Historically, Index Coop has used Set rather than Balancer for our portfolio creation needs. It's not clear why this has to be the case, necessarily, and certain products lend themselves to Balancer pools more than others. For example, behavior-wise, BED targets a 1/3 composition of BTC, ETH, and DPI, which is literally what Balancer does. Intuitively, it seems that Set is better for token-denominated compositions (e.g. market-cap weighted portfolios, leveraged ETFs), and Balancer is better for value-denominated compositions (maintaining 33% in three assets, maintaining an 80/20 stocks/bonds ratio like a target date fund)
However, Balancer has claimed that they can also operate token-denominated portfolios by locking the portfolio composition and not allowing it to trade for 90% of the time, and then changing the weights to a new target weighting and letting the market "arb" you to your desired weighting. The claim is that you save on gas costs and price impacts.
I do not think this is accurate - I think there is a genuine rebalancing cost that is apples-to-apples with Set Protocol rebalancing costs, and I'd like to measure it.
Target deliverable is a spreadsheet that can simulate the multi-step operation of changing a balancer pools' weight and then a subsequent arb back given a constant externally determined spot price for a 2 asset pool.
jackiep00
commented
2 years ago
Historically, Index Coop has used Set rather than Balancer for our portfolio creation needs. It's not clear why this has to be the case, necessarily, and certain products lend themselves to Balancer pools more than others. For example, behavior-wise, BED targets a 1/3 composition of BTC, ETH, and DPI, which is literally what Balancer does. Intuitively, it seems that Set is better for token-denominated compositions (e.g. market-cap weighted portfolios, leveraged ETFs), and Balancer is better for value-denominated compositions (maintaining 33% in three assets, maintaining an 80/20 stocks/bonds ratio like a target date fund)
However, Balancer has claimed that they can also operate token-denominated portfolios by locking the portfolio composition and not allowing it to trade for 90% of the time, and then changing the weights to a new target weighting and letting the market "arb" you to your desired weighting. The claim is that you save on gas costs and price impacts.
I do not think this is accurate - I think there is a genuine rebalancing cost that is apples-to-apples with Set Protocol rebalancing costs, and I'd like to measure it.
Target deliverable is a spreadsheet that can simulate the multi-step operation of changing a balancer pools' weight and then a subsequent arb back given a constant externally determined spot price for a 2 asset pool.
Helpful resources: https://medium.com/balancer-simulations/understanding-balancer-pools-c2b877dcc082 Essentially i am trying to chain Scenario D and Scenario C together: