Help refine this bitcoin ecosystem health check methodology

See Monday’s post on Daily Fintech for background.


We look at 4 indicators

Indicator 1 = Price.** What do investors/traders think? This shows the wisdom of the crowd with skin in the game. Yes, we can also see the madness of crowds in the charts, which is why our health check looks at 4 indicators not one. The problem is that you can assess bitcoin as a store of value in three ways:

  • currency (like USD)
  • commodity (like gold)
  • startup stock (like Facebook 10 years ago).

The latter is where we get the wild forecasts of a single bitcoin being worth $100,000 or $1,000,000 (from about $966 as I put key to pixel). Before dismissing that as crazy, consider the fact that one share of Berkshire Hathaway is worth about $250,000. We don’t normally see this because standard practice is to issue more shares to keep the individual share price low enough.

Like a stock, the supply of bitcoin is fixed. In the case of bitcoin, it is fixed at 21 million. If you own 210,000 bitcoins you own 1%. That is why bitcoin people talk about market capitalization (which is not how you talk about currency).

Bitcoin as an asset class is gaining some momentum. The near zero correlation to other asset classes is attractive. Volatility and liquidity look fine.

This is where the other tests matter. Investors will happily take a punt on a big upside if the downside risk is protected. If you bought bitcoin in 2009 it was like buying founder stock – you have no downside. If you buy in 2017, you spend real money. For downside risk to be protected, bitcoin must be more than tulips. It must be a real currency. Which brings us to Merchant Transaction Volume.

**Indicator 2 = Merchant Transaction Volume. Is bitcoin being used as a currency in the real world? What do merchants and their customers think? This is a good forward indicator – like a blood pressure check. The problem is that this data is tough to get. To track this we need to subtract the transactions done for speculation or money laundering. This data is surprisingly hard to find. One data point on that is helpful is Number of Transactions Excluding Popular Addresses.

This does show some reasonable growth and excludes speculative bursts around events such as Brexit.

Is there a better data point to track? I imagine that big payment processors such as Coinbase, Bitpay and Circle have this data, but do they put this in the public domain?

Indicator 3 = Cost Per Transaction. It ain’t free, but we hope it is cheaper than credit cards. Entrepreneurs will markup cost to provide valuable services, but what is the cost? shows this. The data problem is that this chart is in USD, so the exchange price gets in the way. What we need is cost per transaction in bitcoin (just like we use transaction volume in bitcoin). If anybody knows where to find this, please tell me. Maybe one could compute this from a mix of things like Hash Rate and Difficulty. However, that is a nuance. The big picture is that small transactions are not economic today. Which means some combination of offchain centralized processing and/or use of Sidechains. This is another area where Segregated Witness and Lightning Network rolling out in 2017 will have a big impact.

Indicator 4 = Transaction time. Is this approaching the “human real time” (a few seconds) that consumers expect from digital services? Or will we have to trust centralized intermediaries to make it look like human real time (rather ruining the core proposition of bitcoin)? The Average Transaction Confirmation Time shows this. There is nothing dramatic about this chart, it looks stable as one would expect unless there had been a significant change to the protocol. The problem is simply the numerator, which is in minutes.

This is where we expect to see a lot of change in 2017 as Segregated Witness and Lightning Network roll out. For background on these scalability issues please read this post.

Why 4 indicators are needed.If your doctor only ran one test, they would be negligent. The bitcoin ecosystem is complex and therefore also requires a number of tests.

Tomaso Aste of University College London presented a paper at the P2P Financial Systems Conference last Sept which estimated a $5 transaction cost for bitcoin.

Presentation can be found here.

1 Like

Statistics from Indonesia on the number of users and transaction volume:
The report from the Jakarta Post states that the number of bitcoin users in Indonesia has risen from 80,000 up to 250,000, with a daily transaction value amounting to Rp 20 billion (US$1.48 million).

Thanks @Efi. If that data is accurate it represents about 0.1% of GDP as per my back of envelope:

Daily = US$1.48 million.
Annual = *365 = 540m = 0.5bn
GDP = c 870bn
So Bitcoin = less than 0.1%

At about 1% I guess we can start talking about this going mainstream.

Thanks @Dianacbiggs and welcome to Fintech Genome. I guess we really need Segregated Witness and Lightning Network to bring that $5 cost down.

From Mexico’s bitcoin exchange I read that:
Bitso, saw more trading activity in one month during December than in 2014 and 2015 combined, amid the end of the year price rally.

But why arent the Bitcoin exchanges (or their incestors) experiencing increased trading, reporting statistics?

Also from my sampling of bitcoin bid/ask spreads for my personal trading, I dont observe any tightening of spreads (which relates to cost per transaction) during this rally.
Have bitcoin margin levels changed?

The Chinese bitcoin exchanges will be charging a 0.2% fee.

@Efi if Chinese exchanges add trading fees we will see a drop in trading but the numbers will be more reflective of real world.

Trading volumes did drop but trading became more evenly distributed across exchanges.