Option Question about Delta

An approximation is an approximation and if you know where it starts breaking, it's good enough. For example, I would frequently use spot swap rates weighted by duration to find the forward swap rate - I know it's an approximation that might be off by huge amounts on a steep curve, but in most non-trading cases (e.g. talking to the management) it's good enough. Using delta as probability (and there are a lot of cases where you need an idea of risk neutral probabilities, e.g. in skew analysis) is about as good.

Anyhoo, last question and I will drop the subject.

When you look at the skew, do you think it's "reasonable" to use X-delta risk reversals as a measure (or, my favorite, 25d RR/50d)?

 

MAW, with all due respect:
a) the various points you make are all reasonably well-known;
b) their validity doesn't change the fact that we're discussing an imperfect heuristic, rather than a mathematical proposition;
c) since I know sle quite well, I know he IS in fact an extremely experienced exotics bookrunner (rates, as well as equities), so you are barking up a completely wrong tree here;
d) you're in very poor company.

My point here is that you're not really arguing with the statement made. I don't think anyone ever disputed the fact that delta is NOT everywhere and always the same as probability. The reasons and the math behind this are all reasonably well-known.

 

You may be right Martin.
Of course those various points are all well-known, that's what makes me get started.
Anyway, My bad.

 

To get a ‘pure’ linear exposure to the skew, a simple ( var swap – gamma swap ) spread would be « reasonable » enough.
You then just need to be focused on the relative replicating portfolio values. It’s straighforward to price.

 

Hmm. Actually, I would say that a fixed strike risk reversal is a cleaner skew position for a variety of reasons (to start, a vs-atm or vs-gs would actually start gaining vol exposure if the market moves upward enough). Anyway, THIS is an interesting question that I'd be happy to argue about.

 

No worries... It's just that I think the heuristic/rule of thumb, imperfect as it may be, IS a lot more worthy of a discussion than the various mathematical facts, which are all, as you have stated yourself, neither here nor there. And I mean discussing the rule of thumb, as well as its various imperfections and flaws.

 

This is a "reasonable" measure.

 

This forum has the most interesting discussions.

BTW - which Starbucks are we meeting at this week?

 

Let's try Peet's this week.

 

Us guys with little accounts still use McDonalds. They do have wifi. Set the date and time.