Indeed, my theory for the gradual decline in the price since Feb/06 is basically that: Chinese traders who experience a few losses in a row are selling their coins (since they are not very useful in China) and taking their money out; and not enough new clients are joining, so that there is not enough new money coming in.

Additionally, some Chinese miners who need Yuan may be selling at the exchanges, adding coins and taking yuan out.

The OKCoin one is dated TODAY 12:29 UTC, exactly when the last large  drop (~3050 --> ~2950) started.
The Huobi one is from yesterday 09:30 UTC, when the previous large drop (~3150 --> ~3050) started.

In both cases there may have been a bit of "insider trading".

See my previous post.  The time of the OKCoin notice apparently coincides with the start of the current dump.

It may be that the Chinese exchange CEOs have karpeled a little; they may have known what was coming but waited until the last minute to tell their clients.

I am not sure, but the OKCoin note may be saying that further notices will be shown only to logged-in clients "in order not to affect the normal market conditions".

The Huobi note link is in red, next to a red light bulb, near yhe top of their homepage.  The OKCoin is in red too, next to a loudspeaker icon.  Further notes mat be accessed perhaps by fiddling with the URLs. (Huobi's note 325 is some Western press news, 326 does not exist yet.)

These notices on the Chinese exchange websites may be relevant, especially the OKCoin one below:

Huobi, April 17 17:16 CST = April 17 09:16 UTC
http://www.huobi.com/news/index.php?a=show_notice&id=324

OKCoin, April 18 20:29 CST  = April 18 12:29 UTC
https://www.okcoin.com/t-1008440.html

Não sei se tem muito conteúdo.  

Ele quer testar a teoria de que as pessoas procuram ouro e bitcoin nas mesmas ocasiões, quando há ameaça de alguma crise econômica ou algo assim.

Então ele compara a a curva de preço do ouro com a do bitcoin.   Só de olhar dá pra ver que as duas não poderiam ser mais diferentes, e a teoria não tem fundamento.

Mas ele não desiste, então em vez de olhar para o preço ele olha para a variação do preço - diferença de um dia para outro.  Não sei se faz sentido.

Aí ele calcula correlação estatística entre as duas coisas e  conclui que quando as pessoas gostam de ouro elas não gostam de bitcoins e vice-versa. Mas a correlçaão é muito fraca e não significativa.

Tenho umas críticas:

(1) Não está claro se, ao passar de preços para variações de preço, ele usou escala logarítmica (i.e. o preço subiu/desceu X por cento) ou diferenças absolutas (subiu/desceu X dólares).  No segundo caso a resposta sera muito distorcida pois as diferenças antes de Abr/2013 seriam irrelevantes de tão minúsculas.  Infelizmente, pelo segundo gráfico parece que ele calculou diferenças absolutas.

(2) Calcular correlação cegamente não é sensato, pode haver um único ponto com valor absurdo ("outlier") que mela tudo,  Ele deveria procurar os pontos que mais contribuem para a correlação e decidir se são legítimos ou "outliers".

(3) De quaqluer maneira, os dados de Nov/2013 em diante vão dominar a correlação.  E como nessa época o ouro estava meio parado, a conclusão, sem surpresas, foi que não há correlação.

(4) De qualquer maneira, até Abr/2013 pouquíssimas pessoas conheciam bitcoin e pensavam seriamente nele como investimento semelhante a ouro.  Então não tem sentido procurar essa correlação antes de Abr/2013, ou mesmo antes de Nov/2013.

(5) As grandes subidas de preço em Abr/2013 e Nov/2013 foram claramente causadas pela abertura do mercado de bitcoin na China, e não porque as pessoas valorizaram bitcoin como investimento.  Os Chineses tem usado bitcoin principalmente para especulação de curto prazo ("day trading").  As grandes variações no preçodo bitcoin desde essa época foram todas devidas a mudanças nas regras bancárias e comerciais da China em relação ao bitcoin, que não tem nada a ver com investimento de longo prazo. Logo qualquer correlaçao que houver será por acaso.

(6) Ele acha um "p" de 9.2% e conclui que tem 1 chance em 11 de obter esse valor porcoincidência,  Na verdade é o contrário, p=9.2% significa que tem 91% 90.8% de chance (10 em 11) de obter uma correlação como a que ele obteve, ou maior, sem que exista uma relação de fato.

Ou seja, ele não achou nenhum resulatdo significativo, e não me convenceu.  

ATUALIZAÇÃO:  Escrevi um comentário na página do artigo, apontando alguns desses problemas, mas ele foi apagado. Ó vida.

PS. As for whether they are bearish or bullish at different hours of the day, that is something I actually wanted to check.  Unfortunately I still don't know of a convenient source for the summary data.  (Bitcoincharts has the data for download, but they dont carry Huobi or other Chinese exchange except BTC-China, which has rather small and weird volume.)

EDIT: China is UTC + 8 hours, so 03:00 am of day N in China is 19:00 of day N-1 UTC.

All I know is what I see on the bitcoinwisdom hourly volume charts.  Huobi's clients generally seem to trade substantially until midnight, then gradually stop until ~3:00 am, and gradually start again from 06:00 on.

I don't see any difference on Friday evenings or even Saturdays, but they seem to start later and trade less on Sunday mornings.

OKCoin has a background volume that seems to go on continuously 24/7.  Could be fake, I don't know.

Not at all, it was just a plain PNG delivered from the University server.  The reason to post it as a link was that the bitcointalk image proxiy often truncates PNGs for no obvious reason, and it did so in the preview.  I have re-delivered the Lesson without the link.

(If sending links in private messages is forbidden, why does the software allow it?)

Really? 

I did not intend them as trading advice, thye are just an idea that I have been testing and do not know whether it is worth anything.  I m glad (and relieved) that they worked for you.

Obrigado, é bom saber que estou trazendo alegria ao mundo. 

Oh! Wel, well, well... 

It is decided, then.

In order to avoid collateral damage to innocent readers, I will deliver the Exemplary Lesson to @JayJuanGee by private message.  If he never shows up again in this forum, you will all know why.  If he survives (and I sincerely hope that he will), perhaps he will recover enough to tell you all how bad it was.  In either case, all the other sinners out there will have a chance to repent.

PS. By the way, about the Neo & Bee TV ads broadcast in Cyprus: were they in Greek, or in English? (The ones that were circulated in this forum were in English, created by an UK agency it seems; but I assumed that they would be translated.)

Well, the story has been picked up by some other newspapers, and the arrest warrant seems to have been confirmed by some reporters:

Kathimerini  (CY, weekly)
http://www.kathimerini.gr/761354/article/epikairothta/kosmos/diey8ynths-etaireias-bitcoin-egkateleiye-thn-kypro-meta-apo-apeiles

iCyprus
http://ikypros.com/?p=8302

Shangai Daily (CN)
http://www.shanghaidaily.com/article/article_xinhua.aspx?id=213327

Daily mail (UK)
http://www.dailymail.co.uk/news/article-2603742/British-boss-Bitcoin-virtual-currency-firm-wanted-Cyprus-suspicion-fraud.html

Financial Mirror (CY)
http://www.financialmirror.com/news-details.php?nid=32369


The Cyprus Mail seems to be an old daily newspaper based in Nicosia, the only one in English language.  is is appropriate toc all it "local"?
https://en.wikipedia.org/wiki/List_of_newspapers_in_Cyprus

As for other (Greek language) newspapers in Cyprus, I don know what to search for with Google.  If they did not report on the affair, I would rather conclude that Neo & Bee was not such a big thing in Cyprus after all...

But the most significant thing,for me, is that there are no reports that contradict the Cyprus Mail, other than forum posts by anonymous users and by Mr. Brewster's himself. 

If you care to know, I would give 80% probability that the Cyprus Mail and other newspaper reports are essentially correct.   Sorry, but my life experience taught me that statements by someone in Mr. Brewster's situation should not be given any weight whatsoever.  If he wants to benefit of the "innocent until proven guilty" principle (that is appropriate in courts, but disastrous in everyday life) he should turn himself in.  If he wants his statements to be taken seriously, he should make them to the court, under oath and pain of perjury.

If the number of active wallets N(T) was connected to the price P(T), in four months during which P(T) varied up and down several times, by 40% or more,  we would expect to see at least some corresponding variation in N(T).  I don't see that.

If we are supposed to look only at the long term (3 years) and ignore any deviations spanning 4 months or more, then the statement is not convincing because we get two steadily exponentially quantities, so that any other exponentially increasing quantity -- such as the increasing quantity of CO2 in the air -- would "explain" the increase inP(T) just as well as N(T).


I don't know why some Americans are worried about the American economy; if one looks at the last 600 years, ignoring small variations, it has been a tremendous growth. 


It is a pity that my Exemplary Lesson got punctured, you would definitely need to get it now.

Look, I understand that you are frustrated by this irrelevant price oscillation that has been going on since December.  But it is not my fault, is it?  You must have done something to deserve such punishment from the Bitcoin Goddess.  Did you perchance tread into her temple with dirty shoes?  Break a sacred jade mirror? Kill an albatross? Offend an old teacher?

This thread is starting to read like some Theater of the Absurd  play... 

Asking for evidence in general is not an accusation.  However, the people who have been asking for evidence (at least some of them) explicitly denied that the Cyprus Mail artilcles were evidence.  What else could they be then?

I have my evaluation of the situation, my standards of evidence and my way of making sense of what I read, thank you.  I write my opinion based on that evaluation, and how I got to it.  Everybody is entitled to a different opinion, and no one is obliged to find my arguments convincing.

Some quite emphatic but quite contradictory testimonials have been posted here.  It follows that least some of the people posting here are shameless liars.  Naturally, I consider any factual statement posted here as very weak evidence.  When several such testimonials and external documents agree, and none contradicts them, I may give a high probability to the statement.  When testimonials and/or external information disagree, I must try to guess which is more credible, and assign appropriate probabilites to each version.

I am curious about the schooling and professional history of Mr.  Brewster in England, because they obviously could change a lot my evaluation of the situation.  In fact, I would think that any investor should want to to know that basic information before investing, and anyone who solicits investment is obliged to provide it.

But there does not seem to be a positive correlation between that number and the price, since December:
https://blockchain.info/charts/n-unique-addresses?timespan=180days&showDataPoints=false&daysAverageString=30&show_header=true&scale=0

In fact, from Dec/01 to Feb/28, the correlation seems to be strongly negative - when one increases, the other decreases.

By the way, why would that number N(T) decrease? Does it exclude addresses that are were used before time T but are empty at time T?

Chinese Slumber Method prediction for Friday April 18

Prediction valid for: Friday 2014-04-18, 19:00--19:59 UTC (not before, not after)
Huobi's predicted price: 3001 CNY
Bitstamp's predicted price: 483 USD





[ Plot legend ]

EDIT: legend link fixed

Today's data point was very good, (S = 0.0009, W = 0.982), but substantially below the previously assumed trendline.  It does not seem possible to fit the last three or four points into a plausible trendline. (A parabola might do, but it is hard to justify it in terms of fundamentals.) So we assumed trend break, and a new straight line trend defined by the last two points only; namely, A + B*(d-d0), where d-d0 is the number of days since Apr/16, A = 3185 and B = -92.

The Bitstamp prediction, as usual, is the Huobi prediction divided by the currency conversion factor R, which was assumed to be 6.21 CNY/USD. It was 6.21, 6.23, 6.32, 6.38, 6.21 at the last five Slumber Times.
 
Checking the previous prediction

Prediction was posted on: Wednesday 2014-04-16, 22:59 UTC
Prediction was valid for: Thursday 2014-04-17, 19:00--19:59 UTC (~20 hours later)

Punished for being bull, again:

Huobi's predicted price: 3198 CNY
Huobi's actual price (L+H)/2: 3093 CNY
Error: 105 CNY (~17 USD)

Bitstamp's predicted price: 513 USD
Bitstamp's actual price (L+H)/2: 498 USD
Error: 15 USD
 
NOTE: "You don't even have a clue as to which clue you're missing." -- Miss Manners

This thread, and its "moderated" (?!?) spin-off, have been pretty boring of late.

To those who demand evidence of everything: you are accusing the Cyprus Mail of lying and/or of copy-pasting their article from this forum without bothering to make even a single local phone call to the N&B offices to check.  That is a serious accusation, so it is your turn now to provide the evidence of that.

As for Danny's  background before going to Cyprus, we have only a claim that "he worked as security in nighclubs and pubs in Sheffield UK".  Are there any public records of that?  Did he have any schooling after high school? Did he ever provide a biography? 

I do not think that it will get that far, but they may get discretely "invited" to a friendly but frank talk with some government official who will further clarify the "clarification" that apparently wasn't sufficiently clarifyiing to them. And the the guy will not be from the PBoC.

See my previous posts, this apparently was the cause of the short-lived bump at 04-17 05:12 UTC. 

I belive that these "TOP-UP" cards are not the same thing as "3rd party payment processors depositing directly into Huobi's  bank account".  There seens to be a "4th party" involved, a card-issuing company.  The card company receives the money from Huobi clients via bank wire transfers or the 3rd party payment company, "recharges" the clients' cards, and then the cards can be used somehow to deposit money into the clients' internal Huobi accounts .  Thay is what I fantasized understood from Google Translate's output.

I wonder if the PBoC will be happy with that solution, but I would guess that they did not ask the PBoC beforehand...