Thanks! Well, between the authors of that study, who saw no arbitrage opportuities, and those two traders, who made piles of money with arbitrage, I would rather trust the latter. Don't know why, just a hunch.
By the way, the flaw noted by the critics -- that the study only looked at closing prices every 24 hours -- is also present in the paper from a Czech (?) author that was cited here a while ago, that "proved" that China dit not lead the price changes. That paper did not specify the time resolution of its data; but since it was downloaded from Bitcoincharts.com, it could not be less than 1 minute, and probably was more than that.
As those two traders note, arbitrage between bitcoin exchanges is quite fast: it usually propagates large moves in less than one minute. (Smaller moves may not propagate at all, and trend reversals may be delayed by the spread.) So it is not surprising that he could not detect a clear leader from that data...
Another problem with both the Czech article and the bank report is that they omitted Huobi and OKCoin, using instead BTC-China as representative of the "Chinese" market. But, if (as I believe) the November bubble was due to the adoption of bitcoin by the Mainland Chinese speculators, then BTC-China was more likely to be one of the followers, rather than the leader. Moreover, after last December BTC-China became a very minor player in the Chinese market.
Because of their no-fee policy, OKCoin and Huobi usually have very narrow spreads, and track each other very closely (often les than 0.50 USD apart), down to the smallest bumps and dips, except for very
short fast spikes. That can only be due to very efficient arbitrage between them, possibly by the exchange owners themselves.
From all I have seen, I still believe that Huobi and OKCoin are the market leaders, who have been defining the BTC price, since last November; and that the other exchanges mostly follow them through arbitrage.
Poll
