Bitcoin Prices Follow Dollar with 0.89 Correlation

[bitfair]

Over what time period??

I don't think this correlation is valid - maybe for a very short time period. My own long-term correlation studies (utilizing ALL the data) don't match up with this.

It seems to be a textbook example of what econometricists call "spurious regression", and occurs surprisingly often with timeseries data. [Check the Durbin-Watson statistic to be sure!]

I.e. the high correlation is probably meaningless and there is no relationship.

[BrightAnarchist]

Over what time period??

I don't think this correlation is valid - maybe for a very short time period. My own long-term correlation studies (utilizing ALL the data) don't match up with this.

Just the last 50 days.

OK, that makes more sense. 50 days is pretty much random IMO. The study I ran shows a strong correlation with the S&P, but that was over all the data. The dollar more or less moves inversely to the S&P, hence my surprise at the OP.

[molecular]

There seems to be quite a bit of denial about this. Something like: "Bitcoin couldn't possibly follow a doomed fiat currency."

It's not that bitcoin is following the value of the dollar so much as it's following the dollar's rhythm.

keep in mind the USD Index used here just measure USD in terms of other fiat currencies. If value is lost on all fiat currencies evenly, the USDX won't move.

If you want to talk about the "value of the Dollar", you'd have to look at it in terms of hard assets, maybe stocks and the like. Possibly BTC in the future. One can clearly see (overlooking the more recent move) that the USD is going down down down.

[notme]

One can clearly see that the USD is has been going down down down.

FTFY

I wouldn't overlook the recent moves that are occurring simultaneously in all major markets.

[molecular]

One can clearly see that the USD is has been going down down down.

FTFY

I wouldn't overlook the recent moves that are occurring simultaneously in all major markets.

when everything measured in something moves unisono, maybe it's the something that actually moves.

[notme]

One can clearly see that the USD is has been going down down down.

FTFY

I wouldn't overlook the recent moves that are occurring simultaneously in all major markets.

when everything measured in something moves unisono, maybe it's the something that actually moves.

Right.  Everything priced in fiat has been falling, thus deflation.  Yes, we have had quite some inflation in the recent past, but I wouldn't discount the potential tops forming.

[jago25_98]

I think an important test of Bitcoin is will people outside the USA buy Bitcoin to give a support level if the USA was to drop massively thus leading desperate people to cash in their BTC savings? - how to analyse that other than simply looking at country distribution?

If we look at:
https://en.bitcoin.it/wiki/Bitcoin_Ladder#Top_countries

we see a few BRIC countries...
 key ones for me are Poland & Russia. Brazil seems a bit reliant on USA trade? - as do the other big ones?

 So it's difficult to say what if would take for BTC to decouple to the dollar. One thing's for sure - we need to get Bitcoin spread more multilaterally & economically in order to further secure it. Growing use in USA antagonistic countries such as Venezuela & Russia is bullish. states.


edit: This presupposes a drop in the USA dollar and stocks. While many have been going on that the USA is about to fall for ages you have to remember that it is the global ruler and if times get tough cash tends to flow to the core, hence the USD bull run.

[nrd525]

You can get all sorts of correlations if you shift things around 1,2 or 3 days forward or backwards for short time periods.

Correlation is not causation.

A famous example is that some economists used to think that the solar cycle heavily influenced the economic cycle =)