This dashboard compares the price of UST on Terra's Oracle vs the price of UST in the UST-USDT pool on Uniswap. It tries to drill into the differences in price between Terra and Uniswap and also the causes of these price differences.

Comparative Analysis

As always, let's start off simple by plotting the price of UST (averaged hourly) on Uniswap's UST-USDT pool (using the 0.05% reward liquidity pool since that is the only pool instantiated on Uniswap) and also on Terra (price determined by the Oracle):

https://s3-us-west-2.amazonaws.com/secure.notion-static.com/b6f5a58f-e41d-435f-9d3b-2bed1e6fd950/Screenshot_2021-07-29_at_02.22.14.png

(Interactive Link: https://velocity-app.flipsidecrypto.com/velocity/visuals/dcfb1765-3986-4548-8a3b-146e69e0d2d5/d18e1495-a21c-4cbb-967f-36868114a3f5)

This is interesting. We see that initially UST's price vs USDT (i.e. the amount of USDT required to purchase 1 unit of UST) was quite high and then as the fear of a bank run approached the price deviated significantly from the peg to the downside (this downside was still smaller than the one experienced by the Terra Oracle). It also seems like the Oracle Prices are significantly more noisy on Terra but the deviations in prices last for shorter periods of time. What I mean by this is oscillations from the expected price (of 1 USDT) despite not being as frequent take longer to recover from on Uniswap (as we can see on the 9th of May and the 19th of May). This may possibly be due to the lack of arbitrageurs on Uniswap for the UST-USDT pool and an overall limit on the amount of liquidity available.

The next graphs show us the deviation from 1 for both the Uniswap and Terra prices of UST. This should just clean up what we saw in the last graph and showcase the point of deviations taking longer to recover on Uniswap by colouring a whole area of prolonged deviation on Uniswap pink:

https://s3-us-west-2.amazonaws.com/secure.notion-static.com/8fda843c-718a-41f7-b18e-cdd1a4f4f5e6/Screenshot_2021-07-29_at_03.14.42.png

(Interactive Link: https://velocity-app.flipsidecrypto.com/velocity/visuals/a0d0cbe6-7f9f-49e9-84fc-bdd345d764c3/d18e1495-a21c-4cbb-967f-36868114a3f5)

Another view of the deviation from the target price of $1 for UST on Terra and Uniswap would be to measure the absolute value of deviations from the expected price:

https://s3-us-west-2.amazonaws.com/secure.notion-static.com/9cd0bca5-1bff-4073-b361-1b05adc55702/Screenshot_2021-07-28_at_06.59.47.png

(Interactive Link: https://velocity-app.flipsidecrypto.com/velocity/visuals/a2a7fc84-c13f-4151-ae8f-b7a3bc52cfe7/d18e1495-a21c-4cbb-967f-36868114a3f5)

The graph above shows us the absolute value of the deviation from $1 for UST on both Uniswap and Terra. The use for this chart in my opinion is to see where the deviations lie and how large they tend to be (without needing to look at the direction of deviations). It seems like the largest deviation was on Terra for a small period of time with UST losing its peg by more than 3% and people fearing a bank run in the May crashening. Then there is a cluster of deviations of around 1.5% to 2.5% on both Terra and Uniswap at different points of time - June and July for Terra and twice in May for Uniswap.

This next plot just displays the absolute deviation in price between UST on Terra and UST on Uniswap:

https://s3-us-west-2.amazonaws.com/secure.notion-static.com/1ed8dd5f-d114-48c9-b5e1-7ab79120281e/Screenshot_2021-07-28_at_07.00.22.png