AVAX - 19 March 2022
The Dopex Community Analyst series is a series of strategies, ideas, opinion pieces and educational resource written by independent contributors from the Dopex community. Every month many knowledgeable community analysts share a short analyses of a coin of their choice and share these with Dopex. The goal of these articles is to empower the community and help Dopex increase SSOV volume & deposits. With these articles we hope to provide users with additional information that will help them in making an informed choice of strategy using our products.
Without further ado let’s jump into our next article.
**Long AVAX Convexity (to improve a diversified portfolio risk-reward)
This article has been written by Luca of BWS (BWS aims to democratize access to complex hedge fund options strategies)
Token & Position Introduction
**AVAX long calls strip in order to be LONG AVAX tails ( both left and right) delta hedged with a parametric regime model.
The assumption is that you have avax ecosystem exposure in the portfolio and you want to take advantage of the volatility, especially at the extreme. Be “fearful when others are greedy, and greedy when others are fearful.” ( Warren Buffet)
Pls, see https://www.bondiwealthsecurity.xyz/analytics for whitelisted positions and relative "Beach Ready Score" to get a feel of the kind of portfolio to look for to improve risk-reward characteristics.
Dopex prices Avax option with B&S. In absence of a predictive model you should consider fair pricing ( also transaction cost is super low given Avax, thus not really relevant).
You are not trying to predict vol/play vega ( e.g. compete with GARCH models and relative) just to be in the best conditions to take advantage of smart portfolio rebalancing at portfolio level without panicking on the bottom or FOMO in at the top.
**Position & Analysis
Permanent strategy as far Avax options priced with B&S.**
**Data Exploration
N.a. not a delta strategy. Avax underlying exposure is an assumption.
**First thing first you want to understand the behavior of the time series, you're not looking to predict the price at this stage, you just look to have an idea of the distribution of the time series.
You typically want to chart some volatility-based stats/ bands ( which one does not really matter, Bollinger band, ATR, whatever estimates of deviation will give about the same info).
What you want to do here is to create some prior knowledge to see if the results will make sense.
Frequentists will call this cheating… but in Bayesian terms, you need to build your priors somewhere. The goal is make money, not show a statistically significant p-value.
Here is a possible outcome that you can observe for the stat property of the time series
(definitely not the only one and maybe not the best one).
Seems to be a period of ‘calm’ alternated with a period of ‘dynamic’ and the last ‘ turbulent’ move.
So one solution is using regime models as a signal. **
*Regime Algorithm
***There are many algorithms that can be used here. One of the simpler ones to explain while definitely not the best for out-of-sample performance are Hidden Markov models (HMM).
https://en.wikipedia.org/wiki/Hidden_Markov_model
https://towardsdatascience.com/hidden-markov-models-an-overview-98926404da0e
[https://blog.revolutionanalytics.com/2014/03/r-and-hidden-markov-models.html
](https://blog.revolutionanalytics.com/2014/03/r-and-hidden-markov-models.html)***
Hidden Markov model (HMM)
***HMM can be used as a statistical filter. Applied to a signal, it will extract regimes that are a strong indicator of the state of the market. All non-significant market variations will be ignored, so as to solely highlight an obvious change.
HMM cannot predict the future but it can give an idea of RISK.
3 different risk regimes provide insight into the state of the market.
- The first is considered ‘calm’
- The second ‘dynamic’
- The third ‘turbulent’
**Let’s see how can be applied to our Avax closing price
Volatility Regime signal ( with vol calculated on n = 120 (hours) )
**
Long story short, if you use an off-the-shelves algorithm the signal is quite erratic and you really want to do some cleaning first on the background.
Strategy Implementation
Ok, now we have a signal (even if it's not perfect), but how do we use it to help us to make money?
We're going to use the same behavior that HFT uses, in a low volatility regime we want to be aggressive delta hedging, when vol is high we want to be more passing and let our deltas run.
First thing first you need to calculate the delta ( I shared an excel sheet in https://discord.com/channels/833683475608371201/834138124337807360/947740960442421258), but let’s use Bloomberg here or
*
*
And you can do the same calculation for all strikes.
Your Initial net portfolio would be LONG 4 strikes AVAX calls options vs short AVAX-PERP at your favorite exchange ( initially delta neutral).
You will have this kind of payoff chart ( long strangle-like, even if you are just long call and short futures ). As you see in a static scenario you struggle in a market slightly up ( but lower* than your call strike), but do very well in a crash or a rally.***
* is also a function of time.
***#nrs are based on calculations for a 100M notional portfolio. Yes, we assume no limit in open interest for Avax which unfortunately is almost sold out. **#devsDoSomething
**
If you see ATM will have 2.8M gamma ( for 1% of the underlying move up I would be long avax of 2.8M~ and more short Avax of 2.8M).
Now I can use the regime model to decide whether I want to fully hedge and come back flat delta or let the delta run…. Which is the more creative bit ;) and will not share here.
***However, in a nutshell, even the concept is very similar to vanilla delta hedging you can find in this write up: https://medium.datadriveninvestor.com/how-to-delta-hedge-an-option-part-i-2efc91b24400
*****Possible Outcomes
Base Case Scenario:** you add positive convexity to the portfolio without paying too much Variance premium.
Bad Case: dull market, you lose the premium and you do not make enough money from delta hedging to pay for the option / get pinned.
Good Case: Extreme events happen, left or right side. You reap the benefit of a big move.
Disclosures
At BWS we are Implementing the strategy for users for https://www.bondiwealthsecurity.xyz/ and https://bwsquared.vercel.app/ users in collaboration with Dopex ( launching soon)
More details in https://www.bondiwealthsecurity.xyz/ docs ( coming soon)
—
Views expressed in this article are the author’s own and not reflective of the position or professional views from Dopex.io.
Dopex reimburses analyst contributors with a small payout to partially compensate the time spent on research and writing. For questions feel free to jump into our discord (discord.gg/dopex) and chat with the team or analyst contributors directly.
———
About Dopex
Dopex is a decentralized options protocol that aims to maximize liquidity, minimize losses for option writers and maximize gains for option buyers — all in a passive manner. Dopex uses option pools to allow anyone to earn a yield passively. Offering value to both option sellers and buyers by ensuring fair and optimized option prices across all strike prices and expiries. This is thanks to our own innovative and state-of-the-art option pricing model that replicates volatility smiles.
📱Stay Connected
Follow our official social media accounts and visit our website to stay up to date with everything Dopex. Twitter | Discord | Website
🚨IMPORTANT
Be careful of fake Telegram groups, Discord servers and Twitter accounts trying to impersonate Dopex. Always verify that you are on our official Discord server, Twitter page, or website.
← Back to Blog