What say you?
Thanks for taking the time to read this
The analysis was good on the Ponzi scheme up to a point, but the thinking was too limited. The assumptions in that post are based on the idea that you musst be exchanging BTC for fiat in order to be doing the arbitrage. Certainly that is possible, and since BTC arbs has account slots for BTC and $, a good assumption that this is being done. It isn't necessary or even likely in my mind that that is all thats being done.
As far as BTC itself is concerned, all thats necessary is a means of transport between 1 exchange and another. To most of us, in most markets that means would be the $, but in a world where nearly a hundred different coins exist on different exchanges, its certainly not the only possibility.
Ive often used the features of 1 currency, usually faster transaction times, to transport BTC from one exchange to another. I have at times used that transaction to make money. If that were automated through scripts to work at the push of a button instead of manually calculating and entering the actions as I have done, it would be even more effective.
At this moment there is a spread of $616.94 $618.999 and $624.42 between the main exchanges on BTCwisdom. In this 1 instant today we're looking at an almost 1% difference. Given 50% for overhead as an easy guess for general business practices were talking about a need for BTCarbs to produce 3-4% per day in positive margins to be able to give out that .5-2% gain we see in the accounts. With transport fees figured in they could likely make this happen in perhaps 12 transactions. Smaller would be easier on the markets themselves, while larger would be more efficient for dealing with fees. There again I think the use of alternative coins for transport seems more likely as some of those transaction costs are quite small compared to BTC or converting to fiat.
The use of multiple currencies, and multiple coins also greatly blunts the argument that any use of arbitrage would tank the market from huge transactions. Split that transaction set up 100 different ways, across multiple exchanges, and across multiple coins. You could easily 'hide' quite a large volume of movement that way. I'm not inferring they hide them to avoid detection here, but rather use hide to indicate a lack of disruption at the exchanges used.
How they're doing it on the $ side.. I dont really care to think about that much, as I don't do that myself. On the BTC side though getting a 1% gain would be quite feasible. Getting multiples more then that would be feasible even for me, but the time necessary to develop that automation would not be insignificant. I'm happy to pay them to do it for me.
Now with that said, I don't have a lot of BTC tied up there either. About .5 BTC. I wouldn't even have that much there if I hadnt done my own calculations to see that its doable though.