Tuesday, February 25, 2014

The Promise and Peril of Bitcoin

Economist Susan Athey and venture capitalist Balaji Srinivasan discuss the digital currency.

Bitcoin is a stateless, digital currency that allows people to make transactions with no middlemen — no banks, no transaction fees and no governments. Instead of printing banknotes or minting coins, a list of the registration numbers of each of the "bitcoins" and their owners is kept. In the last week, however, Bitcoin has found itself under siege for a variety of problems, and since then, the price of the coins has tumbled.

Susan Athey is an economics professor at Stanford Graduate School of Business who researches Internet economics. Venture capitalist Balaji Srinivasan founded the Stanford Bitcoin Group. He has a Stanford doctoral degree in electrical engineering.

They spoke with Stanford's Clifton B. Parker about Bitcoin:

Is the Bitcoin protocol trustworthy?

Athey: In terms of trustworthiness, a large community of experts has examined the protocol closely and found it to be secure. Possible weaknesses have been discussed extensively in the Bitcoin community as well as by academics. No protocol can be guaranteed to be perfect, of course.

In recent days, some issues have arisen with a denial-of-service attack that prevented transactions from confirming properly. MtGox, a major Bitcoin exchange, suspended withdrawals for a few days to deal with the issue. This is a reminder that all digital systems have some vulnerabilities.

Given the diminishing role of physical cash in our economy, we can expect problems to arise in all of the different financial systems we have.

My personal view is that it is important for regulators to provide a framework for legitimate Bitcoin businesses to operate within the U.S. legal system … legitimate firms will have the resources and incentive to provide more sophisticated consumer protection and to develop and adopt best practices for security.

Are we ready for Bitcoin on a large scale?

Srinivasan: If the Internet was programmable communication, Bitcoin is programmable money. The analogy is actually a very good one. Before the Internet, in order to deploy a program that used the central communications network, you needed a deal with a network operator like AT&T. With the Internet, individual nodes could write programs to directly communicate with each other, without any central approval required.

In the same way, today, in order to deploy a program that uses the central financial network, you need a deal with a large bank or credit card company. With Bitcoin, individual nodes can write programs to directly send and receive money from each other, without any central approval required.

There are technical improvements that will need to be made over the next few years to allow Bitcoin to become usable on a large scale. This is like the early Internet, which went from dial-up modems to broadband wireless over two decades.

What are the strengths of the Bitcoin system?

Athey: The Bitcoin system has at its core a secure public ledger. An entry on the ledger is not an IOU from a bank, like your bank account balance is. Rather, the ledger entry is the definition of ownership of a bitcoin. Within the Bitcoin protocol, one individual can send bitcoins to another without any middlemen or risk … that is a very powerful concept.

Srinivasan: The strength of Bitcoin is that it is decentralized programmable money. It is like having three APIs [application programming interfaces]: You can check balances, write contracts and make transfers simply by tapping keys on your computer. No supernodes with special permissions like banks are needed to manage these transfers – anyone can be a so-called "miner" and help with the distributed transaction approval process. Thus, if the Internet enabled permissionless innovation, Bitcoin allows permissionless monetization.

What are its weaknesses?

Srinivasan: It is a young protocol. In the fullness of time it will be something like HTTP [Hypertext Transfer Protocol] – there will be apps, hardware and the whole nine yards built on top of it. Right now there are many Bitcoin improvement proposals, which will give you a sense of where the community is focusing development efforts.

Athey: The exchange rate is too volatile now for it to make sense for the average individual to hold bitcoins, except as a speculative investment. This limits its usefulness somewhat.

What is interesting, however, is that the level of the bitcoin exchange rate is not important for Bitcoin to be useful as a medium of transaction. Suppose that an individual wishes to use Bitcoin to transfer value. She can buy bitcoins and send them, and the recipient can cash them out, and for this purpose it does not matter whether the exchange rate is $1,000 per bitcoin or $100 per bitcoin.

If the individuals already have accounts on Bitcoin exchanges and some funds there in their local currency – for example, if they are verified users of a service like CoinBase – they can execute the whole transaction in about 20 minutes. As long as the bitcoin exchange rate does not fluctuate much in that time interval, the dollar value does not matter. It is an efficient mode of transaction.

What about the Bitcoin software?

Athey: The protocol does have some weaknesses like mining. [Mining is the process of adding transaction records to Bitcoin's public ledger of past transactions.] Mining was a very clever way to solve multiple problems at once – it helps distribute the currency in an orderly fashion and provides an incentive for individuals to contribute computing power to maintaining the ledger. However, mining wastes a lot of energy and creates complications as well.

Another weakness is that transactions take 10 minutes or more to confirm.

The fact that Bitcoin is open source brings all of the advantages and disadvantages of open source software … there is no need to rely on a company to maintain it. On the other hand, there is no central entity with which a government regulator can exercise control.

Will Bitcoin grow in popularity in the future?

Athey: I think a useful analogy here is PayPal. PayPal has lower transaction fees than credit cards, and for transactions between PayPal customers who fund using bank accounts or keep balances in their PayPal accounts, the transactions costs are low. PayPal enabled a lot of new use cases for electronic commerce. Small businesses priced out of credit cards were finally able to participate in electronic commerce.

Bitcoin and other virtual currencies can take all of this one step further. International payments for low-value transactions become viable … workers from poor countries who take domestic jobs in rich countries and then send funds home currently pay high fees, especially for small or urgent transactions.

How can the Bitcoin system avoid counterfeiting, pirating or criminal use?

Srinivasan: The fundamental problem that Bitcoin solves is the so-called double-spending problem, intimately tied to counterfeiting. To understand it, let's go through four scenarios.

  1. In a physical cash system, when "A" gives "B" a physical dollar bill, "B" knows that "A" no longer has possession of said bill.
  2. By contrast, in a naive implementation of a digital cash system, if "A" simply emails the serial numbers on that dollar bill to "B," then "A" still has a copy of the serial numbers as well, and "B" knows this. If these serial numbers are being treated as "digital cash," there is a temptation for "A" to spend them again with party "C." Hence the double-spending problem.
  3. The traditional centralized solution to the double-spending problem is for a special node, like a bank, to act as an intermediary. Now when "A" sends money to "B" over the Internet, that central node debits "A" and credits "B," recording these debits and credits in a private ledger.
  4. The decentralized solution, Bitcoin, is to do away with special nodes and instead allow every node in the network to maintain a copy of the ledger of debits and credits. This ledger is called the Bitcoin blockchain.

Athey: In the early days of the Internet, one of the few industries to make money in e-commerce was pornography. In fact, pornography was a large share of Internet activity then. As the Internet matured, pornography did not go away, but the fact that the Internet facilitated pornography did not lead the government to slow it down. Today, the Internet is still used for crime, but law enforcement can catch criminals through their digital trails … if enough is at stake with bitcoins, one can use the digital trails people leave behind to associate addresses with individuals.

One activity in the Bitcoin community is the development of ways to associate identities with Bitcoin addresses. Such a feature might make it easier for businesses to comply with regulations.

What needs to be improved?

Srinivasan: The Bitcoin improvement proposals and the open source repository issues are good resources to see what needs to be improved. Most improvements would be made via code fixes that would be adopted by clients and/or miners in the Bitcoin community, much like people adopt new web servers or open source code.