@ slipperyslope, peter r
do you think transaction volume is largely a lagging indicator?
I am hoping to compute the cross-correlation function between number of transactions,
N, and price when I get some more time. However, it is already clear that at least at the "peaks,"
N is a lagging indicator. If you zoom into the Metcalfe Value graph, you'll see that
N increases for a while after each peak in market cap. I think this makes sense because people continue to furiously send coins to the exchanges to dump!
I believe in the bigger picture, there is no clear leader or lagger. If price increases, then this attracts media attention bringing in new users (causing an increase in
N). If the bitcoin economy grows (
N increases organically), then I think we expect this to cause an increase in price based on the quantity theory of money (since we are growing the bitcoin economy).
EDIT: but still
N seems to increase with less volatility than price. It appears more consistently straight on a log chart.
how much emphasis should we place on coin trade when viewing the correlation between transaction volume and price, or does the exclusion of popular addresses nullify this by disregarding exchanges for the most part? still in that case, how do we determine how much of that volume belongs to personal exchanges of btc/fiat?
I don't think it matters. As long as the % of economic transactions to non-economic transactions remains relatively stable, I don't see why the absolute number matters (and we can't calculate it anyways).
Agreed - absolute value is irrelevant. More interested in a determination of - as you say - economic vs non-economic transaction volume as related to price discovery as a coefficient, something you have already shown to some extent, great work.
As infrastructure evolves and becomes more complex we should see a decrease in the ratio of acquisition vs trade transaction numbers even with an exponential increase in acquisition numbers. Something which is likely obvious to some who have manipulated the data in some way, and would certainly be reasonable. That is my main point of consideration.
Side note -
N drop lagging past price peak can be a variety of reasons - coin dumping at exchanges, coin acquisition from new buyers still taking place as a result of a) exchange administration in verification and fiat transfer b) cotinued mania of ATH + runup (esp if absolute value of acquisition on average is smaller then size of large holder dumps trading the market) etc
EDIT Also agree that bitcoin economy = speculation + trade atm - market is coming to life...