How do Hyperliquid HIP-4 markets differ from perps?

Perp endpoints use the perp `meta` response and integer asset indices. By contrast, Hyperliquid's official outcome docs say outcome assets are built from an `outcome` id plus a binary `side`, with representations like `#<encoding>`, `+<encoding>`, and `100_000_000 + encoding`.

That is a useful signal for researchers: HIP-4 markets are not just one more perp namespace. They are a separate instrument family with their own identity scheme.

Hyperliquid's perp and hyperp docs revolve around continuous exposure, mark prices, and funding logic. The official `outcomeMeta` example instead centers on fields like `expiry`, `targetPrice`, and `Yes` or `No` side specs.

So the research object changes. One side of your work is no longer a continuous contract you can think about in the same way as a perp; it is a defined outcome question with dated terms.

The biggest mistake is to port perp intuition straight across. Outcome markets can still trade on the same venue, but the features you care about and the failure modes you watch are different.

This is exactly the sort of surface where explicit metadata, forward testing, and run provenance matter. A clean chart is much less useful if you lost the actual market question the chart came from.