Graph-Based Approach
A lot of pricing problems look simple until you need broad coverage. You might have thousands of assets, many quote currencies, and uneven liquidity across venues.
A graph-based approach organizes this into a network. Each asset can be treated as a node, and each trusted market rate becomes a link between nodes.
Once you have that network, you can derive cross rates by finding paths through it. The value of the approach is that it scales: you don’t need a direct market for every pair to produce a usable conversion.
Why It Matters
Graph-based methods are a practical way to create wide exchange-rate coverage without relying on every possible direct trading pair.
What makes a graph approach “safe” for pricing?
The graph itself is just a structure; safety comes from the rules around which edges are allowed. Many systems require spot-only markets, freshness limits, and outlier filtering before an edge is considered usable. Without those constraints, a graph can connect everything but still produce unreliable numbers.
How do you choose between multiple paths in a graph?
Often there are many ways to get from one asset to another. Systems can prefer shorter paths, prefer higher-quality venues, or prefer assets like major stablecoins and fiat pairs. The choice policy is important because different paths can produce slightly different results.
Real-World Example
To get SOL/GBP, a system may not rely on a direct SOL/GBP market. Instead, it can use SOL/USDT and USDT/GBP edges in a graph to compute the cross rate.
Graph-Based Approach and Exchange Rates API
Cross-asset exchange rates require a structured way to connect markets. CoinAPI’s Exchange Rates API uses graph concepts to link assets and publish rates across a broad universe.