Recently, there have been many discussions regarding P-Rep reward sharing in various ICON discussion channels. Specifically, ICONists are concerned about P-Rep teams offering high reward shares that augment the protocol-level rewards, in exchange for votes. In this post, we’ll discuss why P-Rep reward sharing and voter bribery are detrimental to the ICON ecosystem. Before we discuss additional rewards, let’s quickly go over ICON’s protocol-level reward structure. In exchange for staking and delegating ICX, ICONists can receive a 2-12% reward per year per category.

The reward percentage per category depends on the percentage of total ICX staked. There are three categories to stake to (Reps, DApps, and EEPs), which means ICONists can expect a reward between 6-36% per year. ICX rewards come from two sources – transaction fees and newly minted ICX. Both of these sources are derived from P-Rep productivity. P-Reps validate transactions and produce blocks, and ICONists will not receive rewards without the productivity of P-Reps. Thus, rewards in the ICON ecosystem are essentially derived from the efforts and productivity of P-Reps without any reward sharing model needed.

Two Models of Reward Sharing

Over the past few months, we’ve seen two models of P-Rep reward sharing emerge from various candidates.

  1. Sharing X% of profits in exchange for an extended lockup period.
  2. Sharing X% of representative rewards without any sort of lockup period.

Model #1 - Altruistic Reward Sharing

This model is altruistic in nature, in the sense that it is beneficial for all parties involved – P-Rep, voter, and the ICON network. Unlike Model #2, which doesn’t take P-Rep expenses into consideration, Model #1 only gives away X% of profits. For example if a P-Rep generates $20,000/month in rewards, has $12,5000 in expenses, and chooses to share 50% of its profits, then voters will receive their share of $3,750/month after the lockup period. In this model, a P-Rep is able to prioritize node stability along with its specific focus – development, marketing, content creation, etc. There is no pressure to cut costs in order to appease voters because this model shares X% of profits only. This model also benefits the voter because it generates a slightly higher annual return rate. Lastly, the ICON network also benefits from this reward model because it encourages long term staking of ICX, which in turn creates a more stable network. Remember, companies and enterprises crave stability.

Model #2 - Destructive Reward Sharing

This model only benefits the P-Rep and the voter, and has the potential to be extremely destructive towards the ICON network. The stability and security of the ICON network plays a huge role in ICX valuation. Thus, it doesn’t make sense to vote for a P-Rep whose reward sharing model has the potential to be destructive towards the ICON network. Before we get into why this model is destructive overall, let’s go over why this model is beneficial for the P-Rep and the voter. Greed is the driving force behind this model, and a P-Rep who utilizes this reward model to garner votes is effectively engaging in bribery – paying for votes with no additional beneficial commitment to the ICON network. In this model, the voter benefits from a much higher reward rate because the ICX used for rewards is sourced from revenue instead of profits.

For example, if a P-Rep generates $20,000/month in rewards and shares 80% of its revenue, then voters will receive their share of $16,000/month. This model is destructive because it incentivizes a P-Rep to cut corners when it comes to node operation and the overall quality of contribution to the ICON network. Since the stability of the ICON network is derived from optimal node operation by P-Reps, a P-Rep that cuts corners to appease voters via high reward schemes is a detriment to the ICON ecosystem. This model completely falls apart when you realize it’s a percent-based model applied to a hyper-volatile asset. When it comes to node operation, certain costs are fixed and known. For example, an AWS server with 32 CPUs and 72 GB of RAM (ICON’s recommended specification) is always going to cost approximately $1,300/month. Similarly, unless the P-Rep team in question is planning to work on a volunteer basis, there will be HR costs as well.

Sample Monthly Cost Breakdown

  • P-Rep Node (Primary) - $1,300
  • P-Rep Node (Backup) - $800
  • Citizen Nodes - $400
  • HR Costs (3 Team Members) - $9,000
  • Business Operations - $1,000
  • Total - $12,500

With the above cost breakdown, which is completely reasonable, a P-Rep offering an 80% reward share would have to generate $64,500/month in rewards in order to break even. ICON is not a money printing machine, and a P-Rep with an 80% reward share will be incentivized to submit a high I_rep to stay afloat. In the end, a high reward share model reduces the security and stability of the ICON network, increases ICX sell pressure, and has the potential to direct a lot of value out of the ICON ecosystem.

Reward Sharing & Behavioral Encouragement

Let’s dig deeper into why Model #2 is so destructive for the ICON ecosystem by discussing the behaviors that it encourages. A P-Rep that prioritizes itself before the stability of the ICON network by bribing voters with a high reward share attracts a few different demographics.

  1. ICONists who want the highest reward rate possible to compound their ICX stake at a faster rate.
  2. ICONists who do not understand the negative impact of a high reward share model, and how it has the potential to kill the network.
  3. Voters who don’t care about ICON, and want the highest amount of tokens to dump as soon as possible.

Demographic #1

In theory, this demographic keeps value within the ICON ecosystem through re-staking. The implications of the actions by this demographic can go both ways depending on what P-Rep the rewards are re-staked to. If it is re-staked to a P-Rep giving away 80% of its rewards, the negative feedback loop will strengthen. On the contrary, re-staking to P-Reps with strong technology, marketing, and business development roadmaps could strengthen the project as a whole. For Demographic #1, I am in favor of the re-staking to other P-Reps because it is an effective method of siphoning value from high reward share model, and injecting it into more productive endeavors.

Demographic #2

This demographic is comprised of ICONists who care about the ICON network, but may not understand the negative economic implications of a high reward rate model. Over the last few weeks, I’ve observed a number of discussions about P-Rep reward sharing on ICON channels, so it’s nice to see community members educating each other about the destructive nature of reward sharing with no strings attached.

Demographic #3

This demographic is the one we should worry about. Participants in this category have one thing in mind – maximizing short term gains. At the protocol level, ICON is already an attractive network because it offers an annual reward rate between 6 and 36% depending on the percentage of network staked. 6-36% per year is a great return, especially when you consider ICON’s fundamentals and enterprise adoption – ICX is not a safe investment, but it’s probably a lot safer than a lot of crypto assets out there. An individual in this demographic is looking to maximize returns on a fundamentally sound token. Thus, he or she will most likely vote for the P-Rep who’s offering the highest reward share. Since this person has no commitment to the ICON network, he or she will likely dump ICX rewards in order to buy other cryptos or material goods. In addition to extracting value out of the ICON’s reward model in an unproductive way, this demographic also places unsubsidized sell pressure on the ICX token. When P-Reps sell ICX to fund DApp development, content creation, business development, and more, the act of selling is subsidized by the potential for future price appreciation of ICX, which may eventually offset the initial sell order in terms of absolute USD value. Conversely, an individual selling ICX sourced from a P-Rep offering a high reward share puts unsubsidized sell pressure on the ICX token – which is bad for both current token value and future token value.

Reward Sharing is an Opportunity Cost

At the end of the day, reward sharing is an opportunity cost. A P-Rep that participates in voter bribery by giving away 80% of rewards is effectively destroying all the potential opportunities that could’ve emerged from directing funding to technological development, market research, onboarding enterprises, and more. In summary, here are the economic implications of no high reward share model versus a high reward share model.

No High Reward Share Model

  • A P-Rep can direct 100% of its rewards straight into endeavors that can have a positive effect on ICX price in the longterm. This is extremely important over the next 2-3 years because the ICON network is still in its infancy. As ICONists, we need to support P-Reps who are committed to giving an ICON a first mover’s advantage.
  • There is no need to inflate I_rep submissions in order for high reward share P-Reps to maintain a profit margin.
  • All parties win in this model – P-Reps get adequate funding, ICONists still receive great staking rewards, and the ICON network enjoys stability and security thanks to adequate node infrastructure.
  • On a side note, I believe reward share models sourced from profits (not revenue) that enforce lockup periods are beneficial to the ICON network because a P-Rep can still direct the necessary funding to productive endeavors while encouraging voters to secure the ICON network.

High Reward Share Model

  • A P-Rep must cut costs in order to deliver on its 80% reward share promise. As demonstrated above, this model doesn’t make much sense when considering the absolute and fixed costs of running a stable and redundant node.
  • A P-Rep is incentivized to inflate its I_rep submission in order to break even or maintain a profit margin.
  • The P-Rep and its voters “win” in this model – both parties receive a larger absolute value of ICX, but this has a negative effect on the ICON network as a whole because value is not directed to the proper places (development, business development, marketing, etc.).

As an ICONist, ask yourself one simple question before casting your vote. Do I vote for a P-Rep who will utilize rewards to develop tools, educate the community, and onboard businesses onto ICON, or do I vote for a P-Rep whose biggest call to action is giving away 80% of its rewards to parties that aren’t committed to furthering the ICON ecosystem? Vote wisely.