> ## Documentation Index
> Fetch the complete documentation index at: https://docs.noonum.ai/llms.txt
> Use this file to discover all available pages before exploring further.

# Portfolio optimization

> How Noonum turns a strategy's ranked holdings into a weighted portfolio, using score-tilt weighting to tilt weight toward higher-scoring companies.

Portfolio optimization turns a strategy's ranked holdings into an investable, weighted portfolio. Noonum does this with score-tilt weighting: holdings that score higher on the signal you choose receive more weight, and lower-scoring holdings receive less.

The score you tilt by is up to you. A common choice is the [conviction score](/base/concepts/signals-and-scores), so companies with the strongest thematic fit carry the most weight.

<Note>
  "Portfolio" here means the weighted result of `POST /strategies/{strategyId}/optimize` — a computed output, not the saved `/portfolios` resource. See [Strategies, holdings, and portfolios](/base/concepts/overview#strategies-holdings-and-portfolios) for how they differ.
</Note>

## Why this style of weighting is valued

<CardGroup cols={3}>
  <Card title="Transparent" icon="eye">
    Each holding's weight follows directly from its score.
  </Card>

  <Card title="Replicable" icon="repeat">
    The same scores produce the same weights every time, so results are easy to reproduce and audit.
  </Card>

  <Card title="Lower turnover" icon="arrow-trend-down">
    Weights move with the scores instead of swinging on small input changes, so portfolios stay more stable between rebalances.
  </Card>
</CardGroup>

<Note>
  Score-based weighting is the approach the ETF and index industry commonly uses to build factor, thematic, and ESG portfolios.
</Note>

## Construct the portfolio

Portfolio optimization runs against the `POST /strategies/{strategyId}/optimize` endpoint, which tilts holding weights by your chosen signal score. See [REST integration](/v2/guides/rest-integration) for a worked request and response.