TLDR
In order to significantly improve the efficiency of the IMX token, we propose a removal of the token rewards lockup period. Gauntlet’s insights are all results of our modeling and analysis, which show strong indications that these changes can drive positive expected value. We recommend a removal for the following reasons:
- More tokens than required are being spent. The current 6 month unlock causes traders to discount their rewards by up to 66.6% leading to an inefficient use of tokens.
- Lockups do not influence long term token holding behavior. There is no data to suggest that the current 6-month lockup leads to long term token holders.
- Removing the token-unlock and reducing the rewards allocated will not impact the token negatively. Removing the lockup is unlikely to cause a a strong impact impact to IMX. Removing the lockup will however slow the longer term rate of inflation as token emissions can be reduced.
Intro/Abstract
Earlier this year, Gauntlet announced its partnership with ImmutableX with a shared goal of optimizing the ongoing Trading Rewards Program to maximize efficiency and impact. The trading rewards program was established as a strategy to bootstrap users and games to onboard onto the ImmutableX ecosystem. “At Immutable, empowering our community through true ownership is a top priority. ImmutableX is recognized as the leading platform for launching web3 games and trading digital assets, and you deserve to share in that success.” Setting aside up to 100,000 IMX tokens a day is a major investment in the community. Through our work with major lending and exchange protocols, we at Gauntlet are committed to ensuring our partners are equipped with data-backed insights to achieve their goals.
Furthermore, in the hyperspeed realm of crypto, this requires staying highly nimble. Protocol strategies must constantly be reevaluated as data provides new and informative lessons on how to improve. In this post, you will see Gauntlet’s modeling of the effects of the token lockup mechanism. Currently, one-third of the rewards are paid out immediately to traders at the end of the trading week, while the remaining two-thirds are unlocked linearly over the next 6 months. One of Gauntlet’s top priorities is to investigate whether the lockup effects are beneficial for the rewards program. Gauntlet will battle test current assumptions and make our recommendations for the community,
The Increased Cost of Token Lockups
One main goal of Gauntlet’s partnership is to increase the efficiency and effectiveness of the rewards program to drive growth to the protocol. The IMX Rewards program should be measured by its ability to encourage new and existing users towards increased participation while distributing rewards efficiently.
A rational trader has two things to consider when deciding whether or not to trade on ImmutableX versus another exchange due to the trading rewards program: cost and revenue.
On the cost side, a trader must pay the fixed fees included in the IMX platform. On the revenue side, the equation becomes more difficult. Traders must discount their revenue against the risks of holding IMX, which may move against them before their tokens unlock. Long unlock periods, coupled with IMX performance volatility, results in traders significantly discounting the value of their earned IMX tokens against the value without an unlock period.
Gauntlet has empirically evaluated the level to which traders are currently discounting their locked-up revenue. This discounting results in IMX spending excess tokens relative to what they could have spent without employing a lockup period. By removing the lockup, our modeling of ImmutableX predicts two paths with positive expected value:
- Spend less and achieve the same results
- Spend the same amount and achieve superior results than the program so far.
Simulation Analysis
To analyze the effects of the reward lockup, Gauntlet used a geometric Brownian motion process to simulate various paths of the IMX token parameterized with recently estimated volatility and zero drift. (Volatility was calculated as the close-to-close method based on the last 60 days of the period).
As the plot above shows, the value of daily IMX trading rewards is subject to high volatility and can make it difficult for traders to estimate their returns accurately over a 6 month period.
Figure 2 shows a simulated distribution of changes in IMX reward value measured in USD from time of trade versus time of token unlock (taking into account the monthly unlocks). Although the distribution shown above presents a heavy right tail, around 61% of the results show a negative change in value. As expected, the mean of the distribution is 0 since we parameterized the GBM with 0 drift. The main takeaway of the distribution is that it shows that the value of the trading rewards can vary quite drastically by the end of the lockup posing an increased risk to traders looking to collect rewards. Thus, we believe this high variability in rewards value leads traders to discount the rewards aggressively in order to offset the downside risk. This discount is likely to increase with both longer lockup times and higher volatility.
Empirical Insights
Another way we can analyze trader discounts is to compare the amount of fees traders are willing to pay versus the value of the rewards at time of trade. Consider a trader that is looking to make a profit from the trading rewards program. For this trader, their return (defined as ratio of reward to fees minus 1) should be greater than 0.
1-DiscountRewardUSDFeesUSD-1> 0
Because rewards are distributed proportionally based on volume, we can assume that profit seeking traders on a given day will continue to trade (and pay fees) up until the point where the above condition is no longer satisfied. Therefore, by using the equation above and looking at historical ratios of rewards and fees, we can estimate the discount that traders are applying to the rewards.
Discount = 1-FeesUSDRewardUSD
Looking at data from recent months, the estimates show that traders have discounted IMX rewards anywhere between 55%-80%. Since these numbers seem to be centered around ~66.6%, this might suggest that a common strategy amongst traders is to trade up to the point where the fees can be covered by the first 33.3% rewards which are unlocked in the same week, and then all future unlocked tokens can be considered as profit. From a rational trader perspective, this would mean that the most aggressive discount should be at 66.6% given current unlock mechanics. Traders appear to be taking a low risk approach and covering themselves from the downside risk of fees ever being higher than rewards.
Based on this insight, we can visually compare the immediate unlock schedule against the schedule of rewards under various assumptions where traders are discounting at 55%, 66.6%, and also in the worst case where traders don’t discount at all.
Based on this evidence, Gauntlet’s analysis shows a positive expected impact where removing the lockup mechanism from trading rewards can allow traders to no longer discount their returns and increase rewards efficiency by up to 66.6%.
Lockups do not provide value to ImmutableX
At the time of developing the incentives program, the theory behind creating a lock up period made sense. There are two primary goals that the Foundation was originally trying to achieve with their lockup periods:
- Lockups can reduce the likelihood of IMX holders to sell the tokens
- Lockups can prevent substantial impact by limiting circulating supply
Gauntlet has taken a quantitative look at all of these points and determined that none have empirically generated real value to the Foudation and Protocol, especially in relation to the high cost of increased spending that lockups cause.
Lockups do not result in IMX holders that are less likely to sell
One possible argument in favor of lockups may be that traders are less likely to sell the IMX tokens on the market after the unlock is removed. One could argue that traders are using the incentives program as a cheaper means to acquire IMX tokens to hold. The lockup periods may encourage emissions only to traders who will continue to hold the tokens after the lockup period ends. Unfortunately, this argument doesn’t seem strong empirically.
Gauntlet looked at the IMX token withdrawals from ImmutableX relative to the payments of unlocked tokens.
We see a clear correlation between withdrawals and unlock periods where withdrawals clearly outpace unlocked payments. Rather than staking tokens on the ImmutableX platform, traders tend to withdraw those tokens to sell on the open market.
Given that tokens are being sold at the same rate they’re being unlocked it is unlikely that traders with a lockup are more likely to hold versus traders without.
Removing the lockup is unlikely to create substantial impact
The greatest concern about removing the lockup period is that the lockups are “protecting” the IMX performance. The idea here is that if tokens that are paid out are locked up, they won’t be added to the general circulation and thus won’t impact the performance until unlocked.
In general, we’d expect the inflationary nature of increasing the IMX token circulation to be already considered. Traders already know the max supply cap and the schedule that tokens will be released as outlined in the ImmutableX Whitepaper and can consider this accordingly.
The fear of unlocking tokens too quickly is not in reducing the expectations of inflation, but rather in preventing unlocking too many tokens at once. Unlocking too many tokens may cause an impact against available liquidity in a given day. One could imagine a scenario where the Foundation releases too many tokens at once, and crashes the performance for that day due to too little liquidity. This crash would likely be quickly corrected by traders buying back discounted IMX tokens; however, it may be possible that the crash could reduce confidence in IMX's long-term prospects, impacting the token’s performance over a longer time horizon.
To assess this risk, Gauntlet looked at the ratio of historical IMX payouts to the overall daily trading volume on exchanges and compared it to the counterfactual if the tokens had been paid out without a lockup.
In the worst-case scenario, we see only ~5% of daily volume in tokens may be paid out without a lockup and the same incentive schedule, a relatively small amount that the market should be capable of absorbing. In reality, since removing the lockup should allow the Foundation to lower their token spend up to 66.6%. The actual emissions payout scenario will likely be much more optimistic.
Gauntlet looked at the statistical relationship between historical token unlocks and token withdrawals on IMX performance for that day to ensure that this amount of emissions won't result in significant daily impact.
For token withdrawals (a proxy for selling) we see no statistically significant relationship between withdrawals and changes in IMX performance for that day, even despite some outliers at almost 10% of daily trading volume.
The situation looks similar for payouts themselves where there is no statistically significant relationship between unlocks/payouts and IMX changes.
Empirically the fear that IMX rewards payouts without a lockup will cause any significant damage to the IMX performance is overblown. The market is either able to absorb the sale of IMX tokens or more realistically traders are rationally selling out of their IMX position over a period of time to avoid performance impact.
What is damaging however is the inefficient spend of IMX. The Foundation can reduce inflation by spending less IMX or spending it more effectively without the lockup. The inflation is already considered, but the spend efficiency is not. If the goal is to encourage sustainable volume and token longevity, removing the unlock is one of the most efficient ways to accomplish this.
Removing the lockup is unlikely to cause a strong impact to IMX. Removing the lockup will however slow the longer term rate of inflation as token emissions can be reduced.
Conclusion and Next Steps:
Gauntlet recommends removing the lockup to benefit the Foundation and Protocol’s long term, sustainable growth. Inefficient rewards spending can be effective for bootstrapping new ecosystems, but the Foundation is now at a scale where it should consider operating differently to achieve its next growth chapter. While lockups appear beneficial to growth at a topical level, Gauntlet’s analysis indicates that it has not been an effective lever with regards to token holder growth nor reduced sell pressure. As such, our analysis indicates that there are negative effects to IMX token action from the current construct.
Gauntlet will put this to a snapshot vote and welcome any feedback or questions from the community.