In a new paper with Benny Sudakov from @ETH_en, we compare Restaking to Proof-of-Stake (PoS) protocol. First, we look into the sufficient condition that a restaking graph is secure. Under this condition, transforming a restaking graph into secure PoS protocols is always possible.

Next, we introduce a new measure -- restaking savings: minimum extra stake for validators to be able to transform any secure restaking graph into secure PoS protocols. Transformation is done by splitting all validator stakes among secured services.

We derive lower and upper bounds on restaking savings, as a function of a restaking graph. In particular, we show that it asymptotically grows as a square root of number of validators.

In a complementary question, initially secure PoS protocols are given. Aggregating them into secure restaking graph is not always possible.

We introduce PoS savings: minimum extra stake for validators in the aggregate restaking graph so that the graph becomes secure.

PoS savings is lower bounded by a constant, and is upper bounded by the largest degree in the restaking graph among services. A wide gap between lower and upper bounds remains.

For more details, check out the paper: . Any feedback is welcome.