Floating Collateral Requirments based on Realized (or Implied) Volatility

First post; my view, coming from a futures/options trading background, is that certain mAssets (such as SPY, AMZN) are highly overcollateralized while others (GME, AMC, VXY) are arguably undercollateralized.

Would it be possible to calculate a rolling 30-day Realized Volatility from the oracle feed, or, ideally, the 30-day expiry ATM implied volatility (not sure of data source) to generate a floating collateral requirement unique to each mAsset’s varying risk profile? This would be a route to dramatically improve capital efficiency in the ecosystem in a manner that I think mAsset buyers may not object to, since it would help convergence.

I’m interested in what others think about this, but I believe it’s reasonable that assets like SPY & AMZN would actually be closer to a 25% minimum rather than current 130-150%. Meanwhile based on realized & historical vol in recent weeks GME capital requirement maybe should have been higher than where it is currently set. This is intuitive to me but I am interested in other perspectives.

Would a proposal like this be feasible/any support?

This is reasonable. At the moment I don’t think there are any standards on determining collateral requirements and they are fixed. I like your idea. Could you expand on how you are deriving 25% for SPY & AMZN? What frequency would be good for the changes to become operative (your talking about real time which might be the right answer)? How would this work in a “circuit breaker” type event or black swan where the “tighter” floating value becomes unsafe?

Imagine 800% like on Shadows.

The options premiums (at the money straddles) are a great place to derive real-time market risk as they tend to respond quickly to changing conditions. SPY and AMZN at the moment have very low implied annualized volatility of around 25% for AMZN, and (currently) below 20% for near-term SPY. Keep in mind that collateral needs to cover a black swan upside move only (until inverse mAssets arrive).

Of course it’s up to everyone how much of a buffer is desired, but in my view the current 200% requirement for VIXY is riskier than a theoretical 20% requirement for SPY. Currently the implied volatility on VIXY is around 105% across the term structure. In any case, most would agree it’s much more likely that VIXY suddenly goes to $30 than that SPY suddenly goes to $500.

An option would be to take a running average of annualized realized volatility over 30 periods @ 1-minute intervals, 5-minute, 30-minute and daily, and then taking the average of those values to get a baseline, and then adding a buffer multiplier if desired.

Would need significant engagement from developers if something like this would gain any traction.

Not to overload the thread but allowing for auto-burn or swap-buy & burn as an alternative to liquidators showing up to buy debt could make tighter collateral viable. Swap-buying should, of course, only happen if the swap premium is less than than the 20% liquidation penalty…

The penalty amount should also probably be a floating value using a similar metric (real-time vol) to reward liquidators variably based on how dangerous the debt is. Who is incentivized to buy GME debt @a 20% premium if the volatility is @ 1000% (which is what it approached during previous run-ups)?