See what price alone can’t show.

A latent state is an underlying condition we cannot observe directly. We infer which states are plausible from the evidence we can observe.

Updated belief = likelihood × prior belief, then normalized.

• The two ground coordinates, z₁ and z₂, describe an abstract latent space. They are not named market indicators.
• The surface height shows conditional probability density after one synthetic observation. Probability is the volume under the surface over a region; height alone is not a probability or a return.
• The two peaks show that different explanations can remain plausible after observing the same evidence. They do not define PraedInvest’s market regimes.
• The curved strands trace the surface. The stronger accent strands help reveal its shape; their color adds no separate signal.
• The contours beneath the surface connect locations with equal density. They are projected onto a separate plane for legibility, not confidence bands.

This is a static, synthetic Bayesian example: a two-component Gaussian prior updated with one noisy observation. The drawing shows a bounded view of an unbounded latent space. It is not a PraedInvest model estimate, proprietary methodology, forecast or trading signal.