Stake Bleeding

The basis for stake bleeding is the long-range attack of PoS systems. This attack works on PoS chains that implement a sort of “longest chain” rule. Nodes select the canonical chain by comparing the stake that was deposited within a certain fork.

Let’s assume an attacker that owns a minority of the total stake. He can secretly create a fork and publish it after he withdrew his deposit from the original chain. The problem is that, since he is the only block producer in the attack chain, he earns all the transaction fees. This, in turn, increases his stake in this fork It is conceivable that his stake is higher than the stake in the original chain after a while. Other block producers would switch to the attacker chain.

Checkpoints can mitigate this problem. But if the checkpoint frequency is too low, enough transactions are left to earn fees from.

You can find further information about stake bleeding here: https://ieeexplore.ieee.org/document/8525396