So with the halvening coming up this year, I got inspired to do some data analysis on the bitcoin supply. I already knew that we are ahead of the blocking "schedule" if you calculate the number of blocks since genesis and multiple by 10 minutes each, but I wasn't terribly familiar with the road we took to get to the current point.
I figured I might as well share the information here, although I think I'm the biggest numbers nerd in gethashing. Still, maybe someone will find some of this interesting. Plus, I'll use pictures!
So here's the basic graph everyone has seen of the bitcoin supply vs time. I should mention the data-set I'm using is from blockchain.info's graph of supply vs time, so my highest accuracy is a 24 hour period. That data-set is available here.
Most people have probably seen this before, and blockchain.info has a better picture anyway, but it was the starting point of my analysis.
So based on this graph, we can see some basic features of the history of the btc supply. Namely, it started off pretty poorly. We were way behind for a good chunk of btc's existence. However, we can also see on the graph that it started to trend upward and never looked back. We not only overcame the initial deficit, but tracked well with the expected supply all the way till the first halving. At that point it seemed like we beat the supply estimate continually.
Ultimately, this graph wasn't quite accurate enough to give a good picture, so I ran more numbers and generated expected vs actual blocks found rather than total supply.
This graph gives us a decent view of the phenomenon since it removes the slope changes caused by the halvening. In this graph we can clearly see the negative trend and the upswing. What's really interesting to me is that the upswing stopped after a bit and we stayed level with the expected supply for a while. Still, at a certain point we started diverging from the expected supply. The one thing that bugged me about this chart is that I could tell it wasn't a consistent slope, but I couldn't see exactly where the changes were occurring. This led to the final graph:
This graph shows the difference between the number of expected blocks and the number of blocks actually generated over time. Here, we see clearly the periods from the other graphs. There is the initial lagging of the mining power from January 2009 to January 2010 (roughly), with the sharp upturn from January 2010 to May 2011. At that point, the network managed to stabilize somewhat, as we see the supply and the miners balancing each other fairly well between May 2011 and January 2013. In January 2013, we see another huge rise which lasts until September 2014. We also see the beginnings of another rise that we've all seen over the last few months.
It's important to remember that this graph is showing the portion of the network that is re-factored every 2 weeks to be stable. That means that when you see a steady increase over a long period of time, the network is constantly being beaten at it's estimates. In mining terms, thats A LOT of hashpower coming online consistently.
One of the reasons I chose to use the supply as my basis for investigation is because this is the only data that's actually raw. When you look at hashrate charts, those are all calculated based on block times. There may have been 300PH on the network at a given time, or there may have been a very lucky 150 PH. The supply, however, is a real number, not calculated, aside from the summation.
So after looking at the data, there are a few things we can say for certain.
We are ahead of schedule with the bitcoin rewards. By my calculations we are currently 164.44 days ahead of schedule, and we look to shave even more off. This primarily means halvings will occur sooner than expected. The current estimate for the halvening is 2016-07-18 16:55:31 UTC. Originally, this halvening was going to occur in 2017.
We have not seen large amounts of hashpower taken offline very frequently. Or at least more has come online to replace it quickly. The last halving occurred on Nov 28th, 2012. If you look at the graph, you'll see nothing really. There is an extremely slight down-trend, which may have been masked by the uptrend in early 2013, but it doesn't seem like much got turned off. Now, obviously we're talking about far different levels of profitability given the price at the time. For comparison, here's the btc price chart overlay.
There's some really interesting interactions between the price and the supply, but I would suggest no-one tries too hard to link them, the data doesn't really support it.
Some final numbers:
We are currently 23,679 blocks ahead, which at 144 blocks per day, puts us at 164.44 days ahead. We are ahead of the projected supply by 711,850 bitcoins. At today's price of roughly $430/btc that's $306,085,500 in market cap that has been introduced over the estimated amount.
We are living in interesting times, my friends