Very good point. If you were to delta hedge, you would capture the exact difference between implied and realized volatility. It is the cleanest way to get a pure volatility exposure.
As our approach is extremely path dependent, it is hard to know in advance if you would have made more money by delta hedging or not. In some cases the underlying could drift from one boundary to the next and end up right in the middle - you would have lost a lot of money delta hedging.
Now if it ended right by one of the straddle boundaries, delta hedging would have make you keep a lot of money.
About delta hedging and its purported shortcomings excluding expenses, wouldn't it be fairly easy to write a computer program that would hedge automatically? In turn reducing the time and effort needed to do this, possibly even making it more error-prone?
I’m sure there are layers and I also don’t have the overview to be smart about this, but I’d still like you to expand on the second part.
Rebalancing options? What I had in mind was buying/selling (with a certain frequency) an appropriate amount of the underlying so that the options+equity bundle has delta equal to zero. Is there more to it?
The fallacy about delta hedging is that delta is the only thing to hedge. If you have to rebalance your deltas, volatility itself has likely moved, and you will have to rebalance it, too, to keep the desired exposure. For instance, if the market has moved 3% and you were trading options at the money, hedging the delta is a good start, but your vega profile is significantly different: Do you still want these options? Are you going to switch them for another position instead?
I would definitely not outsource the task of managing the options part to an algorithm. And if there is space for something manual in the process, why not keep the flexibility of doing everything manually?
Because when I say delta hedging is prone to error, it is not because you could buy instead of selling or make a mistake in the number of shares (although these things will also happen). But more because we often end up doing it, disregarding the entire book as a whole, and making suboptimal decisions in the process.
Now you could indeed make all the options trades you need and then have an algo hedging your delta no matter what, at the end of the day. But my point is that trading options is much more like playing chess than an efficient process you would perfectly automate, unfortunately.
Thanks for sharing the note. You mentioned that delta hedging might make selling vol on stuff like SPY/QQQ/LQD/XLU profitable too. Sure, it will require some more intervention and favorable trading costs. But would it be possible for you to elucidate on some approaches that can be explored for such delta hedged VRP trading?
Hi there - thank you for reaching out. The VRP is just an approach to identify setups where implied volatility is too expensive. If you decide to enter the trade AND delta hedge it, the approach wouldn't be any different than any other delta hedging methods. You would need to sell/buy some of the underlying to make sure you stay neutral. To minimize the cost, you could do that once a day - but it would really depend on the number of day to expiry.
I guess my real question was - for the identified VRP tickers, if we were to delta hedge them (instead of holding them for 7/14/23 DTE), how would the expectancy and edge change? Is there more edge in keeping such positions delta neutral via re-balancing?
Very good point. If you were to delta hedge, you would capture the exact difference between implied and realized volatility. It is the cleanest way to get a pure volatility exposure.
As our approach is extremely path dependent, it is hard to know in advance if you would have made more money by delta hedging or not. In some cases the underlying could drift from one boundary to the next and end up right in the middle - you would have lost a lot of money delta hedging.
Now if it ended right by one of the straddle boundaries, delta hedging would have make you keep a lot of money.
About delta hedging and its purported shortcomings excluding expenses, wouldn't it be fairly easy to write a computer program that would hedge automatically? In turn reducing the time and effort needed to do this, possibly even making it more error-prone?
If it was easy, everybody could do it ;)
In theory delta hedging is simple. In practice it often involves rebalancing certain options and this operation cannot easily be automated.
I’m sure there are layers and I also don’t have the overview to be smart about this, but I’d still like you to expand on the second part.
Rebalancing options? What I had in mind was buying/selling (with a certain frequency) an appropriate amount of the underlying so that the options+equity bundle has delta equal to zero. Is there more to it?
The fallacy about delta hedging is that delta is the only thing to hedge. If you have to rebalance your deltas, volatility itself has likely moved, and you will have to rebalance it, too, to keep the desired exposure. For instance, if the market has moved 3% and you were trading options at the money, hedging the delta is a good start, but your vega profile is significantly different: Do you still want these options? Are you going to switch them for another position instead?
I would definitely not outsource the task of managing the options part to an algorithm. And if there is space for something manual in the process, why not keep the flexibility of doing everything manually?
Because when I say delta hedging is prone to error, it is not because you could buy instead of selling or make a mistake in the number of shares (although these things will also happen). But more because we often end up doing it, disregarding the entire book as a whole, and making suboptimal decisions in the process.
Now you could indeed make all the options trades you need and then have an algo hedging your delta no matter what, at the end of the day. But my point is that trading options is much more like playing chess than an efficient process you would perfectly automate, unfortunately.
Ok, get it now, thanks!
Thanks for sharing the note. You mentioned that delta hedging might make selling vol on stuff like SPY/QQQ/LQD/XLU profitable too. Sure, it will require some more intervention and favorable trading costs. But would it be possible for you to elucidate on some approaches that can be explored for such delta hedged VRP trading?
Thanks.
Hi there - thank you for reaching out. The VRP is just an approach to identify setups where implied volatility is too expensive. If you decide to enter the trade AND delta hedge it, the approach wouldn't be any different than any other delta hedging methods. You would need to sell/buy some of the underlying to make sure you stay neutral. To minimize the cost, you could do that once a day - but it would really depend on the number of day to expiry.
Thanks for your response.
I guess my real question was - for the identified VRP tickers, if we were to delta hedge them (instead of holding them for 7/14/23 DTE), how would the expectancy and edge change? Is there more edge in keeping such positions delta neutral via re-balancing?