Price Improvement
Price improvement measures how much better an RFQ quote is compared to the best available alternative venue. It is the core value proposition of the RFQ model — by soliciting competing quotes from professional market makers, takers can achieve better execution than passive AMMs or order books, especially on larger trades.
Formula
Price improvement is calculated by comparing the RFQ output against the best venue output:
improvement = (rfqAmountOut - bestVenueAmountOut) / bestVenueAmountOutThe result is expressed as a percentage or in basis points (bps), where 1 bps = 0.01%.
improvementBps = improvement * 10_000Example: If the best venue offers 9,950 USDC output and the RFQ quote offers 9,965 USDC:
improvement = (9965 - 9950) / 9950 = 0.001507...
improvementBps = 15.07 bpsThis means the RFQ delivered 15.07 basis points of price improvement over the best alternative.
Positive improvement means the RFQ is better. Negative improvement means a venue would have given you more output — in this case, the UI will still display the venue results so you can make an informed decision.
Where Improvement Is Displayed
Quote Cards
Each quote card includes a vs AMM Baseline comparison row showing the percentage difference between the quote’s output and the best AMM estimate. A green +X.XX% with an upward trend icon indicates the quote beats the AMM; a grey value indicates it does not.
Savings Banner
In the Venue Comparison panel, when the best RFQ quote exceeds all venue outputs, a green callout displays the absolute token savings and percentage improvement:
+15.23 USDC saved with RFQ vs HyperCore Spot (0.15% better execution)
The denominator for this percentage is the best venue output, making the number directly interpretable as “how much extra you got by using the RFQ.”
Execution Success Screen
After a successful fill, the execution panel shows a price improvement callout if the RFQ beat the best venue:
+12.5 bps better than best venue
This appears in a green badge within the success overlay, giving you confirmation that the RFQ delivered real value.
Mid-Price Benchmark
In addition to venue-vs-venue comparison, every quote is benchmarked against the universal mid-price derived from the HyperCore L2 orderbook. This provides a venue-neutral reference:
referenceOut = amountIn * (midPriceIn / midPriceOut)The vs Mid-Price row on each quote card shows how much slippage you are taking relative to this theoretical zero-impact price. Lower numbers are better.
This mid-price benchmark ensures that slippage numbers across all venues (HyperCore, DEX, RFQ) are computed against the same reference and are directly comparable.
How Improvement Feeds Into Points
The points engine uses price improvement to compute a continuous improvement multiplier that scales the points earned on each fill:
multiplier = 1 + clamp(improvementBps, -20, +50) / 100This produces a range of 0.80x to 1.50x (clamped to 0.80x - 1.60x after final bounds check):
| Improvement (bps) | Multiplier |
|---|---|
| -20 or worse | 0.80x |
| 0 (equal to venue) | 1.00x |
| +10 | 1.10x |
| +25 | 1.25x |
| +50 or better | 1.50x |
Fills where the benchmark is unavailable (e.g. the venue comparison could not reach any venue) receive a penalty multiplier of 0.90x instead of 1.00x, to discourage gaming by intentionally trading in conditions where benchmarks cannot be computed.
The improvement multiplier stacks with other multipliers (privacy, repeat decay, NFT boost) to produce the final points earned per fill. See the Points section for the complete formula.
Interpretation Guidelines
| Improvement Range | Interpretation |
|---|---|
| > +20 bps | Excellent — significant savings over venues |
| +5 to +20 bps | Good — RFQ is clearly winning |
| 0 to +5 bps | Marginal — RFQ is competitive but close |
| -5 to 0 bps | Neutral — venues are slightly better at this size |
| < -5 bps | Consider using a venue directly for this trade |
For large trades ($100k+), even small basis point improvements translate to meaningful absolute savings. A 10 bps improvement on a $500k trade saves $50.
Partial Fill Considerations
When a venue can only partially fill the requested size, the comparison adjusts:
- The venue’s output is scaled to the filled portion only.
- The RFQ’s full output is compared against the venue’s partial output, which would naturally favour the RFQ.
- The UI displays a “Partial (XX.X%)” badge on the venue result so you can see the fill rate.
This ensures that partial fill results do not artificially inflate the improvement number. The savings callout still reflects the real-world advantage of being able to get a full fill via the RFQ.