in previous discussions, you chose to include Chinese volume 1:1 in your analysis (i.e. you consider it as "real" as the volume on the other exchanges), while I consider large parts of Chinese volume meaningless for most analysis, at least when comparing cross exchanges or aggregating volume.
I am still not convinced that a "deflation factor" is justified there.
Volume is interesting, I presume, as a measure of liquidity. Zero fees bring in more traders and also allows traders to trade on smaller spreads. So there are both more coins in the market, and the same coins get traded many more times per day.
Should the second effect be discounted, or should it be counted as measure of liquidity, just like the first one?
Repeated trading does not provide liquidity for large buys or sells, but it shoud work for smaller ones. With zero fees, a trader who just bought at 500$ may be willing to sell again for 501$. With fees, the same trader would probably hold back. Thus, even with the same traders holding the same positions, a zero fee market would provide higher liquidity than one with fees.
Does this make sense?
In terms of market liquidity, you are absolutely right of course. But I'll try to clarify where I'm coming from when I look at volume as someone who tries to get trading insight from it:
1) volume now in relation to previous volume, same exchange, usually short history only. The much beloved "classical" TA that you appreciate so much. The simplest example is that of a capitulation and corresponding volume, for example. In that case, I don't care about volume being human or bot, based on the assumption that bot volume will represent an approximately equal share of volume during all periods (within some reasonable time frame).
2) aggregate volume over several exchanges, sometimes over a long history. This is the case where I heavily discount bot volume. Keep in mind why this volume analysis even exists: because for traders/speculators, it is perhaps the only available window into the actual on-exchange money flow (unless you happen to know the owners of mtgox or bitstamp, perhaps
). Back in the golden days, it was easy to do volume price analysis like that, across eras and exchanges: there was only one relevant exchange. With the advent of zero fee exchanges, this methods obviously doesn't work anymore - automatized trading volume with very low cost (zero fee, aiming at high frequency of small profits) quite obviously (to me at least) does not bear the same relation to money flow into Bitcoin as does a large, costly human order.
How to go from the assumptions in 2) is far from obvious, but perhaps it gives you an idea why I'm interested in ways of adjusting CNY volume in the first place.
