Examining the Supply of Dai
An Analysis into CDP Demand
Last week, Hasu posted a great analysis of Maker Dai and its ability to scale. In summary, he explained how unlike other stablecoins, the supply is not directly linked to the demand due to the lack of direct redemption for the underlying peg. Rather it is the demand to take out a collateralized loan primarily driven by the desire to lever existing position, manage project treasuries or avoid capital gains tax. In my prior analysis and valuation of Maker, I incorrectly assumed the supply of Dai would be related to demand for it. Here I run analysis on various scenarios tracking the demand for loans assuming Multi Collateral Dai is implemented and wrapped bitcoin (wBTC) can be used as collateral.
Note: This is a simplifying assumption as there will likely be multiple large-cap cryptoassets collateralized. This can be thought of as a catch all for cryptoassets collateralized on MCD since they generally move with a high correlation.
Similar to my last report, I will be using scenario analyses to look at how various input changes affect the supply of Dai given they are highly subjective and prone to large fluctuations. This allows you to view the projected Dai supply using assumptions you believe to be reasonable.
Sensitivity Analysis 1: Fluctuations in Ether and Bitcoin Price
The first analysis I ran is looking at how the supply of Dai changes given increases in the price of the underlying collateral. The rationale behind this is as the price increases more users will want to take out CDPs denominated in larger USD amounts. If you have $10,000 in ether which is later worth $50,000 you will take advantage of the increased access to liquidity. Instead of using a fixed percent of ether locked up, I used the Dai outstanding as a percent of the total ether market cap to account for this. (Currently ~0.5% according to Makerscan)
As you can see the price of the collateral has significant effects on the amount of Dai that will be created through CDPs. The introduction of potential collateral with a market cap similar to that of bitcoin increases the Dai debt from ~$70 million to over $400 million. Of course this will not happen immediately, but likely over the same time period it took ether to have 0.5% of its market cap taken out as Dai via CDP.
Sensitivity Analysis 2: Fluctuation in the Debt/Market Cap %
In this analysis I kept price fixed and looked at how changing the debt taken out as a percentage of market cap effects the Dai supply. Rather than use approximate current prices, I used bitcoin and ether prices around half of prior all-time highs of $500 and $10,000 respectively. (If you’re reading this you probably believe we’ll see those levels inevitably).
Again these inputs create a lot of variability, but demonstrate the levels of outstanding Dai that could be reached. This suggests we could see a supply increase around 60x current levels, which would have a direct increase in the MKR burned to repay these loans. As mentioned in my valuation these interest payments act as pseudo cash flows to MKR holders which can be discounted and provide an estimate of fair value for MKR.
In conclusion, Dai is unlike fiat-collateralized stablecoins in that it does not scale with increased demand for the stablecoin. However, I disagree with Hasu in that it is not scalable since a growing cryptoasset ecosystem will increase the demand for collateralized loans. These models look at various inputs and demonstrate the potential growth in Dai, yet they don’t look at what would happen in a situation where we see prices above prior all time highs. Another factor not taken into account is the tokenization of real world assets. This is a divisive topic but if it’s something you believe could happen, it would increase the potential collateral by orders of magnitude creating vastly more Dai as well as value to MKR. Regardless, MakerDAO is one of the more exciting projects in the space and I’m looking forward to following its progress as it continues to mature.
If you’d like copies of this model as well as my prior valuation model, feel free to shoot me an email at firstname.lastname@example.org.