I wonder how much of the rocket is a result of XRP holders switching to Bitcoin.
The rise from the "Dead Sea Valley" (Apr/4--May/19, Huobi at ~2750) seems to have happened in a few isolated buying episodes. Between them the price seems to have wandered either at random or with momentum gained in those spurts.
I can see five such events (1)--(5) on Huobi:
! Start Date+Time ! Length ! Rise ! Amount !
Num ! (UTC) ! (hrs) ! (CNY) ! (kBTC) ! Obs
----+-----------------+--------+-------+--------+----------------------
(1) | May/20 02:00 | 2 | 50 | 10 |
(2) | " 14:00 | 3 | 170 | 18 |
| May/21 | | | | slow rise then flat
(3) | May/22 05:00 | 3 | 90 | 21 |
| May/23 | | | | ramp
| May/24 | | | | slow dip and recovery
(4) | May/25 04:00 | 3 | 250 | 25 |
| May/26 | | | | slow rise
| May/27 | | | | dip and recovery
| May/28 | | | | random drift
| May/29 | | | | random drift
(5) | May/30 09:00 | 4 | 220 | 25 |
To get China local time, add 8 hours to UTC.
The numbers in the "Amount" column are not very meaningful, though: they are just the total Huobi trade volume in those time intervals. They are almost certainly over-estimates since they must include lots of coins sold at Huobi by arbitragers, as well as coins bought and sold multiple times during the rise. The presence of arbitrage also means that one cannot add the amounts from different exchanges.
This pattern of isolated spurts does not seem to be caused by a broad rise of optimism in the market (except for the optimism induced by the spurts themselves).
Perhaps one or two of the spurts were sudden decisions by rich individual traders, for random motives. (A suitcase of cash being successfully delivered to a Chinese exchange, perhaps? A client tryng to escape through the BTC door from Bitstamp's more demanding KYC procedures?)
Or perhaps each spurt is a trader of group of traders obtaining some insider information (such as the imminent opening of the offshore versions of Huobi and OKCoin, trading CNH insteadof CNY?)