In-Depth Analysis Validates Bumper's Market Functionality with BTC vs ETH Insights

One of the final steps in moving the protocol forward was to understand the differences between the BTC and ETH market yields and premiums. This has come up a couple of times in our community, and although the belief was that this was just down to how the markets work, we wanted to run to prove this through deeper analysis.

Specifically, we wanted to know, based on the current state of the protocol, if these market differences are valid.

And the answer is, Yes! Bumper is working as it should. **cracks open a can of SOLO**

‍

Displaying data

Early users will know that the Markets Overview section at the top of the dashboard was one of the interface improvements we’ve delivered since launching on Arbitrum. Our backend telemetry shows a range of protocol health data points, but displaying this in the front end, as weighted averages, highlighted the difference between the two markets more obviously and, also brought more clarity for users on likely premium costs and yields.

Bumper, as you’ll know, is a two-sided risk market where users are free to interact with the protocol as they choose. So, while the protocol spools up and we gain traction it’s only natural that some data, especially averages, looks a little lumpier.

With positions ending less frequently, it takes longer for the protocol to show the result of rebalancing on the front end. In the case of yields, for example, the current average term is 66 days for the ETH market; so, until a user exits their position, the yield won’t be realised.

‍

Observed Yield and Premium Differences 

Ultimately, the differences can be explained by fewer BTC Takers with lower risk relative to the ETH Market. 

Although it’s not quite as simple as that as there are additional factors involved, such as the amount protected, floor levels, term lengths and frequency of the price update from the oracles. The premium in the BTC market updates around twice as frequently as the ETH market. Looking at Deribit options data, and also our own Black-Scholes calculations, the Implied Volatility (IV) differs greatly between the two assets.

We know how much our fans love some charts, so these might help illustrate what we saw.

The chart below shows the Liability of the two markets. This is the total amount of claims that can be made at any point. These are taken from mid-December till early April. The horizontal dotted lines are the averages across the whole period, but, if you only consider the second half of the plot, we can see that the Liability for ETH is much higher.

‍

Below is a plot of the incremental premium that is charged. While BTC is lower overall, for the second half of the plot it’s increasing at the same rate as ETH.

‍

And here is the Liability Risk Factor. Again, the second half shows that the risk in the BTC market is much lower than in the ETH market, which should flow through to lower premia and lower yields.

‍

All the while you can see the levels of debt are about the same for both (see below).

‍

Parameterisation

For a period after launch, BTC's 'premium scale' was set too low. It's now been adjusted, and combined with the fact that it updates twice as frequently, all other things being equal, BTC premia is expected to be higher than that of ETH. Overlay this with the term length of positions, while some may have been longer terms during this lower period, then it would permanently reduce their premium and make it show as lower.

‍

Black-Scholes Comparison

Bumper premiums are regularly compared to Black-Scholes pricing and, while we’ve shown that we are generally cheaper, deviating too far from BSM is an indicator that we might need to retune the market settings.

We’ve since put a function in place which checks this to see if premiums are drifting outside of an ideal range in our simulation, which now mirrors exactly what happens on-chain, to give us the full picture.

‍

Continual Optimisation

As we continue to move forward we expect to continually be optimising and making improvements to Bumper. The previous position size limits have also been increased to illustrate our confidence in the analysis outcomes. Our growing suite of AI tools continue to aid progress alongside our excellent engineers and the simulation. 

In summary, we’re happy to announce that the protocol has a clean bill of health. Onward and upward.

‍

‍