Imagine that we have inverse iAsset to very mAsset, such that the multiplication product of the two prices always equate to a constant.

For example if we have mAPPL at 100 we all ways have iAPPL at 10000/mAPPL = 100, and if mAPPL doubles to 200 the iAPPL will halve to 50, and vice versa.

This provides a way to “short” a certain asset, and if we have staking reward for iAsset-UST, it will counter balance the demand for mAsset-UST, as this will help the mAsset track the oracle price better.

Furthermore if we dynamically adjust the staking reward such that, when the price of mAsset is higher than oracle price, we make the staking reward of iAsset-UST higher than the staking reward of mAsset-UST, driving up the demand for iAsset, and dropping the price of mAsset in the process. and vice versa.

inverse assets as you propose probably wouldn’t work because you would have a winning strategy by dynamically hedging a delta neutral position of the asset and its inverse

Even without rewards, you would have a winning strategy if peg was exact.
If you hold 1 m-btc and btc^2 inverse-m-btc (with dynamic rebalancing) you are pocketing something that depends on btc volatility.
You can see this by applying Ito formula to get the stochastic differential equation of the inverse asset.

Sure, you’re just volatility trading, but this is not a bad thing and doable on any developed platform (or even cross platform). Allowing for dynamic hedging is probably a good thing as well probably.

In the case of looking for staking/LP rewards and given the fact that rewards are much higher right now and a non-exact peg, this shouldn’t pose any significant problem.

Yes, but the difference is that inverse assets are not presently tradeable anywhere, so by making them tradeable, you either change global market structure or you lose the peg

I agree with the fact that the originally proposed inverse asset structure is not great, but inverse assets structured in a different way should not be extremely different from product types on Synthetix/Bitmex/etc.

So I have put out a rough spec internally for a mechanism somewhat similar to Synthetix, but there’s still a few things that need to be discussed or come to a consensus. I’m not a big fan of the price ceiling/floor with constant deprecation and relisting, since it would be very cumbersome to do this all the time.

I will write something up publicly sometime mid-week or so

@Sihyeok I don’t want to be off topic here, but did you see the proposal I just made on automatic supply adjustment to fix the premium problem? Can’t find a flaw in it but don’t understand things enough to be even remotely sure

i dont think inversed asset prices would make a diference. there would still exist a premium, just to the other side of the spectrum.
so for example if apple’s oracle price is 100usd. terra price is trading at a 10% premium which means 110usd. the inverse price would be at 90usd. if oracle goes to 110, terra would be 120 and the inversed price would be 80usd.

premium exists cause of the minter risking their collateral, they still need to put collateral in if they mint inverse assets.