Thread by @2lambro

A short story on why the @Umami $GLP vault failed and how did its users get rekt:

#GMX

If you are still reading you probably know how $GLP works and where the yield comes from. You probably also know that the GLP vault strat is supposed to safeguard against volatility by hedging the underlying assets in GLP, but does it actually do what it says on the tin?

The hedging strategy basically goes like this:
Assuming $GLP asset allocation = 50% Stables/25% $BTC/25% $ETH. When the vault buys $100 of GLP for 20% yield, it also opens shorts on Tracer for $25 of $ETH and $BTC for say 5% interest. Boom, you earn 15% delta neutral yield.

But Ser why doesn’t it work?
It's because GLP is a basket of tokens with constant changing weight. The weight of the tokens change as users mint/redeem $GLP, when open interest on #gmx changes or when @GMX_IO is utilised as a swap because there is 0 slippage 🐋🥰

Imagine if the weighting of #ETH within $GLP went down but @Umami didn’t rebalance in time. This will result in an over- hedge on ETH. If ETH went up in price, the excess short ETH position will REKT umami's PnL.

Because of the nature of Tracer's mechanism, umami can only rebalance every 8 hrs. Asset weighting can move significantly within each 8 hr interval, with this risk exacerbated during volatile market conditions (like we are seeing now)

Using Tracer also means there is a cap on how much they can hedge as well hence the 5m cap. (Thank God for saving the Degen’s aping into delta neutral)

The other downside is that tracer’s "EFFECTIVE LEVERAGE" does changes depending on if the price of the underlying asset increases or decreases since the last rebalance as well.

#Umami GLP Vault is not a bad product and there isn't a better designed strat out there yet, when keep an eye out for these features in future vaults or improvements:

-Active monitoring of changes in GLP weighting
-Tighter rebalancing timeframe
-Improved capital efficiency via debt ratio management

Follow me for more #alpha