hey all, anyone here ever tried (or is already using) a local volatility model? I am trying to implement it with dupire's formula, following the method in this webisite: https://financetrainingcourse.com/e...cal-volatility-surfaces-in-excel-final-steps/ however after I run the calculation I get "impossible" vols, for one strike for instance I get a value of 22%, one strike after that it jumps to 259%, then goes back to 23%. Any advice? Is there something inherently wrong in the method the website is proposing? Are there known limitations in this model? Does it work only for specific options classes? And the second question would be.. let's admit that the model works, one thing that I can't wrap my head around is this: local vol is supposed to be the vol that you get when spot goes to a certain strike..correct? but how does the model account for the timing of the move? I expect that the vol would be different if the strike was reached tomorrow vs one month from now. Can someone explain to me the concept in layman terms?
You first should check whether your pricing data you input into the model are correct? Sometimes bid and offers are stale and don't represent actual supply and demand
Limited markets mean lumpy vol -- take an approximation. I do something similar, and calculate local spreads a $5, $10, $15, and $20, and can adjudge the accuracy of any of those 4 by whether it's in line with the other three. It's what you gotta do. I have not heard that anywhere, and your follow-up question would be my first question, too. IV seeks to give us the impossible Future Volatility -- so looking forward, it's where the market *thinks* that some event might go, or stairstep to, or plummet from. If the market had gotten there already (i.e., "spot"), then the vol portion of the pricing has a completely different horizon at which to aim. (And so, "When?!?!?" comes to mind...) FWIW, I think of local vol as event-driven, closer to the market price, and market-driven (spreads, OI-exits, mechanical stuff, etc), further away. (And so, I'm not surprised by any crazy shit I see out in the sticks...)
look at here,https://birga-trade.com/zhizhilev.html this is specifically a mathematical model - https://birga-trade.com/zhizhilev52.html this author is much more believable unlike others, in his calculations on modeling and volatility there is only one error, and in your version there are at least 3 errors The problem with any mathematical model is that the authors use the phrase “in the form of a random process forming filter” or “white noise”, in fact there is no place for randomness in the market, everything is clear and regular, just no one can find this secondary pattern
Setting aside the hilarity of recommending, without context, a Russian book on an English-speaking forum... a quick scan of this thing produces (among the endless boring generalities and circular "logic") such howlers as "the optimal scheme of investment requires an analysis of the entirety of the financial market which produces a single solution such that no other solution produces a better result." The rest of it is at least this turgid and obfuscatory. (It also seems that the Kalman filter is the one and only sure method of producing this solution. I guess a single pass through "the entirety of the financial market" is just too quick for words, so we need to apply recursive estimation just so it doesn't blow right by us...) I'll note that it wouldn't surprise me to find out that the average Russian investor is impressed by this thing; their level of sophistication may be easily gleaned from the ads scattered around the above site. E.g.: "For 500 rubles, you could Eat fast food Or place it in an option and head for the best restaurant in town <button>INVEST</button>"
I’m not interested in your opinion, the person asked, I answered, earlier I saw your post on my blog, and on this basis I can conclude that you are a sucker
Did you think that I was talking to you? How cute. How long have you had the delusion that the world revolves around you, and have you looked for professional help for that? Clue: I was laughing at you, as are most other people. Perhaps we shouldn't laugh at anyone with this amount of obvious damage, but your random nonsense here is just too droll.
I will give you one clue the same math problem can be solved in three ways 2 + 2/2= 3 (2 + 2) / 2= 2 2+ (2/2)= 3 If you want to calculate market volatility Find the only reason for this volatility that mathematicians stubbornly do not see and you will get confirmation automatically And then it will be easy to calculate this volatility at any time interval, for example - like this #35 Today at 10:23 AM PS And do not ask for advice from these morons, it is better to seek the answer yourself
I am too dumb, what is local volatility? If I have a set of volatility surfaces, why can't I just do a simple interpolation to get what I need?
Resto, It is clear to all reasonable readers on here that your 'answer' was totally not useful, if not outright confusing... but this has been your modus operandi all along....