legendary
Activity: 1470
Merit: 1007
Nice find, Adrian-x.
Too bad, it looks like chodpaba is on hiatus again. I get the impression he (she) is most active during relatively uneventful periods, experimenting with new indicators to see which way the situation resolves. I guess from a theoretical point of view, "exciting" moments like the one now aren't that interesting -- plus, there's plenty of trading to do :D Anyway, consider posting more often again, chodpaba.
Re: (global) trade volume vs. market cap.
Total USD trading volume on the big 3 exchanges for the past 365 days (bitcoincharts data) (converted to USD for btcchina, assuming fixed rate of 0.15): 2.7B mtgox, 1.4B bitstamp, 2.4B btcchina. Note that some people dislike taking the trading volume at btcchina at face value because of the 0 trading fees, and perhaps they're right to a degree and btcchina's volume has to be slightly discounted, but probably not by much, so let's just add it all up for simplicity to: 6.5B USD.
Calculating market cap based on the slow 1d SMA100 across the big 3 = 414 USD, so ~5B market cap.
Taking the faster 1d SMA30 = 895 USD, we get a ~11B market cap.
Considering that we peaked at 1200 USD for a market cap of 14B USD, twice as much as the (approximate) global trading volume, a correction was perhaps no surprise.
* * *
[long-ish rant]
Chodpaba is right of course in ridiculing the (naive) idea that market cap (current unit price * total no. of units) somehow needs to correspond to the total amount (of USD) that had to be deposited at the exchanges and used for paying for BTC. It's a recurring theme on crackpot economy shows: "Why, if there's 1000 stocks, and you pay me 10 Dollar for one, then suddenly they're worth 10k USD, CREATED OUT OF THIN AIR. So obviously the stock market is a sham."
So he's right of course, but not entirely so: he's leaving out the part that there is in fact some subtle relation between the market cap and trading volume, and perhaps even between different types of trading volume (the point is simply that this relation doesn't need to be a 1:1 ratio). Example: In terms of trading volume, it makes no difference whether the same N participants keep trading the same USD and BTC back and forth, altering the price and the distribution of USD and BTC in the process, or whether a constant influx of new participants brings in new money. This is a somewhat artificial example of course: a constant influx of new investors will almost certainly yield a higher price and a higher volume, but at least in principle the same volume number could be based on a smaller "insular" group or growing group of investors. I think it is self-evident however that the two scenarios are not the same at all for the long-term price development...
Which is why indicators like CMF, PVT and OBV exist in the first place: the market cap cannot be taken at face volume, we don't have complete information on all market participants (on an individual level), so people have been struggling to come with ways to gauge how money enters (or leaves) the market for a long time. Those measures are somewhat crude (they all essentially say: if price rises, the volume is good, add it. if price falls, subtract it. minor differences apply, e.g. adding volume proportional to price change, but in principle that's what they do), but they probably have to be that way, too much information is missing.
I've been toying around with ideas how to do something with a similar goal in mind, gauging what I call 'money at stake'. One very preliminary idea takes into account not just the executed trades, but the order book as well. The obvious objections to the order book apply of course: incomplete picture of available funds, and no guarantee of sincerity. It still seems like a promising approach: an attempt that tries to determine "how much money is actually there", and "at which price it is willing to enter the market" cannot completely ignore the order book I believe.
[/long-ish rant]
