1. This is one of the best resources I've come across for implementing emergency hedges using options in a cost effective manner. Now more than ever I think Hari's wisdom can be applied to manage risk within the crypto options space especially before things get interesting...
2. As @zackvoell mentioned in this note, #BTC 180 day rolling realized vol is at nearly a 2 year low. Vol has several characteristic features across every market - one of them is the concept of mean-reversion.
3. We may not know when, but vol tends to go back to its average long-term value. Since realized vol is so low right now, I'd be risk-averse to place large short vol trades. It feels as though things have quieted down a little too much - seems a bit off.
4. Furthermore, using @GenesisVol's excellent analytics, we can see that current vol up to 180 days is below the median and 25th percentile. Again, this doesn't give me a lot of confidence for having net short vol exposures.
5. Recently we've seen several crypto funds focused on selling options for "generating yield". This strategy is often described as collecting pennies infront of a steam-roller. @nntaleb has a grim analogy for short vol strats: "A turkey has a great life until Thanksgiving day".
6. These short-vol trades end up in this type of catastrophe when unbounded tail-risks aren't hedged. I'm confident these funds have sufficient risk-management policies in place to hedge their downsides accordingly - otherwise that's just gambling.
7. Suppose a trader wants to hedge #BTC covered call tail risk. Hedging with puts is costly as these are high in demand from people wanting to protect the downside. Even with major moves down, rolling OTM puts likely won't breakeven due to the cost.
8. Instead, Hari proposes another method - selling a 1x2 put ratio (aka put backspread). As an example, this consists of selling 1x 25d put and buying 2x 10d puts. This gives us exposure to being long downside tail risk while selling expensive put skew along the way.
9. The nice thing about this strategy is that you can place this trade for minimal cost (sometimes you can even get a net credit). This prevents a trader from burning their capital with expensive hedges while staying protected from the extreme down moves.
10. Based on Hari's research, both the 10d and 25d SPX puts trade closely in line, so by selling 1x 25d put and buying 2x 10d puts, you can get a nice payoff in case things get *really* bad on the downside.
11. Crypto isn't the same market as the S&P500 so the exact delta values to choose for the trade will likely be different than 25/10. The point here is to present strategies for hedging rather than exact parameters.
12. A key drawback of this strategy is that it will lose money for "normal" down moves. A 10% move down in #BTC may not be enough for this trade to be in the green, but once things really start falling down (similar to Black Thursday), this position will kick in.
13. It generally isn't wise to hold this position to expiry due to the massive theta bleed, rather it's best to roll over to a longer dated maturity. Managing tx costs and slippage is key for this - that's why large funds need to use a service like @tradeparadigm.
14. On the other hand, if a trader wants to hedge right-tailed risk (prices going up a lot), they can enter a call backspread. Same thing here except we use calls - sell 1x 25d call and buy 2x 10d calls. This strategy will only deliver if prices *really* go up.
In summary:
- Current realized vol is very low relative to its long-term average (expect mean-reversion)
- Having open-ended tail risk is extremely dangerous in a short vol trade
- Hedging the tails cheaply allows one to be short vol but stay protected from extreme moves
Some useful links for more info on these strategies:
optionsplaybook.com/option-strateg…

optionsplaybook.com/option-strateg…

I'd be interested to learn how other players in this space think about hedging in this market. @kyled116 @fb_gravitysucks @rambo1stbld @vkik94 @JSterz

• • •

Missing some Tweet in this thread? You can try to force a refresh
 

Keep Current with Samneet Chepal

Samneet Chepal Profile picture

Stay in touch and get notified when new unrolls are available from this author!

Read all threads

This Thread may be Removed Anytime!

PDF

Twitter may remove this content at anytime! Save it as PDF for later use!

Try unrolling a thread yourself!

how to unroll video
  1. Follow @ThreadReaderApp to mention us!

  2. From a Twitter thread mention us with a keyword "unroll"
@threadreaderapp unroll

Practice here first or read more on our help page!

More from @samchepal

2 Oct
1/ Learned a lot about variance swaps by reading through @EmanuelDerman's awesome paper. This inspired me to replicate a variance swap term structure for #BTC by using options data from @DeribitExchange. Image
2/ During my research I read about the first #BTC variance swap between @GSR_io and @BlockTower which occurred in the summer of 2019. Given the lack of public data for these swaps, the only real way to get a decent price estimate is to use a replicating portfolio of options.
3/ These variance swaps allow for traders to make outright bets on volatility^2. Instead of using options (ie: straddles), with these products there is no need to delta-hedge. The payoff is as follows:

(Realized Variance - Strike Variance) x Notional
Read 6 tweets
1 Oct
1/ Spent some time exploring the market implied distribution for #BTC options trading on @DeribitExchange. This was a bit trickier than I expected but learned some interesting things along the way...
2/ I came across a closed form risk-neutral probability density (RND) solution from @EGHaug's detailed book on options pricing. I was surprised to learn that the RND is just the 2nd derivative of the option value wrt strike price. For those interested below is the formula. Image
3/ The limited number of options for #BTC Dec-25-2020 required me to linearly interpolate the IV across theoretical strikes. Instead of just 18 actual IV values, now we are able to estimate nearly 3,000+ IV data points as shown below. This will allow for a smoother RND plot. Image
Read 7 tweets
30 Sep
#BTC daily returns are not normal! After running a Gaussian kernel density estimation and comparing this to its respective normal distribution, we can see that #BTC has a lot of kurtosis as shown by the "peakedness" near the centre. Image
Also, Black Thursday and other extreme events occur much more often than a normal distribution would predict. This can be seen in the weight of the tails of the estimated distribution.
Theoretically, if the market is pricing in a normal dist the trader can make money here.

As a recap, the estimated dist has a greater probability of staying in the centre than the normal dist would predict. Also, the estimated dist has fatter tails than the normal dist.
Read 8 tweets
29 Sep
I've been thinking about the dangers of shorting long-dated options. These options cost more mostly due to their additional time value. Recall an option’s value = intrinsic value + time-value, therefore options with more time left will cost more.
Below are put prices for Oct-2020/11k vs. March-2021/11k. Despite both being OTM, March costs ~3x more than Oct. This is because the March option has 177 days vs. Oct only having 30 days left. March allows us to have optionality for 147 more days which is reflected in its price. ImageImage
It can be tempting to short these longer-dated options given their higher premium. If all goes to plan and these options expire OTM, then there is a nice premium which can be collected (or as the saying goes these days - we can "harvest" these premiums).
Read 8 tweets
29 Sep
Using vanilla options trading on @deribit, we can manually price exotic derivatives for #BTC. A digital option (aka binary option) is priced based on the probability that the spot price > strike at maturity. We should expect the price of the digital to equal option delta. Image
We can price the digital by constructing unit call spreads which have a max payoff of $1 - this constrains our range so we can focus on interpreting the probabilities. This is done by normalizing the diff between the call spread by an adjustment factor.
For very low strikes, we can see there is some notable difference between the two. This approach is an approximation so we can expect some deviation from the overall delta values. Note - I'm assuming the strike is 11k in this case.
Read 4 tweets
26 Sep
I've been looking at @deribit's #BTC option flow data. I'm always interested in exploring how the market is positioned on an aggregated greeks level. Fortunately, extracting the trade direction is possible using @deribit's API.
This analysis only looks at the previous 1000 trades which covers the past ~20 hours. Therefore, this doesn't take into consideration previous trades beforehand and is not entirely representative of market positioning. It's more like a quick glance of what is going on now.
Below is the aggregated delta exposure for the most recent 1000 #BTC option trades. This was calculated by taking the deltas of calls/puts and summing them up based on whether the option was bought or sold. Image
Read 6 tweets

Did Thread Reader help you today?

Support us! We are indie developers!


This site is made by just two indie developers on a laptop doing marketing, support and development! Read more about the story.

Become a Premium Member ($3/month or $30/year) and get exclusive features!

Become Premium

Too expensive? Make a small donation by buying us coffee ($5) or help with server cost ($10)

Donate via Paypal Become our Patreon

Thank you for your support!

Follow Us on Twitter!