The rise was from 600$ at 05:43 UTC to 630$ at 07:00.  Then it dropped 10$ in the next 10 minutes. Whatever displeased the Goddess must have happened not long before 07:00. 

It fell 7-10$ on Bitstamp, BTC-e, Bitfinex 15 minutes ago.

Hm, price fell a bit again.  What happened? Did someone offend an old prof?  I told you that the Bitcoin Goddess does not like that...

* Huobi and OKCoin jumped from nothing at 05:43 UTC, Bitfinex only at 05:44, BTC-e maybe at 05:47.  So Bitfinex and BTC-e can be eliminated. 

* At Bitstamp there was a trade of ~400 BTC at 04:43 but it did not have immediate effect; the price started rising over the next 3 minutes with very little volume; whereas at Huobi and OKCoin trade stayed very high after it started.

* The the initial rise at Bitstamp from 600$ overshoot but quickly fell back, stabilizing at 613$ (13$ increase), while at Huobi and OKCoin the rise was ~70 yuan which is a bit more than 10$.  The difference of 3$ can be explained by "hysteresis", since the real spread at Bitstamp is usually larger than that.

That may indeed happen, but it does not explain why they kept trading vigorously while the price was nearly stable, varying by only 2-3 yuan in each minute.  Why wasn't this volume an hour ago, when price was varying by the same amount?

In contrast Bitstamp has been nearly dead for most of the time.  Obviously it is not the Bitstamp traders who are moving the market.

Perhaps some Chinese investor tweeted to his Chinese followers that he grabbed some lots at the USMS auction for 620 USD/BTC.

On Bitstamp, BTC-e, and Bitfinex, there was a surge in volume (peaking at 300-700 BTC/min) while the price was shooting up, but when the rally stalled then volume dropped again to the very low levels it had before (10-30 BTC/min).

On Huobi and OKCoin, volume surged to about the same levels (~600 BTC/min) during the rally, but then it remained high (~120 BTC/min) while the price wandered between ~3800 and ~3820.

Need I say that I see that as evidence that the rally happened in China, and the trades in the West were mainly arbitrage?

That is why there are so many scams in bitcoinland...  Most bitcoiners aparently expect from exchanges and bitcoin ventures only the same level of business ethics that they expect from a drugs dealer... 

But an exchange that does unethical things will of course "put the clients' money in question", at the very least by spoiling their "luck" at trades.

An exchange that uses a dirty trick against some other client may next use a different dirty trick on you.  That is why one should not tolerate ANY unethical behavior from  them, against ANYONE.  If clients don't mind other clients being scammed, the exchange will have no reason not to scam them, one by one.

That is plenty of time to take advantage of the clients, big or small, even without any fancy psychological heuristics.

Say for example that the current spread is 600$ -- 610$.  At 10:45:00 Alice places a buy order for 10 BTC at 605$.  At 10:45:03, before Alice's bid gets posted to the world, Bob places a sell order for 10 BTC at 600$.   Carl, the exchange owner, then inserts between the two orders his own sell order for 10 BTC at 605$, and his own buy order for 10 BTC at 601$.  Only then he posts the trades to the world, namely

  10:45:01 sell 10 BTC at 605$
  10:45:03 sell 10 BTC at 601$

The first trasaction being from Carl to Alice, the second one from Bob to Carl.  Both Alice and Bob will be happy with the liquidity and the result, but Bob will get only 6010$ for his coins.  Whereas in a honest exchange the only transaction would be

  10:45:03 sell 10 BTC at 605$

from Bob to Alice. In this case Bob would get 6050 for the coins.  The 40$ that Bob failed to earn in the first case are pocketed by Carl.

Certainly many other such tricks are possible, e.g. when a client places a single buy order to scoop up N sell orders at different prices, Carl can run ahead and buy the N-1 lowest sells, then place a single sell order just below the last one.  Or whatever.

That does not seem to bother the typical bitcoin scammer.  I don't know if Danny Brewster got front page coverage anywhere, even in Cyprus; and he obviously did not get much worried about it.  The dozens (hundreds by now?) of lesser bitcoin scammers got no coverage even in the "bitcoin media".

As I said, one would still require a trusted audit to certify that clients are real and that each client got a single client code.  The dirty tricks that I imagine are happening would be thousands of small "thefts of good luck" per day, as illustrated above, that could not be masked by using a few friendly clients in place of Carl.

It also has a quote by the Citibank analyst that seems rather naive.  No one here expects the 30'000 coins to be dumped on the market in a short time, right?

And the prof seems to assume that the auction price will be known right after the closing, but the USMS documents say that someone will have to submit a FOIA to get that information.  (Statements by participants cannot be trusted, they will care much about the effect on price.)

And the prof also claims that the USMS wants to auction the coins before they lose value, whereas I do not see any haste (the auction was authorized quite a while ago) and do not believe they would have that motivation (the USMS has no mandate or obligation to schedule auctions for maximum profit, all it cares is to have a fair auction with enough bidders).

Am I the only one here who does not know what "420" mean, other than it is related to drugs and is not a collection of 10 different answers to the Ultimate Question of Life, the Universe and Everything?

That is because stock exchanges are strictly regulated and audited by the SEC, and many unethical tricks are crimes with stiff sentences.   I was assuming that bitcoin exchanges would remain free from such regulations for a while still.  If they are to be regulated the same way as stock exchanges, then the UCCs may add little to public confidence (but would still be very useful for market analysis).

That is actually very wrong.  The exchange knows the client orders before other clients, so it can scoop up good trades that otherwise would have been exploited by them.  (I believe this is what is called front-running, yes?).  Any money that the exchange makes this way comes from the pockets of their clients, but they would not notice it -- they would just have less luck than they would in an honest exchange, and as a whole they would lose more money than just the trading fees.

I suspect that all exchanges are doing this (and/or other dirty tricks). If something is not illegal, cannot be detected, and is extremely profitable, why would a businessman not do it?

A standard audit of the books and accounts would not detect this.  But with certfied UCCs in the logs, people will notice that one client (or a few "friends of the house") will have extra luck and will always seem to know juicy orders beforehand.

The intended purpose of the unique client code idea was not to prevent criminal use of bitcoins by clients,  but to prevent certain unethical behavior by the exchange, possibly with complicity of some privileged clients.  (Of course it woudl only work  if the uiqueness of the codes was certified by independent, reputable auditors.)

Unique client codes would  also immensely improve our knowledge of the market.  For instance we could tell how many coins have been bought for each given price and not sold yet; that would provide an estimate of long-range liquidity that may be much more rreliable than the order book.  We could tell whether the market is concentrating or distributing bitcoin ownership.  How many active traders there are in each exchange.  And much more.

Wasn't there a movie where the girl said "Love is never having to say you're sorry or to delete your posts"?

Is there also in Dutch a saying like "an idle mind is the workshop of the Devil"?

Since everybody has me on ignore by now, no one will mind me observing that the price on OKCoin and Huobi started to rise several minutes before that big buy at Bitfinex.  Thus that buy may be just an arbitrage trade, delayed by the trading fees at Bitfinex and perhaps by residual price mismatch.

That rise in the Chinese exchanges was almost predictable: now is 6:40 am in China, about the time when the first traders "wake up".  I have not checked statistically, but somehow I got the impression that the early-risers tend to be more bullish than average.  So the Chinese day tends to begin with a price rise, that may be reversed later.  But, again, this is only an impression.

Could it be just frustration with a BTC price that doesn't seem to know where it wants to go?

Indeed.  This finding seems to render the IW team analysis of the blockchain irrelevant.  They looked in the wrong place, so no wonder they did not find anything.  

Assuming that BK was unaware of the flow between the hot and cold wallets, this finding also restores the credibility to his claim.  Back to square zero?

However, wouldn't the hot wallet be caught by the recursive script during its first pass?

I don't know how  InstaWallet worked.  Is it possible that BK was the victim of a phishing-style attack?  Say, he was led to a fake IW site, or the real IW server was hacked to divert some client deposits to an address that did not belong to InstaWallet, and omit those accesses from the database?

Indeed, considering the number of people he has fished embarassing things about, including people with means (and history) of suing people, it is highly likely that he got at least a letter from a lawyer threatening a lawsuit.

Or he simply may have given up hope of recovering his lost bitcoins, and got fed up with a community that, by and large, is indifferent to crime.  When they do not side with the criminals against ther victims.

And why can't they use such an account to pay the merchant?

What does Coinbase do to avoid fees on dollar deposits, that merchants cannot do?

Or are you considering international purchases?  In that case going through bitcoin could save money over paying directly with dollars; but bitcoin-based payment processors do not yet reach overseas, do they?  (And for most overseas purchases, transportation and customs taxes will dwarf the credit card fees.)

Yes, many traders would not like even that limited increment in transparency. But others may like it, if it makes them feel more confident about the exchange.  Exchanges with and without unique client codes could coexist, so that clients could choose.   That is something that the market (or should it be the meta-market?  ) seems capable of deciding.

But I don't think that they would be necessarily inconvenient, even for big traders.   In your example, other traders would not know whether the buying whale has a preset max price, or how much he intends to buy; or whether the small traders will keep on buying too.  And such a buying whale would be sharing the market with many other traders -- including big selling whales, for which the same reasoning would lead to the opposite effect.

Moreover, it is a zero-sum game (in the short run at least), so any change in the rules will have no net effect on profits when averaged over all traders.  To evaluate such a change one must use other criteria, e.g. whether it seems more "fair" to traders and thus avoids them blaming the exchange for heir bad luck.

Interesting, but that is only the minimum of course.  Many bidders probably entered with much more than 25'000$, offering higher prices.  Reversing the math, if the bids for the ~30 kBTC that SM submitted to the USMS were near market price, and every one of the 42 syndicate bidders managed to get a slice in those bids, they must have pledged about 18 M$ in all, or ~430'000$ each, on average.

It occurred to me now that 25'000$ is also the minimum that one can invest in the SMBIT fund.  Has anyone seen the actual terms of the SM syndicate?  Could it be that the winning bidders will get SMBIT shares, instead of the actual bitcoins?