How PMarket Arena Clears a Market

The browser arena is an interactive approximation with a very specific job: keep the important forces visible while running continuously in a web page.

It is not a full exchange simulator. That boundary is important. The arena makes mechanisms visible. The paper evidence comes from the reproducible research pipeline.

One Tick at a Time

Each tick follows the same conceptual loop:

1. the hidden truth path moves,
2. agents update beliefs or pressure,
3. shocks and interventions modify the market state,
4. liquidity and price impact transform pressure into price,
5. the UI records price, volume, spread, stress, and interpretation.

The important design choice is that every tick has the same readable shape. The arena is not trying to hide market dynamics behind one final chart. It is exposing the loop that creates the chart.

The Clearing Equation

The large-scale research engine uses an analytical batch-clearing idea:

$$p^* = \frac{\sum_i w_i \hat{p}_i + D_{\mathrm{panic}}} {\sum_i w_i}$$

The browser arena extends the same intuition with demo-specific terms for shocks, liquidity stress, trend pressure, and intervention effects:

$$p_t = \mathrm{clip}_{[0,1]} \left( p^{\mathrm{belief}}_t + I^{\mathrm{trend}}_t + I^{\mathrm{panic}}_t + I^{\mathrm{shock}}_t + I^{\mathrm{arb}}_t + \eta_t \right)$$

The point is not that this is a final market design. The point is that every force has a name and a visual trace.

Why Not Simulate a Full Order Book in the Browser?

A full order book is valuable for microstructure research, but it is not always the right tool for public explanation. A browser demo needs to run indefinitely, react to user interventions, stay responsive, and surface the main mechanism without burying the reader in queue mechanics.

The arena chooses a cohort-level model because it is explainable and interactive. The research repo still keeps the more formal engines and reproducible figures.

What the UI Should Teach

The market chart should teach price and truth tracking. The volume bars should teach participation and stress. The agent distribution should teach composition. The intervention log should teach causality. The commentary panel should teach interpretation.

The UI is therefore part of the theory. It decides which forces become visible.

With the engine exposed, failure becomes easier to discuss. A market can be alive, active, and visually busy while its price becomes a worse signal.