Efficient portfolio management

A straightforward way to look at Enjoyoor’s entire liquidity is to consider a portfolio of assets where collateral tokens have positive portfolio weights, and synthetic gigaAsset tokens have negative portfolio weights. With this portfolio analogy, we have the mighty modern portfolio management theory at our disposal, where we can build portfolios with different mean-variance characteristics, analyze asset correlations, pairwise cointegrations, and be very flexible regarding the portfolio objectives we target. As an example of a meaningful objective to pursue, consider the following risk optimization problem:

E.g. minimize the portfolio variance where one of the assets has a negative weight (mimicking synthetic debt). To illustrate, we’ve modeled a portfolio of 10 popular ERC-20 tokens and added ETH short to the portfolio to mimic debt in synthetic ETH. Here is how the portfolio performed on the unseen test set:

Figure 4: cumulative returns of ETH and model portfolio over the past year

Interestingly enough, the optimal weight for the ETH short turns out to be around 0.53 or 53%. This is not achievable with collateralized portfolios, as the entire collateral value is only 47% of the portfolio. This is already not a collateral-backed synthetic system, but rather a synthetically hedged portfolio, and is extremely capital efficient. Yet such a portfolio is very hard to manage, as hedging always incurs additional costs. So, initially, Enjoyoors will operate as an overcollateralized system. We then aim to gradually migrate from an overcollateralized system to a hedged portfolio approach as we build liquidity and momentum.