[DAO Discussion] Treasury Hedged GLP Investment

Hi All,

I propose to allocate a portion of the treasury to a hedged GLP strategy in order to generate market neutral returns for wMemo holders.


  • Post redemption treasury value is roughly $135mn, of which $81.5mn is in stables → these stables should be put to work
  • We can generate good yield from a proven platform in a market neutral way by using a hedged investment into GLP on GMX


  • GLP is a token comprised of a basket of coins on the GMX trading platform
  • GLP is the counterparty to every trade on GMX (trader goes long btc, GLP is short btc) (trader loses money, GLP wins money)
  • GLP earns 70% of trading fees on GMX
  • GLP is primarily comprised of BTC, ETH, and stablecoins
  • Volatility in GLP value comes from price exposure to the cryptocurrencies in the basket (BTC, ETH, etc) as well as the gains and losses of traders (if traders are net long BTC, BTC goes up, GLP value can drop even though GLP has BTC exposure)


  • As the GLP investment has exposure to the price of its underlying basket of assets, some hedge shall be made in order to earn the yield regardless of wider market movements
  • Bitcoin has historically been heavily correlated to the price of GLP, and a simple BTC short can effectively hedge the wider exposure of coins in GLP


  • I have extensively modelled this trade using crystal ball monte carlo simulations in excel, key assumptions are presented in the slide below (happy to provide model access to those who are interested)
  • From April - May of this year I executed this trade and earned roughly 20% on an annualized basis (happy to provide proof to team members)

Estimate Returns

  • GLP roughly generates a 15% APR in fee revenue (average between Avalanche and Arbitrum APR) and additionally earns some vested GMX (known as esGMX) which is illiquid but provides some additional yield. I assign esGMX no value
  • My modelling suggests a median return of 9% and mean return of 14% - I believe a reasonable return expectation to be closer to 9% based on my own experience


  • Significant smart contract risk. Would appreciate feedback from those with dev experience on relative risk for GLP
  • Risk of traders winning outsize profits at expense of GLP value (unlikely to be significant in long run)
  • Risk of GLP declining in value due to ETH significant price deviation from BTC (unlikely imo)

Investment Recommendation

  • I propose that the treasury allocates 1/4rd of its stablecoins to GLP on Arbitrum, and hedges 75% of the position with a BTC short on dydx using 2x leverage
  • For active discussion - is the return of 9% or even less worth the potential SC risk?



This is something we have talking about in discord a lot it would be great to have and all but we wont have someone to manage it.

Also is a hedge completely necessary?

I would personally vote against this investment if it were not hedged. Hedging risk is key in this market.

Fair enough, just curious what it would look like :thinking:

Am GLP holder personally, so I obviously think this is good. I am all for it.
Why Arbitrum though? Wouldn’t it just be easier for us to be on avax, since we’re avax native?

Also, let’s pick up one of the NFT’s.

1 Like

I don’t like the AVAX exposure in avalanche GLP, too volatile.

Also I think Arbitrum odyssey will fuel trading fees on Arbitrum in the near term. Long term, Arbitrum is lower fees and faster than avalanche and so should have higher fees for GLP.


Glad someone finally took the initiative on this. There is a similar DAO discussion from Tika Marsalas, and GLP seemed to have significant support based on the informal votes. I think a hedged position would be great and we can slowly adjust the hedge as the market turns more bullish. We need to be looking at more hedged/delta neutral farming strategies right now in this market. Given GLP is roughly 50% directional I think we could reasonably expect a 15-20% minimum annual return (including the hedged capital)–although GLP does fluctuate a bit from 30-40% APR (on AVAX chain)…after including the hedged capital were looking at another 33% decrease in returns if you hedge the directional portion of GLP (roughly 50%) at a 1:1 ratio using a non-leveraged hedge. Probably worth it given we have potential further downside risk in the market.

FYI: Arbitrum yields have recently nuked…avax is more than double. To me GLP on avax seems like a better option. I assume you ran your model through Arbitrum (for GLP), if you run it through Avax do you get a similar rate of return? Also, does your concern regarding the volatility of Avax really matter that much if you’re hedging at a near equal rate to the directional allocation of GLP. Wouldn’t the higher ROI of GLP on Avax outweigh the volatility risk?

Edit: I see your advocating for a 75:100 short vs GLP ratio allocation after reading it again and looking at your simulation model. I think this is fine but we will likely need to adjust once market corrects.


The risk of invest into GLP is mainly SC & governance risk instead of price volatility. The latter can be easily hedged by literally shorting (spot or perp) AVAX as it is normally the most volatile assets among the others while the former cannot be hedged.

If wonderland want to go big on GLP, i think it to equally important to acquire certain share of governance power (by purchasing GMX) to safeguard the investment.

I prefer AVAX to ARBITRUM, because of the following reasons:

  • AVAX is a proven blockchain with barely any downtime;
  • GLP on AVAX is less exposure to altcoin (AVAX only)
  • GLP on ARBITRUM has exposure to more altcoin (LINK, UNI, FRAX);
  • AVAX has native USDC
  • AVAX has higher yield

To be fair you should not think its good since it dilutes yer yields xd

I’m going to play the bad guy since no one else wants to and it’s important imo. In this model you assume that BTC will on average decrease by 15% annually. Will you show us the results with other market assumptions (both mean and volatility)? Modeling out only 1 BTC assumption seems extreeeeeemely bad (even if Monte Carlo sims have varying results). 15% is a rather tame move for a crypto, I’d like to see a more comprehensive reporting.

Additionally, since the median return is 9% that means the 14% average is being skewed by a few extremely high return scenarios. In fact, there seems to be a 30+% probability this strategy loses money. This begs the question, why would this strategy be preferable to just farming in pure stables?

1 Like

I agree, the strategy as outlined may not be worth the risk in comparison to just farming stables if you only expect a 9% return on Arbitrum. At that point I think Avax GLP is a much better option–and I think I can safely make the assumption that Sifu feels the same given his own portfolio is GLP on AVAX. My question to you is why are you evaluating the long-term benefits of Arbitrum regarding this investment? If avax is more profitable in the here and now, why not invest on that chain and if Arbitum is more profitable later we can switch chains? If were dumping the Avax weekly for stables and hedging the volatility your concerns don’t seem valid. Maybe you can provide some clarification here…

I made an assumption that Arbitrum odyssey would increase the trading fees for GLP on Arbitrum relative to Avalanche.

Fair point that higher returns now merits just going with Avalanche and I agree. No reason to speculate on potentially higher returns in the future when there are higher returns in the here and now. I agree, Avalanche makes more sense.

Great points and thank you for looking into the literature closely.

The reason the strategy has a 30% probability of losing money is likely due to the very high std. deviation of daily bitcoin returns. My computer is not very fast and so I ran about 50 trials using 300 samples each trial. If I ran more samples or removed the outliers, I believe we would get a higher probability of positive returns.

Not sure that only 9% returns merits the risk, agree with you there.

This topic was automatically closed 7 days after the last reply. New replies are no longer allowed.