The HOPR Sale on LBP: Analytics and Reflections
The HOPR public token launch was a blast! For a while, the HOPR token was even ranked as one of the top 20 tokens by trading volume. But what else happened during the public sale? This post will provide a recap, along with some of the most interesting facts and figures.
It’s easy to forget just how much of a leap into the unknown the HOPR launch was. By allowing the DAO to determine the distribution method for the token launch, we at the HOPR Association had no control over anything beyond the number of tokens available.
Even after the DAO vote, there were still a lot of open questions: How many people would participate in the presale? Would people understand the complexities of the LBP and wait until the price dropped before buying in? How would the presale, which received by far the biggest allocation, affect the progress of the LBP? How many people would liquidate their HOPR tokens from the presale once they unlocked?
We had intuitions about all of these, but one thing you quickly learn in crypto is you can’t second-guess how the markets will behave, especially around a token launch. But another thing you quickly learn about crypto is you don’t have to guess: all the chain data is publicly available for analysis. So let’s look at how the sale went down, and which if any of our hunches turned out to be correct. This analysis uses only publicly available data extracted from the Ethereum and xDai blockchains.
Timeline
The HOPR token launch consisted of three phases: the pre-sale on xDai Chain, the Balancer Liquidity Bootstrapping Pool (LBP) and the transition to a Uniswap liquidity pool. In this analysis we’ll be focusing on Phase 2: the LBP. An analysis of the Phase 1 Pre-Sale, conducted shortly after it concluded, can be found here.
LBP price and volume
Analyzing a LBP is a little tricky because, unlike a more standard sale, a LBP is an automated market maker (AMM). In addition to the complications introduced by the token weightings regularly updating, it means participants aren’t just limited to buying: they can also sell tokens back into the pool.
How could people sell HOPR during the initial distribution? Well, there were two options: 1) Buy HOPR in the LBP and the sell them shortly afterwards 2) Convert HOPR that were purchased in the pre-sale (token launch phase 1) from xDai chain to ETH mainnet over a “bridge” after they were unlocked.
A side note: This second option was always intended to be possible, although not everyone realized it, due to some unfortunate miscommunication about what it meant for the LBP to “finish”. Once the LBP reached its final weightings for the two tokens, it didn’t close, it just transitioned to being a more standard liquidity pool until the Uniswap pool was established. This continued ability to trade was important because otherwise there would have been a period of almost 24hrs when the price of the HOPR token was effectively zero, which would have filtered through to a lot of automated data aggregators and produced confusion at a crucial stage in the token lifecycle. We’re sorry for anyone who misunderstood and didn’t realize they could sell during this period, but as we’ll see later most people who rushed to sell at this point actually ended up selling at the lowest prices.
One of our main worries when communicating the DAO’s plan for the launch was that people wouldn’t understand the LBP mechanism. LBPs have a price-setting function, with constant downward price pressure from an artificially high initial price. This is a clever way to let the market find an acceptable price, but it doesn’t mesh well with crypto sentiments like hype, FOMO, and “get in now, ask questions later”. We explained this repeatedly in our blogs and on our various channels, but we were still expecting at least a few trades in the opening stages.
To our surprise, a significant amount of trading happened at the start of the LBP. This heat pushed the token price up from 1.92 DAI/HOPR (opening price) to its highest point (2.33 DAI/HOPR). After the first hour, the rush of buy-ins tapered off and the token price started to drop as the LBP mechanisms took over.
Hourly trading volumes rose and fell periodically, of course, with the lowest trading volume each day at around 02:00 CET. But the graph shows a gradual and expected decrease in price as the LBP was re-weighted over the three-day period.
Then, around 14:00 CET on Feb 27th, there was a sudden increase in trading volume and a drop in token price, forming a “V” shape. This is when pre-sale tokens were unlocked on xDai Chain. Some presale token holders quickly bridged their tokens back to Ethereum mainnet and sold them on Balancer LBP.
This had a strong short-term effect: the token price dived from 0.767 DAI/HOPR to 0.470 DAI/HOPR within an hour, its lowest value during the entire LBP. But this dip was short-lived: the price bounced back to 0.658 DAI/HOPR after another 50 minutes.
Activity was high during this period: 4.61 million and 3.26 million HOPR tokens were traded during these two hours, respectively. This is a clear majority (82.3%) of pre-sale trades. Volume-weighted price dropped from 1.23 DAI/HOPR to 1.12 DAI/HOPR.
Presale withdrawal
As seen in the LBP price chart, there was some selling pressure coming from the presale tokens. But what percentage of the presale tokens ended up in the LBP? Was it just a few people cashing in, or did it represent a major sell-off? Below is a breakdown of the presale token withdrawal.
50 million HOPR tokens were available for purchase in the presale. At the conclusion of the presale (23rd of Feb), just 1.38 million tokens were left unpurchased. These were bridged back to Ethereum mainnet by the DAO to add to the initial liquidity for the LBP.
39.7m xHOPR (81.7% of the pre-sale) were claimed on the first day of pre-sale unlock. This smashed through the threshold for raising the personal cap, so it remained at 16,000 xHOPR per participant.
Looking at the unlock history, we see that the first three hours after the tokens were unlocked was by far the busiest claim period. Around 78% of the claims that day (accounting for 64% of the total claimed tokens) occurred within the first three hours.
Presale tokens and the LBP
After presale tokens (xHOPR) have been claimed on xDai chain, they can be
- bridged back to mainnet for trading on exchanges;
- traded on exchanges which live on xDai chain;
- wrapped to wxHOPR for use in a HOPR node.
Option 1 is what caused the price drop at the end of the LBP. We can use data from the Omnibridge to look at how many presale tokens took this route.
A total of 48.62 million HOPR tokens were sold in presale. Of these, 42.80 million were claimed before the LBP ended. Roughly two thirds of these (28.70 million) were part of a cross-chain conversion attempt, taking them across the bridge from xDai to mainnet.
We say “attempt” because, of these initiated transactions, only 28.32 million tokens were successfully bridged. The unsuccessful attempts were mostly due to participants failing to pull the bridged tokens from the mainnet side.
This might seem to support a theory of a sell-off among DAO participants. However, of these bridged tokens, just 9.56 million HOPR (33.8 % of bridged tokens) were sold into the LBP.
The funnel plot below shows these figures together.
LBP Traders
Looking at the LBP as a whole, 5,879 transactions were made by 3,630 unique investors. Many accounts transacted just once, but there were also people actively trading throughout the duration of the LBP. We’ve identified 325 “traders” who made at least one buy and at least one sell transaction with the pool.
The bubble chart above illustrates the profit and loss of LBP traders. Dots above the 45° line represent cases where effective sell price was greater than effective buy price, so the trader made a profit.
Bubbles below the line represent accounts that made a loss. The size of the bubble represents the absolute size of each effective gain/loss. The highest gain was $45,750.09. The largest loss was $285,663.66.
These effective prices consider gas costs. Most effective buy/sell prices were below 2.5 DAI/ HOPR. However, two accounts managed to spend more on gas than tokens! These were treated as outliers and excluded from the chart. (More on those crazy gas prices in a little bit.)
Of the 325 trader accounts, only 78 profited from this trading, which is unsurprising given the downward price pressure of the re-weighting. It’s also important not to read too much into these figures. Shortly after the LBP opened, several smaller platforms opened up HOPR trading pairs, so the LBP doesn’t represent 100% of all HOPR trades during this period.
Focusing on the period after the presale tokens unlocked, we see there were 1364 LBP trades between presale unlock and the pool closing. Of these, just 500 were to sell HOPR, and only 357 could be identified as coming from accounts holding presale tokens. 9.5m HOPR were exchanged for 5.8m DAI, an average price of 0.6 DAI/HOPR.
So while a portion of presale participants did choose to sell, it was a minority, and the downward price pressure this created was soon absorbed by buyers.
Exorbitant Gas Costs
Finally, it’s worth taking a look at the gas expenditure involved in a major event like this. As veterans of the Ethereum space since the early years when even complicated interactions would cost just a few dollars, it’s still hard to wrap our heads around current prices, so we didn’t really have any intuitions here except that the numbers would be big. While holding the presale on xDAI Chain mitigated this problem for pre-sale participants, no such provision could be made for the LBP.
Taking an average Ethereum price of 1500 USD/ETH, the gas costs associated with trades during the whole LBP period totaled more than half a million dollars: an astonishing $578,121.41 to be precise.
If that wasn’t upsetting enough, while participants spent $373,323.21 on successful transactions, they burned an eye-watering $204,798.20 on failed transactions, more than a third of the total. These failed transactions were mostly due to “out of gas” errors and price slippage.
It’s not clear what lessons we can learn from these particular stats, except that there needs to be more information available to ETH users about how to calculate and set appropriate gas levels for transactions and the implications of slippage when using DEXs. For those of you who prefer the raw data as opposed to the narrative above, you may find those here.
Lessons Learned
So what have we learned? First, it’s extremely gratifying to see how smoothly everything went. The DAO had a lot of freedom to decide on the distribution mechanism, which in the end had a lot of moving parts across multiple chains. The LBP in particular was a complex choice from the DAO, but most participants had no difficulty, and it seems to have done a good job of establishing an initial price for HOPR tokens without requiring the HOPR Association to just pick an arbitrary number.
It was great to see that participation in the DAO carried over to the presale, with no need to trigger the mechanism for lifting the personal cap. We’re especially grateful that confidence in HOPR is strong enough that the prevailing sentiment has clearly been to hodl. It would have been perfectly understandable (albeit a little disappointing) if a lot more presale participants had cashed in once their tokens unlocked, and it was wonderful to see only modest pressure here which was quickly absorbed. HOPR is committed to decentralized governance, but there’s no denying that handing the reins to the Genesis DAO was a risk. It’s amazing to see that the community more than warranted the confidence we placed in them.
Qianchen “Q” Yu
HOPR Decentralized Technology Architect
Website: https://www.hoprnet.org
Twitter: https://twitter.com/hoprnet
Telegram: https://t.me/hoprnet
LinkedIn: https://www.linkedin.com/company/hoprnet
