Skip to main content
19 events
when toggle format what by license comment
Aug 16, 2018 at 1:26 comment added Hans @Quantuple: Would you like to take a look at a similar question regarding the interest rate risk premium quant.stackexchange.com/q/41286/6686?
Aug 6, 2018 at 19:59 comment added Hans @Quantuple: Thank you for the reference. I will prove the ergodicity for my setting later.
Aug 6, 2018 at 18:32 history edited Hans CC BY-SA 4.0
added 3 characters in body
Aug 6, 2018 at 6:58 comment added Quantuple Yes, I've already seen this technique used e.g. here: Dimitroff G., Szimayer A., Wagner A., Quanto Option Pricing in the Parsimonious Heston Model.
Aug 4, 2018 at 16:52 comment added Hans @Quantuple: I do have a specific form for the stochastic volatility SDE as well as the market price of volatility risk. I read your answer you linked to which, as you say, shows the specific form of the risk premium. However, as I have realized and as your answer implies, the final express is in the risk neutral measure and thus does not give a means to compute the risk premium. This has to be done in the real world measure. I am edited further my question to compute the volatility risk premium via the long time asymptotics for a mean reverting variance.
Aug 4, 2018 at 16:51 history edited Hans CC BY-SA 4.0
added 340 characters in body
Aug 3, 2018 at 8:06 comment added Quantuple Yes or edits make better sense. It's possible to do that easily if you're willing to make an assumption on the "form" of the market price of volatility risk like Heston did in his original paper. Have you had a look at this: quant.stackexchange.com/questions/37722/pricing-vix-futures/…
Aug 1, 2018 at 20:55 history edited Hans CC BY-SA 4.0
added 106 characters in body
Aug 1, 2018 at 18:35 comment added Hans @Quantuple: I now think my rationale is wrong and wrote as much at the end of in my question. But please scrutinize my argument anyway.
Aug 1, 2018 at 18:33 history edited Hans CC BY-SA 4.0
added 205 characters in body
Aug 1, 2018 at 17:22 comment added Hans @Quantuple: Your understanding is mostly correct but with some difference to my intended argument. I have edited my question to let it stand on a more specific ground. Please review. Thank you.
Aug 1, 2018 at 17:22 history edited Hans CC BY-SA 4.0
Put the question in specific terms.
Aug 1, 2018 at 17:14 history edited Hans CC BY-SA 4.0
Put the question in specific terms.
Aug 1, 2018 at 12:50 comment added Quantuple So assuming $VIX^2$ represents the fair strike of a fresh-start, idealised (i.e. continuous returns sampling) variance swap maturing in 30 days then indeed: $VIX^2 = \frac{1}{T} \Bbb{E}^\Bbb{Q} \left[ \int_0^T d\langle \ln S \rangle_t \right]$ with $T=30/DCC$ years. And the idea would be to write that: $VRP = VIX^2 - \frac{1}{T} \Bbb{E}^P \left[ \int_0^T d\langle \ln S \rangle_t \right] = \frac{1}{T} \int_0^T \left( \Bbb{E}^Q[v_t] - \Bbb{E}^P[v_t] \right) dt$ assuming that $d\ln S_t = \cdot dt + v_t dW_t$. Did I understand that well?
Aug 1, 2018 at 7:55 comment added Quantuple It's not entirely clear to me what you're trying to achieve. What is the variance process $(v_t)_{t \geq 0}$ which you are talking about? Is this in the framework of a particular model? What do you define as the variance risk premium?
Aug 1, 2018 at 6:02 comment added Hans @Nitin: I agree with your last sentence. But what do you mean by the question mark following your "no"? Also, what is your conclusion? Do you agree with my rationale as stated in my last sentence that my setup is correct?
Aug 1, 2018 at 3:42 history edited Hans CC BY-SA 4.0
added 8 characters in body
Aug 1, 2018 at 2:52 comment added user217285 "My concern is that VIX is a traded asset" -- no? It's given by a formula that is a function of traded asset prices, but it is not traded itself.
Aug 1, 2018 at 1:26 history asked Hans CC BY-SA 4.0