Improving Arbitrage & pAsset Stability On Exchanges — Status Proposed Never Implemented
PIP 8 Describes a Variable Conversion Rate Into PEG Tokens
TLDR; In order for pAssets such as pUSD to hold their value on exchanges, compared to their value on the PegNet, arbitrage has to function well and respond to market signals. The purpose of this 8th PegNet Improvement Proposal “PIP 8", is to introduce a variable PEG conversion limit so that arbitragers on the PegNet can more quickly respond when pUSD for example is either above or below 1% of its PegNet reference price of $1.00
The Arbitrage Cycle Today on PegNet
As a reminder this is how the arbitrage cycle works on PegNet.
Today when an arbitrager wants to follow this pattern above they can buy cheap pUSD for example off an exchange and then send it to their PegNet wallet. However in the next step they can only try and convert a small portion of their pUSD into PEG due to the 5,000 PEG per block limit introduced when PEG prices went free floating. Two way unlimited PEG conversions have a known attack vector, so some limit is required.
It was anticipated that 5,000 PEG per block might be too little supply in order for timely arbitrage of pAssets, but the conversion rule was kept simple in the first version in order to gather data and it was discussed and widely agreed that a “buffer” would be introduced in the future to allow for more flexible arbitrage.
This slow arbitrage cycle has now been witnessed and it has confirmed that 5,000 PEG conversion per block is too restrictive in order to quickly return pAssets to their reference prices. Recently pUSD fell under $1.00 and it took about 30 days for the price of pUSD on the exchanges to return to $1.00. With only 5,000 PEG available per block, it took this long for 21,600,000 PEG to be converted.
Upgrading Arbitrage Going Forward
This proposal articulates the means of creating this “buffer” that was previously envisioned by introducing a variable PEG conversion rate that responds to the market.
1. The growth of the PEG conversion limit of 5,000 per block will increase by 500 PEG per block when the pAsset exchange prices are more than 1% off of the PegNet reference price.
2. This 500 PEG increase will continue (provided all the PEG is converted into during each block) for up to 288 blocks. Thus there is a cap of 149,000 PEG per block available for conversion should the price gap persist for 2 days.
3. After the pAsset prices on exchanges return to within 1% of their PegNet reference price, then the amount of PEG available for conversion will decrease by 500 PEG per block until it returns to the normal 5,000 PEG per block level after 288 blocks.
You can see the math here on this Google Sheet.
This gradual increase and gradual decrease over a few days is intended to add flexibility for when more arbitrage is required, without re opening the two way unlimited conversions attack vector.
The Attack Vector That PegNet Has to Avoid
1. If conversions into PEG and out of PEG were unlimited an attacker could push down the price of PEG on exchanges (due to its current low liquidity) and buy say 1 Million PEG cheaply, for $1,000.
2. Next the attacker could push up the price of PEG by 2X on the exchange and then frictionlessly convert their 1 Million PEG to $ 2,000 of pUSD inside their PegNet wallet.
3. The attacker then simply repeats this process. They can push down the price of PEG again by 2X and convert their $2,000 of pUSD into 2,000,000 PEG.
4. The attacker can then pump the price of PEG by 2X & convert their 2 Million PEG into $4,000 of pUSD.
With every pump and dump cycle the attacker is able to double his money. Even as PEG liquidity increases, a smaller push of 20% up and down each cycle would allow the attacker to double their money every 5 cycles.
Cycles can be as short as 2 or 3 blocks (20 to 30 minutes). So before long (a few days or weeks) the attacker can have a practically unlimited number of pUSD, PEG or any pAsset in the system & swamp / dilute out everyone else who mined or bought their PEG & pAssets fairly.
This is a REAL attack vector that was witnessed at a small scale early on in the history of PegNet with an attacker who was pushing around the price of FCT. The attacker repeated the attack for 6 cycles & in that short period managed to make about a 25% gain in only a few hours.
This is the reason the developers of PegNet proposed and introduced one way conversions on less liquid / smaller market cap tokens under $100 million in market cap and under $5 million in daily trading volume. FCT & PEG are both currently under these required liquidity minimums and so conversions INTO FCT & PEG were made one way in order to close the attack vector.
Conclusion / Striking A Balance
In the case of PEG one way conversions can not be absolute, as some conversion into PEG has to be allowed for the arbitrage cycle to function as intended. So the question becomes how much conversions into PEG should be allowed.
This is the balance PegNet has to strike.
1. Offering arbitragers enough flexibility to convert into more PEG when the market signals that pAssets need more arbitrage.
2. Without re opening the PegNet up to an attacker who can enrich themselves by manipulating the price of PEG up & down in short periods of time.
I believe this proposal strikes a good balance with the parameters proposed and with the cap on increasing PEG conversions after 2 days (288 blocks). These parameters are based on the data we have so far about how quickly pAssets are returning to their reference prices. These parameters may well have to be tweaked in the future as we gain more real world data, but in the mean time “PIP 8" offers a significant improvement over the current static 5,000 PEG conversion limit.