiers don't originate with the data vendor. They come from the exchanges and are passed through. Every major provider has the same structure.

Is free market data licence-free? No. A public database is an aggregator, not a rights holder. The series can carry copyright restrictions, or be removed at the rights holder's request.

Will the asset classes come back? Possibly. The asset types stay in the data model and the strategies are asset-agnostic. Two routes: enough demand to carry the licence — or tokenised equities, which trade on public blockchains and pass through no exchange licence.

Study the Past — Improve your Future. 🥋

Try it yourself

Run the backtest with your own parameters and time ranges.

Run backtest →
📬

Don't miss new blog posts

One short email per new post — strategies, backtests, market analysis. No spam, unsubscribe with one click anytime.

By subscribing you accept our privacy policy. We use Resend for delivery. Double opt-in confirmation required.

Comments (0)

Join free to post comments.

Sign up →

No comments yet. Be the first!