Similar to epoch 0 in dydx, wash traders were present to pump volume up to ridiculous levels prior to closure on MCDEX. Now, I don’t think we should sniff every wash trader because it will be logistically difficult and cumbersome, but we should do so when obvious wash trading occurs, especially on such large and obvious case. Otherwise it would significantly hurt the spirit of fair and decentralisation values, pushing a large amount of MCB to for-profit actors that would harm the long term value of MCB.
Now, let me bring your attention to a specific timeframe:
During the last hour prior to the end of Epoch 0, and in this case, on the ETH-BUSD pair there were activity that we should consider as ‘wash’ and remove them from receiving eligible rewards. I hope you would hear my case and inform me if it is wrong. As you can see in the graph above, there is unusual volume traded despite only 18 unique addresses who traded during the time period.
Firstly: The timing. It was no coincidence that specific addresses decided to trade at this particular hour. It was the final hour of the first epoch. I don’t think we can determine such trading activity is in line with normal trading because the volatility and volume of ETH traded on CEX during this hour was no different from the prevailing or succeeding hours.
Secondly: 2 pairs of address both have very similar fee generated. Of the 18 accounts that traded during this hour, only both pairs had such similar trading fees, with delta percentages of only 0.05%. No other address have such similar deltas. The closeness of the trading fee will denote the closeness of the trading VOLUME as well. This means that during this hour long period, both pairs of address have VERY similar volumes traded.
Thirdly: Now in terms of trading observation, both pairs utilise different strategies.
(A) For the yellow and pink pairs, long and shorts from each respective pair were opened at very close times apart, with the SAME notional. Few minutes later, both pairs were then closed, again, at very close times apart, SAME notional as well. They racked up $4.4mil+ in volume.
(B) For the blue and green pairs, this one is interesting. A long and a short is opened one address at the same time. Few minutes later, this address will close this position, while the other address will open a long / short with a Different notional. This happens several times over the hour. Over $600m+ of volume was racked up from this two address, constituting more than 90% of the volume during this HOUR. Now, i only investigated for the hour prior to end of epoch 0, but it highly suggest they also faked volume for the DAY prior to the end of epoch 0, around the same time as well. This is serious as it constitutes to almost 50% of the volume for the ETH-BUSD pair over 1 WEEK, which is HALF an epoch. This is Massive. I don’t think it is fair to distribute a LARGE proportion of the MCB rewards to a single sole individuals / EOA. I have strong suspicions that they also manipulated the BTC-BUSD pair as well in similar fashion.
Proposal: Look, its hard to deal with every wash trader that comes along the team’s way. There will always be actors looking to game the system. It isnt worth the effort when it doesnt significantly harm the system of fair distribution. I don’t want to penalise the pink and yellow addresses because while it is obvious, the material impact is not large. For the blue and green addresses however, which are likely bots (as they utilise a smart contract to trade and not directly via the UI), we should penalise them and not distribute them rewards. We will be distributing Too Much rewards to a concentrated individual where I feel there is material impact. I hope @jie and @jean1 , can make an exception to this bot because it is harming the MCDEX protocol so early on.