I'm not looking to long-term hold, so the premium isn't really an issue. It's the godawful slippage on bitstamp for anything bigger than tiny amounts.

The only thing that prevents me from signing up is, well, that it'd be idiotic to do so, I guess.

I must be the only person on earth right now who contemplates beginning to trade on/sending fiat to mtgox

Hardly. 2 months ago, mtgox taking the lead during surges meant twice or more the volume of the next biggest exchange. These days it means about 50% more.

It's too early to declare the mtgox era entirely over. But it's getting closer.

Did I mention I don't really know much about statistical methods? So I'll need to ask some dumb questions: that beta distribution is not motivated by anything bitcoin specific, right? You're assuming it is a good representation of peak/reversal likelihoods as a function of extreme price change, correct?. Also, I don't see how you get the actual multiplier that graph. What's the PDF that ranges from 0 to 2.5?

Another idea/question: your random walk approach worked quite okay with price change as input. What do you think are the chances that you can get some complimentary information (maybe estimating reversal likelihood) by applying it to a different domain, say, btc volume, dollar volume, OBV. Not at all sure if that could work, but I'm wondering if price (change) is the only useful input. Any intuition about this?

Instead of looking at walls unfiltered, I prefer to see how the bid/ask ratio changes over time. 3 days ago I posted here that this ratio went up drastically, and price followed (125 to 128).

The reversal of that ratio, starting yesterday, was even sharper. And price is already following again. Here's the corresponding 1w graph from coinorama:



How do you read this? And don't just say "manipulation", but specify what's the goal of it. 'cause it sure doesn't look like straight-forward pump-and-dump, the increase was too incremental, and there was no huge sell-off at the peak either.

I admit, I'm unsure what the reason is behind this. Could be an attempt to artificially prop up price, before the July uptrend finally breaks down. Could be an attempt to actively *bring down* the July uptrend. Could be plain old market insecurity ("is the trend over? or not?"). Your guess is as good as mine.

I see it now. On the daily test I still can't really make out the difference between median and actual price, but I looked at the weekly result zoomed in and it's pretty clear there.

What is interesting is that weekly works quite well as a leading predictor even during volatile times, like the extreme run-up in before the April crash, but it completely fails when it comes to crashes -- the April 10th one, but even slightly less drastic ones, like the one on September 4th. (if I'm reading the chart correctly now)

Sorry if that's stating the obvious, but it's just that the gap is pretty noticeable. The method works well while a trend is continuing, but as it is now, it seems to have no chance to predict a sudden reversal.

Still don't get it. I can't see a blue line (for actual price) in the topleft chart. And for test 4), if you backtested one day at a time starting about a year ago, shouldn't there be 365 of the lower histograms, each with a 1 day prediction period?

I admit defeat. How do I read those latest results? Are the "sliding" test results only visible in the upper log chart? If so, how can I compare test result with actual price?

Why yes, you're right of course. Just this morning I bought my usual cup of coffee with a gold doubloon.

If you like to think of current day usage of gold as non-niche, that's fine by me. That's just terminology. My prediction is that the current financial system might undergo subtle changes as the result of bitcoin saturation, but that I don't buy the idea that we're going to live in an inflation free, zero-taxation utopia of Randian proportions. In that sense bitcoin's current purpose will remain "niche", even if it will be widely adopted.

I'll paraphrase: we're at 140 now. Going back to either 127 or 129 within the next few days would be a (at least minor) reversal - or if you prefer, call it retraction. And I don't see any signs pointing towards such a reversal. If at all, I see the uptrend becoming stronger again (main point: price and order book development on bitstamp)

I think I liked your other indicator better. But I'll say that much: if the (short term) reversal that would be needed to reach your numbers takes place over the weekend, I'll give your new method all the praise it deserves. I don't see any signs for that happening, but then again, I usually only pick up trends once they started forming, while you seem to attempt to look ahead a bit more.

Seems to hold, so far. From around 125 (on the 16th, when the ratio started going up) to slightly below 128 today. Small, but steady.

Don't confuse "market forces" with "Internet libertarians", i.e. us. The latter may drive early bitcoin adoption, but the former are probably more important in the long run. Don't lose all hope, though. There will be a niche usage opportunity for truly anonymous wealth transfer/storage, either in the form of another cryptocurrency, mixing services, etc. Think Tor network... available, used, but not mainstream.

I have that line, embroidered and framed, hanging right over my bed.

+1

Also, A/D is a tricky beast. On mtgox, it continued to go up all the way from April to July, and only started falling (!) right before the July reversal. Since then, it mostly flatlined, with a slight downward tendency, while price went from 70 to 140.

On bitstamp, on the other hand, it has been going up almost without interruption ever since July. I simply don't see how an unfiltered reading of A/D is helpful at the moment... gox fiat problems might have to do with that divergence.

There's also the problem that bitcoin transactions are far from "anonymous", which is the premise from which the entire argument starts. In reality, I believe, governments' reactions to bitcoin saturation will be similar to their reactions to Internet saturation: accept (mostly) the overall increase in freedom and citizens' possibility to circumvent certain laws, but pour substantial resources into detecting (and prosecuting) violations that they (and perhaps their electorate as well) consider especially grave, example: child pornography.

Applied to bitcoin this could mean that low level tax evasion will become easier in a Bitcoin environment, and might become more prevalent, but those who try the same with much larger sums will probably be surprised by the creativity governments can develop when they feel threatened at the core. Personally, I don't even expect government size to shrink substantially as a result, just that the *possibility* will exist to more easily avoid taxes, of which only a minority will make use in most countries.

Yeah, I wrote earlier already that I'm pretty pissed off at both bitcoincharts (for not giving a fuck about the second biggest exchange) *and* bitstamp, for not getting in contact with bitcoincharts and solving the problem.

For now, bitcoinwisdom is a (far from optimal) substitute.

Really interesting. Thanks for sharing this. By the way, I can see the images fine.

Some notes/observations:

the 95% interval is, hm, sort of useless for our purposes, isn't it? The range is huge (as you would expect), so not much information can be gained from that. Except maybe to tell the uber-bulls that a 100k $/btc price next year is exceedingly unlikely :P

at a glance, the median appears to be more accurate than the mean. What is really puzzling is that in your backtest looking 1 year ahead, a) mean and median diverge a lot (350 vs 149), and b) the meanmedian is actually remarkably accurate. Just a lucky guess, or not, I don't know. But that's a pretty remarkable result for a relatively simple model, no?

Would be interesting to see how the model works in a "sliding" backtest, say: 1 week prediction, starting 1 year ago, moving forward 1 week each iteration. Looking at the result of your backtest, the median might turn out to be a pretty good predictor of the price one week later, which for trading purposes would be already very useful, but it would be *really* interesting to see how the model performs a week before April 10 2013, or a week before one of the other wild price swings in May.

EDIT: wrote 'mean', meant 'median'

Nice find. But I have no way of verifying if that's really the reason why bitcoinchart isn't pulling data anymore. Other sites still get the recent bitstamp data (bitcoinwisdom, for example), so either they implemented the fix you linked to, or the bitcoinchart situation is caused by something else.

Fun!

I'll have to give a bit of a range though...

Monday, September 23, 2013: 134$ to 146$