tpeng

People ITT are not making a distinction between different types of HFT. Katsuyama and Lewis do not criticize all forms of computerized trading, but specifically computerized scalping that is aided and abetted by exchanges. Their chief complaint is that HFT uses more current prices than the exchanges themselves and they use this to scalp other traders.

http://www.zerohedge.com/news/2014-04-03/bats-admits-ceo-lie...

This is actually very simple. Natural buyers and sellers do not need intermediaries, but intermediaries do need the natural traders. So if the natural traders can coordinate, they should be able to set rules that favor themselves and disfavor intermediaries. I won't say that what HFT does is "unfair" (capitalism does not contemplate fairness), but I think it's highly ironic that HFT and their supporters are complaining how "unfair" it is that natural traders are working together, and yes, marketing their new exchange.

Bitcoin doesn't eliminate security risks, it pushes the costs onto consumers. It's inconceivable to me that regular users can be trusted to maintain and secure their own wallets, so the only solution is an online wallet. But those businesses will need to be paid for their service, and face security risks as well, which under the current system are ultimately transferred to users. For bitcoin to work, we need a system that guarantees that consumers will not bear losses from fraud or hacking.

There's a big difference between the dot-com IPOs of the late 90s and the IPOs of today: revenues. Some may express surprise that Candy Crush generated $1 billion of revenue, or that Facebook generates $8 billion / yr; and some may not agree with the multiples assigned to these companies by the markets; but neither of those thoughts makes the current environment comparable to 1999, when pre-revenue companies with unworkable business plans IPO'ed at high valuations. What I don't see in this article is any analysis of the growth potential, long-term margins, and defensibility that these businesses possess. That kind of analysis is crucial to a valuation discussion.

My thoughts on the BAML report:

The calculation of BTC's value as a currency uses the correct methodology, which I had previously outlined in this post: https://news.ycombinator.com/item?id=6829518

BAML uses M1 (cash + checking accounts) instead of M2 (M1 + savings accounts + time deposits). Using M2 is a lot more aggressive but it also captures the "store of value" value. M2 is really a best case scenario for BTC owners, as it assumes that users will hold their savings in BTC. BAML is fine to be conservative with this.

BAML's inclusion of Bitcoin's value as a money transmission business is absurd. While transfer fees accrue to the shareholders of Western Union, the transfer fees of BTC transfers do not accrue to BTC owners. In fact, the transfer fees are quite small to zero and they accrue to the miners. So if BTC does take share from WU, etc., it only shrinks the pie.

Some posters on HN and elsewhere are saying that BAML is "too conservative", but they have not posted their own numbers and assumptions.

The reality is that BTC's current real economy is nowhere near the size needed to justify the current BTC price, and enormous growth of this economy is already baked into the price. While it's not impossible that this growth can occur, there are real reasons to believe that it won't.

BTC value is directly related to its velocity and the value of the goods and services that are transacted in BTC. I held the purchasing power of USD constant, simply to derive a value of BTC in 2013 dollars. If USD experiences a large shift in PP (unlikely), that would affect BTC/USD. But either way that's not an inherent source of value to BTC.

This is my last comment, as I cannot possibly respond to an endless string of unfounded criticisms.

We could debate what the velocity of bitcoin is/will be, but the relationship between velocity, money supply, and transaction volume is definitional.

http://en.wikipedia.org/wiki/Velocity_of_money

Price-implied expectations

http://www.amazon.com/Expectations-Investing-Reading-Prices-...

There's no obvious reason why BTC velocity should go lower than USD velocity. If anything, its electronic nature and potential for fast settlement times should indicate a much higher velocity than USD M2.

A higher money velocity => lower BTC/USD, but yes, the current price implies a huge future growth of the bitcoin economy. This is what sam is saying and I agree. I happen to think that we will not see that growth/adoption, but that is a different question.

The way I value bitcoin is to take the fully diluted market cap (21 million * $850 ~= $18 billion) and apply a money velocity which I assumed to be similar to US M2 velocity of 1.6 ($18 billion * 1.6 ~= $28.5 billion). Since global ecommerce transactions are about $1 trillion / yr this implies that bitcoin should penetrate 2.85% of ecommerce. Big assumption here is that bitcoin does not work for real-world transactions as the confirmation time is > 15 min. This may be oversimplified because a lot of real-world transactions don't need to be real time. I'm also excluding time value effects which you can insert back in if you like.

From the data I have seen, the bitcoin real economy could be orders of magnitude smaller than that implied by the current price, although there is no way to know precisely how large it is. This is including "illegal" transactions (I'm not sure why sam is discounting drug transactions, unless he means to imply that these will be shutdown. It does make sense to exclude gambling transactions as they are extremely high velocity).

Note that sam's point about merchants immediately converting BTC back to USD serves to increase the money velocity (perhaps by >10x). This would mean that BTC is pricing in a real economy even larger than $28.5 billion. Based on current BTC-denominated real transactions, fair value of BTC is at most only a few dollars.