Proof of Work and Order-Lifetime Rules for Exchange Abuse
Summary
The document considers whether exchanges could use proof-of-work challenges to discourage rapid order submission and cancellation, potentially limiting information probing, manipulation, or advantages from proximity. It explains that proof of work is more common where a service does not trust its clients, while exchanges typically rely on authenticated, contractual relationships with participants. Adding computation to order handling also conflicts with the low-latency design exchanges seek.
As an alternative, the answer proposes a minimum or fixed order lifetime, so orders remain available for a set period instead of being withdrawn immediately. This could reduce rapid order-book probing and misleading signals, though the document provides no empirical test of the proposal or analysis of its effects on liquidity and risk. For a hypothetical public, trustless exchange, it suggests escrow of margin as a more relevant mechanism than interface-level proof of work.
Key ideas
- Proof of work can impose a computational cost on actions, but that cost conflicts with exchange demands for low latency.
- Exchanges generally manage client access through authentication and contractual relationships.
- A fixed minimum order lifetime could reduce rapid cancellations used to probe order-book information.
- The proposed order-lifetime rule is not supported by empirical evidence in the document.
- A trustless exchange may need margin escrow to support transactions.
Tags
Full text
# Are proof-of-work systems used by exchanges? # Are proof-of-work systems used by exchanges? Are there any proof-of-work systems used by exchanges? If not, are there any interesting proposals worth reading? In principle, there might be applications for proof-of-work systems in preventing algorithms from flooding exchanges either to manipulate prices or gain information, or perhaps to limit the advantages of physical proximity. An exchange might require a proof-of-work computation when withdrawing an order, for example, which makes placing and withdrawing a large number of orders unsafe. ## Answer by Phil H (score 2, accepted) https://quant.stackexchange.com/a/10098 Proof of work systems are generally used where you do not trust the client; the Bitcoin one is used to slow down the generation of new coins and is adaptive; if hardware speeds up, the work gets harder. By contrast, an exchange has a contractual agreement with the client, and can require it to authenticate, encrypt etc. The central problem, though, is that exchanges want to be low-latency, which is the opposite of a proof-of-work system. A more convincing solution to the current problems of HFT is to have a minimum (or fixed) order lifetime; any order is valid for x seconds. Much of the HFT activity involves submitting and cancelling orders very quickly just to test the water and discover information about the order book, or to fool less sophisticated algorithms. By fixing the order lifetime, the orders can actually be filled. On the other hand, if there was a kind of Bitcoin-style exchange with a public interface, there might be cause for trust-less transactions, but I suspect that would involve escrow of margin, rather than proof of work on the interface.
Shown in full with attribution under the source's licence. Licence: CC BY-SA 4.0 (Stack Exchange)
This summary was written by Stratmill's research agent from the original; it is not a copy of the source.