Bitcoin Payment Processing after a Fork

Bitaccess’s software powers bitcoin services in over 10 countries, through Bitcoin ATM’s, retail purchases, compliance and wallets. In serving such a diverse customer base, we get a rather unique look at the Bitcoin market. It is becoming increasingly likely we will see a contentious fork of the Satoshi Blockchain (a.k.a. Bitcoin) this year; lets take a closer look at what this will mean for Bitcoin payments.

What’s Payment Processing?

When it comes to Bitcoin payment processing, here are the three main categories:

  1. Selling Bitcoins to end users (ATMs, Bitaccess, ShapeShift)
  2. Letting end users pay for services with Bitcoins (Bitpay, ATMs, Purse.io)
  3. Selling Bitcoin IOU’s to users (Ex. Coinbase, LocalBitcoins, Xapo, Exchanges)

Selling (#1) and receiving (#2) bitcoins requires a functioning blockchain to deliver excellent customer service to customers. Bitcoin IOU’s (#3) generally occurs in central databases, not on the blockchain, and therefore is less susceptible to hard forks or network events. 

Does Bitaccess support a specific fork?

Bitaccess is a customer focused company, and therefore take a neutral stance to the proposed bitcoin clients. We have many customers who support Bitcoin Core, and others who support Bitcoin Unlimited. We would prefer to not have to choose. If a fork does occur, we will continue to support all our customers to the best of our ability. We aim to make our customers happy.

What would a Bitcoin fork look like?

The potential fork of the Bitcoin network would theoretically occur when Bitcoin Unlimited reaches 80% of the overall network hash-rate. We did a short back-of-the-napkin simulation of what that would look like from a transaction perspective.

The details of this simulation are available here. For the technically minded, here is a short list of the assumptions (skip this if you don’t like details):

  • The Bitcoin Unlimited fork would occur shortly before/after a difficulty re-adjustment to cause maximum damage to the minority chain.
  • The hash-rate is split 80% towards Bitcoin Unlimited, and 20% towards bitcoin core, which remains flat over the period.
  • There is an ongoing, flat demand of ~1963 transactions every 10 minutes to be added to the network. This is the current 30 day rolling average number of transactions in block in the current bitcoin network. This ignores any current transaction backlog or transactions currently unconfirmed. Average Transactions per block over the past 2 years
  • Average Transactions per block over the past 2 yearsCurrent network hash-rate of(3616339074GH/s), and estimated upcoming difficulty (498250730553)
  • Not considered: reduced demand of bitcoin transactions, split of transaction demand across chains, potential attacks on minority chain, minority chain difficulty/PoW fork, sudden segwit activation, etc.
  • No consideration of the resulting centralization/decentralization of each network, legal aspects surrounding this

What would happen to Bitcoin payments after the fork?

The immediate effects after the fork would be quite painful for those using the minority chain (Bitcoin core). First, block confirmations would immediately start to take approximately 49.3 minutes.

Services that send bitcoins to users need to have confirmed bitcoins ready to deliver to end users. If it suddenly takes 5 times longer to confirm transactions, Bitcoin brokerages would need about ~5x more bitcoins in their Hot wallets to continue to serve the same number of customers.

Services that accept Bitcoins from customers would be in worse shape. These services would have to wait extended periods of time to deliver products to their customers (assuming they wait for at least 1 block confirmation). This would be exasperated by the fact that there would be significantly higher fees required to have a transaction included in a block, which end users have been historically bad at setting correctly. 

The majority chain (Bitcoin unlimited) would see 12.3 minute blocks immediately after the fork. Assuming the block size is increased to just 2MB, there would be no predicted transaction backlog based on regular activity. That being said, it is unclear how many customers would actually the majority chain (Bitcoin Unlimited).

Difficulty Adjustment

The bitcoin blockchain automatically adjusts its difficulty every 2016 blocks. If there are no significant changes in hash-rate, this means every 14 days. Given the split of hash-rate, the majority chain (Bitcoin unlimited) would see its next adjustment in 17.25 days. At that point, blocks would take 10 minutes again.

Given the slowed speed of block generation on the minority chain (Bitcoin core), it would take approximately **69 days(( for the difficulty to adjust back to normal.

During that time, a transaction backlog would start to build up. If we assume that the demand for bitcoin transactions remains constant, every day, we would see approximately ~220,000 bitcoin transactions go unconfirmed. These are transactions that would have been confirmed if the on the main chain if there was not a fork.

By the time the minority chain (Bitcoin core) finally re-adjusts, there would be more than 1.5 Million transactions in this backlog. In reality, we can assume that a number of users would abandon the chain, in essence creating a period of around 69 days of lost economic activity. 

Most start-ups can’t sustain 69 days of flat growth, let alone shrinking. It is debatable that the minority chain would be able to hold onto a significant number of economically relevant nodes during this time without immediate corrective action.

Lost Economic Activity: Changes in Behavior

Fundamentally, if a fork occurs, both blockchain’s would continue to function, and there would be a shift in customer behavior. The major affect for bitcoin payment processing will be the immediate loss of economic activity.

This means that Bitcoin ATM’s would slow down worldwide, being unable to accept bitcoins from customers, and unable to reliably send bitcoins to customers. Payment processors such as Bitpay would also have significantly increased processing times, and an increase in customers who’s transactions never confirm. 

The only option that Bitcoin payment processing companies would have to continue to the current demand of on-chain transactions reliably would be to start issuing Bitcoin Core IOU’s to customers, or to use a different blockchain.

Should we fork?

We at Bitaccess are prepared to offer our customers reliable, stable service no matter what happens. It will be a difficult time for everyone, but Bitcoin has been through tough times in the past and has emerged stronger as a result. If a fork is what is needed for the community to achieve a diverging set of goals, it is best it happen now, rather than in the future when we have even more customers to keep happy. 

We are going to lose customers during any fork, lets keep it to a minimum.