There are some expensive hardware solutions to capture all network traffic and timestamp it at nanosecond level. E.g.,
https://www.endace.com/endace-high-speed-network-recorders.h...
There are decentralized exchanges that are becoming more popular (Bitsquare for one, check out http://www.bitsquare.io), but there's a chicken-and-egg problem for liquidity. Not a lot of people currently trade on the decentralized exchanges, so liquidity is thin (and until there's liquidity, larger traders will likely not trade there). However, there isn't any counterparty risk, other than if you buy BTC for USD, and the person on the other end charges back the bank transfer. When you trade crypto-to-crypto, the risk is minimal (barring a 51% attack reversing the history on the cryptocurrency you purchased).
(Sell BTC for real money, surely?)
So there isn't any counterparty risk apart from when it's most critical, actually cashing out?
That should show up on your 10GHz scope, although the probe setup is a little fiddly.
Or you could just ask your 10G network card to timestamp packets for you. Only valid for packets sent back to the same card, unless you go to heroic lengths to synchronise between systems.
(The exchange issue is called "counterparty risk", and it's the one area Bitcoin advocates tend to handwave away with reputation systems)
Maybe if you're a day or mechanical trader. Although, I personally don't know if that's actually true. Have you tried it? Can you share any data or citations?
> Exchanges exist for trading quickly
Agreed
> they must be centralized due to this.
Have you encountered any definite proof of your conclusion?
Or swing trader or any kind of trader looking to time the market. Trading requires the ability to execute in near real-time. Block chain transactions and worse, fiat transfers between parties, takes far too long for trading to be effective.
> Have you encountered any definite proof of your conclusion?
They must be centralized to obtain the necessary speed because currently block chain transactions are too slow. Proof is unnecessary, logic is more than ample. One can't trade effectively if ones capital is hung up in "transit" between parties. Exchanges exist specifically to provide fast trading. There wouldn't be any exchanges if block chain transactions were fast enough.
We don't know that blockchain transactions are inherently slow and we haven't verified that no decentralized architectural/algorithmic/structural solutions exist to coordinating trades. In fact, I can think of one or two. So I'm asking you for data, not hypothesis. Do you have any info from your research about trade order execution times on p2p exchanges over time or userbase size? Or is it just your conclusion from anecdote?
On a serious note....
> Engineers at Sydney-based Metamako LP and Exablaze Pty. Ltd., and Chicago-based xCelor LLC are rolling out switches that take around four nanoseconds—four billionths of a second—for messages to transit from one side to the other, sending data from exchanges to electronic traders.
Wow, can anyone shed some light on how they time this? What sort of clock, I'm assuming its some fancy hardware that you use to benchmark at the 4 nano second level? Or do you just claim this and assume that no one can prove you wrong?
As to Bitfinex, on the weekend a few friends and I were batting around ideas on how to reliably trade bitcoins without having to worry about being robbed by the exchange^H^H^H^H^H err worrying about the exchange being hacked and passing their mistake onto you.
The top ideas we had were:
- only trade on US domiciled exchanges, the US government tends to poke their nose into everything and this is one case where you benefit.
- phone the exchange before you start trading. If you can't get someone on the phone, that's a very bad sign:)
- Similarly, if the C level executives of the exchange aren't well known and always around at the regular bitcoin industry conferences then that's a bad, though this fails the Mt Gox test.
- we tried to figure out if you could pull the coins out each night and put them back in each day, but even then, you'd run the risk of the hack happening while you had coins held by the exchange.
But we came to the conclusion that there just isn't any sane way to trade bitcoins currently where you can eliminate "exchange risk". If anyone does know of a way please let me know.