The problem with short gamma

I would say this: volatility is easy to define, but there are numerous different definitions that can lead to confusion. In the context of what you're talking about, you *have* to define volatility on the basis of how often you delta hedge.

If you put on a straddle and *do not delta hedge* until expiration, but manage to lose money despite low realized volatility... then the model is not broken, but your understanding of it is.

If realized volatility is lower than IV as calculated using end-of-day prices, then you would have made money regardless of final price if (and only if) you had hedged on an end-of-day basis. If you do not hedge on a end-of-day basis, then what Bloomberg or other sites/services pass around as "realized volatility" is irrelevant to you.

The realized volatility of *your portfolio* = sqrt(log return from day 0 to day T). And if this number is larger than the IV of your original position, you will lose money.
 
1) I agree with Martinghoul in that there are clear definitions of volatility which most participants can agree on.
2) most index options are purely electronically traded, no pit involvement
3) I also dont subscribe to any sort of "manipulation scheme". The market as an aggregate dictates price action but liquid indexes and their futures are simply too large to be manipulated by single entities. I am slowly getting your whole point is to steer others towards this manipulation argument, be it only in single stock options.
4) Your first point in this thread was your complaint that BS assumes zero auto correlation. Guess what BS suffers much larger deficiencies from some other assumptions. So what, everyone knows it and treats it accordingly.
5) It does not matter what UUUU or UDUD means to you. After UUUU the next sequence may be UDUD. The question you need to ask yourself is what you gonna do right now about your position. I suggested it does not pay to look at the past, it pays to understand what the position of the market currently is and depending on that whether those who hold large options positions have a motivation to hedge more or less frequency and at what levels. This is often very unclear and muddy but sometimes it can make a huge difference in your own hedging and price assessment.

6) According to you, "My point and beautifully exemplified by Carr is that depending on your position and how you hedge, you see different volatility, and you are back to trading the underlying. This is true even in model-free scenarios and has nothing to do with stochastic vols."

-> you have it the other way around. How you see assess volatility determines your hedges. And this heavily depends on your models, in effect you dont trade the underlying you trade your model. So we see things different even from the start.

In thus I dont see a point to digress further unless we can agree that hedges are a function of the assumed volatility and not the other way around. Again I think your whole point is the manipulation scheme and I see where you try to get to but I simply disagree...



Quote from nitro:

Index options are a different animal. For one, many of those are pit traded, and you can see the flow coming into the pit and calibrate accordingly. Very difficult for a single firm to control the underlying as well (although in this light volume environment it may be easier than I think). This is probably your experience and it makes sense that your point of view is such. I also have experience with MMing index options (SPX). Single equity options are what I am talking about, or even index options (ETFs actually) if you don't have access to option flow, i.e., you are in a pit.

But even with that explanation, you are keying on my "complaint". READ WHAT PETER CARR SAYS AND EITHER YOU AGREE WITH IT OR NOT! Forget about what I am saying. First let's see if you agree with that or not. Then we can move to my "complaint".
More...
 
Options are four-dimensional--deltas are influenced by volatility, time and gamma. That said, one can reverse gamma scalp using short straddles by maintaining a delta neutral to delta bias (in the direction of the trend). I have been all over this, and let's face it, directionality cannot be ignored. Let's assume that the only strategy in your arsenal is the short straddle. Let's say you know that the underlying will stay within the break-even area (ATM strikes +-total premium received). Do you make any adjustments? Theoretically, no. You know this trade will be a winner, but you do not know how much profit you will make. McMillan suggests an aggressive treatment here: Buy back the winning short option if it has lost 75-80% of its value. Then, any rebound will allow the other short option to be more profitable. Now, suppose you know the underlying will be unidirectional and will exceed your break-even points, but you do not know how far the underlying will go. One option (pardon the pun) is to not enter a short straddle-just skip this period. Two, actually go long or short, either using the appropriate option or the underlying--changing your strategy. Or, three figure out a way to make money with your short straddle. This may include exiting the short option with any bounce or pull-back and allowing the other option to run its course. Two, you can cover the losing option before it gets to the break-even point. Altho the ITM short loses money, the other short and the underlying make up for the loss. Three, you can either roll up/down the winners. Lastly, create an iron butterfly instead of a short straddle. Even though you are paying an inflated price for the insurance, it is worth it. Why? What if volatility jumps? The long put's value will jump and be an adequate hedge along with the short call. If volatility drops, then your short options lose their value quicker. You can even adjust the IB in ways similar to the short straddle.

It is foolish to just trade one dimension (ie volatility). Directionality plays a huge part and must also be considered. Options are four-dimensional. Two, you can't just hold these trades to expiration. If you have a decent profit in the first three days, exit the trade and look for another one--there is no buy and hold in options either. Same as the underlying--you may buy and hold for days and at most weeks. Anyway, just my $.02, for the tuition I have paid over the past six years was much higher.
 
There has been some good comments. Thanks.

I am wondering, asiaprop, martinhoul, heech, jwcapital, which of you are primarily traders? Do you derive your income primarily or only from trading, or are you guys programmers, analysts, quants, etc in your firm?
 
Quote from nitro:

There has been some good comments. Thanks.

I am wondering, asiaprop, martinhoul, heech, jwcapital, which of you are primarily traders? Do you derive your income primarily or only from trading, or are you guys programmers, analysts, quants, etc in your firm?
More...
All of the above?

I just launched my CPO earlier this year. It consists of a black box trading application I developed/programmed, on the basis of my own model/view of volatility.
 
I am no options expert. There are guys on here that understand greeks way way way more than I do.

Personally, I have never found greeks useful. I use verticle credit spreads and sometimes ratio credit spreads to trade the futures market.

To me the greeks are more of a way for mathematicians to understand what is going on and try to find 'optimal' methodolagies. In my extremely humble opinion they have no causal relationship to the option's behavior. Please correct me if I am wrong, just that in my trading style I don't feel like I have suffered from ignoring the greeks.
 
You must log in or register to reply here.
Back
Top