What are
the ways to bet on volatility rise in S&P 500? You may first think of put
buying or VIX products buying. The problem with these positions is they are
expensive. Everyone is buying puts just in case, so they are overpriced,
sometimes tens of times from the model. VIX products are not cheap either,
having average trend against long. Even direct and straight way – index shorting
– will play against you, because on low volatility market growth is strongest.
There is
another way – far OTM calls selling. This approach has several advantages:
1. Option
time decay works on you.
2. When
volatility is low, calls are usually overbought (at least comparing to the Cognitum Option Pricer model), selling them is mathematically justified.
3. Index
behaviour on market growth is fairly predictable - there are trend lines which
work as barriers.
I would
like to illustrate the last point. Let’s take a look at the current state of S&P 500:
Blue
vertical lines are expiration dates of SPXpm contracts.
There are
two trend lines from different timeframes, they "limit" the index growth. The
probability of a strong growth above those lines is fairly low. So selling
calls with strikes 10-20 higher those lines can be safe enough. It is better to
sell it when market approachs the lines and stabilizes in slow growth on low
volatility.
For more
cautious traders who is uncomfortable having naked sells I can recommend
hedging with buying far calls on contract 1-2 month later, same position size. In
case of a strong growth profits from far expiration will compensate losses from
closer expiration, and time decay will be your profit.
If your
broker requires a lot of margin to sell options, you may try to sell vertical
spreads instead, though this position will be less effective and transaction costs will be higher.
I am really surprised that you found calls are undervalued. Using a similar an option model similar to your product, I found puts to be overvalued almost always and calls to be undervalued during low volatility regimes. The undervalue-ness of calls are even more persistent for calls expiring more than 1 year away.
ReplyDeleteIt is possible to have different estimates for different normalization window lengths. I use one I've tested to be ok for contract durations of my interest.
Delete