I have recently got a request on clarifying The Rock Liquidity paragraph from my Rock Liquidity Pool Pinkpaper. I admit it is a bit too short for its importance in understanding the need for such a mechanics. Unfortunately, the full explanation would be a post of its own. Wait, now it is a post of its own!
What if we crowdfunded 20M Hive for the bridge? Seed the bridge with 10M at each side at watch people run at the currently preferred side. After spending 50k SWAP.HIVE (taking out roughly 50k HIVE), our Liquidity Pool is at 9.95M vs 10.05M which sets the price at 0.99. The run cools down and people start using the default gateway instead as our price is no longer competitive.
This paragraph painted a fictional world where someone enters the market with a traditional Liquidity Pool and realises what an overkill it was (in terms of money needed to operate). Theorists enjoyed. Practically oriented people probably prefer to see Rock Liquidity Pool in action instead - so let's do that:
Here we deal with a boring situation where personal preferences between having HIVE and having SWAP.HIVE are well balanced over the population resulting in a fair frice of SWAP.HIVE of 1 HIVE. We entered the market with our dynamically priced Rock Liquidity Pool and set our fee to 0.3%. So people who want to swap have three options now:
Fineprint: Hive-Engine fee has gone down to 0.75% since the original article was published but the old numbers are kept for simplicity. In the same spirit, different discounted bridges have different fees so they are not going one-sided all at the same time (the ones that offer the lowest fees will be emptied first). HiveUpMe from the WINE token team is the newest addition.
Everyone is going for the DBs and we earn nothing because we are too expensive. Then an event happens. Say dCity releases a new earning opportunity and personal preferences between having HIVE and SWAP.HIVE change.
The DBs start trading one way accumulating HIVE and running out of SWAP.HIVE. Two scenarios can happen.
DBs have enough liquidity to accomodate everyone. The rush ends, and a similar event on the other side of the gateway incentivises the trading in the opposite direction, balancing their books. We yawn.
DBs dry out and the demand continues. Now, people have to buy our SWAP.HIVE with first small transaction priced at 1.003 HIVE (still good enough for them). But RLP is a Liquidity Pool and that means pricing changes as the two sides get unbalanced. The transactions pile up (in the same direction) and our pricing keeps going up. We earn.
By the time we sell for 1.006, there are SWAP.HIVE holders gathering on the street. Why? They smell an opportunity. Our books indicate a fair rate of 1.003 which means we sell for 1.006 and also buy for 1.000 - another pip up, we sell for 1.007 and buy for 1.001. If you are truly indifferent, why not go and sell your SWAP.HIVE to us getting an extra 0.1%?
This is the kind of market RLP serves well. The only downside is if it takes too long, SWAP.HIVE holders will no longer be happy with 0.1% and they start waiting for better deals. They can get as much as 1.004 if our RLP indicates 1.007 is the fair rate but not more - that is because we already sell at 1.010 at that point at people might as well go to the HEG instead. So everyone indifferent is going to sell all their SWAP.HIVE to us at just below 1.004 keeping our business live for another while.
If the rush is still on, we just end up on the sidelines as well and watch HEG print enough new SWAP.HIVE to saturate the market. And the best part - as the rate continuously slides back to 1.000 we are the first to open shop as the SWAP.HIVE holders start selling their newly not-so-wanted tokens to us at 1.004 and not the DBs at 0.998.
Obviously, our fee does not have to be fixed and it can change based on the fair price but it has been intentionally kept simple throughout the article just as the other numbers.
But where is the 9.95M-10.05M figure from the quote? Well, we could have done all of that with a traditional liquidity pool (if we had those 20M at hand). But once we observed where our money printer gets stuck, we can see having 100k actual liquidity and smart price computation aka 19.9M fake liquidity does the same job as 20M liquidity in traditional LP mechanics.
