I keep three DeFi dashboards open while managing positions. On a quiet Sunday I pulled up my uniBTC position on all three simultaneously and wrote down each yield figure before checking any of the others.
Three different numbers. Same wallet. Same position. Same moment.
I traced each one back to its source.
The first dashboard was reading the exchange rate appreciation of uniBTC against WBTC. It treated the rate differential as a yield equivalent. The second was pulling the vault's listed APY from Bedrock's own API. The third was counting points accumulation from an active incentive program running alongside the vault.
All three were reading something real. None of them were technically wrong. But they were reading three completely different layers of Bedrock's architecture and presenting each number as if it was the answer to the single question "how much is this position earning."
The non-rebasing design creates exactly this problem. Because yield accrues in the exchange rate rather than wallet balance, every third-party tool that reads a DeFi position has to decide what it considers yield. The exchange rate differential, the protocol's own APY figure, and external incentive accumulation are three different answers to the same surface question. Different tools made different choices, and none of them labeled which choice they made.
This is a composability gap more than a documentation gap. Bedrock's yield architecture is more sophisticated than what most portfolio dashboards were built to read. The ones that handle it correctly are the ones that explicitly account for non-rebasing behavior. Most don't, and they will silently show you a different number depending on which layer they happened to read.
Know which layer you're looking at before you make a decision based on the number.