IVolatility Egar Service

alassio · Nov 25, 2005

Quote from mysticman:

Go back and read your post where you said outliers were killing you, or maybe that was IV_Trader. If you are short an ATM component straddle and that stock makes a 2 std. dev. move while the index stays the same, that is a big loser for your reverse dispersion....

I wonder whether the variant trying to profit from component/index call skew differences behaves the same in this szenario: If you sell OTM component calls (not ATM straddles) and hedge with OTM index calls where the IV differences are big enough to show significant skew differences the total position should be for a credit. The hope is to keep part of that credit in most szenarios.
I assumed the worst case a strong move of all components against you, where you are properly hedged with the index calls. But maybe the worst case szenario instead is a single component with a strong rally and the index mostly unchanged. However, in this case the premiums received from the other components should still mostly cover the loss from the runaway stock.
It can of course also make sense to use credit spreads instead of naked shorts or butterflies instead of short straddles for additional protection/gamma in case of strong moves.
Has anybody already simulated such szenarios?

Profitaker · Nov 25, 2005

Quote from alassio:

Has anybody already simulated such szenarios?

Yes, I'm running a live model started last Friday (18th Nov)...and this was a snapshot of the portfolio 5 minutes ago.
<TABLE border=2>

<TR><TD> Stock 18 Nov </TD> <TD> 25-Nov </TD> <TD> % Chg </TD> <TD> Strike </TD> <TD> Qty </TD> <TD> Prem </TD> <TD> Value Â£ </TD> <TD> 25-Nov </TD> <TD> P / L </td></TR>
<TR><TD> BP 644 </TD> <TD> 658.5 </TD> <TD> 2.3% </TD> <TD> 650 </TD> <TD> -16 </TD> <TD> 14 </TD> <TD> 2240 </TD> <TD> 19 </TD> <TD> -800 </td></TR>
<TR><TD> Shell "A" 1814 </TD> <TD> 1833 </TD> <TD> 1.0% </TD> <TD> 1800 </TD> <TD> -2 </TD> <TD> 50 </TD> <TD> 1000 </TD> <TD> 58 </TD> <TD> -160 </td></TR>
<TR><TD> RBS 1683 </TD> <TD> 1690 </TD> <TD> 0.4% </TD> <TD> 1700 </TD> <TD> -12 </TD> <TD> 27.5 </TD> <TD> 3300 </TD> <TD> 26 </TD> <TD> 180 </td></TR>
<TR><TD> Barc 612 </TD> <TD> 604 </TD> <TD> -1.3% </TD> <TD> 600 </TD> <TD> -12 </TD> <TD> 23 </TD> <TD> 2760 </TD> <TD> 16.5 </TD> <TD> 780 </td></TR>
<TR><TD> Glaxo 1497 </TD> <TD> 1441 </TD> <TD> -3.7% </TD> <TD> 1500 </TD> <TD> -4 </TD> <TD> 35 </TD> <TD> 1400 </TD> <TD> 10.5 </TD> <TD> 980 </td></TR>
<TR><TD> Astraz 2615 </TD> <TD> 2662 </TD> <TD> 1.8% </TD> <TD> 2600 </TD> <TD> -4 </TD> <TD> 78 </TD> <TD> 3120 </TD> <TD> 136.5 </TD> <TD> -2340 </td></TR>
<TR><TD> Vod 128 </TD> <TD> 128 </TD> <TD> 0.0% </TD> <TD> 120 </TD> <TD> -20 </TD> <TD> 8.5 </TD> <TD> 1700 </TD> <TD> 9 </TD> <TD> -100 </td></TR>
<TR><TD> UKX 5499 </TD> <TD> 5520.80 </TD> <TD> 0.4% </TD> <TD> 5525 </TD> <TD> 12 </TD> <TD> 55.5 </TD> <TD> -6660 </TD> <TD> 54.5 </TD> <TD> -120 </td></TR>
<TR><TD> ...... </TD> <TD> ...... </TD> <TD> ...... </TD> <TD> ...... </TD> <TD> ...... </TD> <TD> Total </TD> <TD> Â£8860 </TD> <TD> Total </TD> <TD> Â£-1580 </td></TR>
</TABLE>

All Call options, Dec05 expiry. The ratios probably need tweeking further, but this is the basis of what I hope to run in December.

Mystic, get back to you soon.

Profitaker · Nov 25, 2005

Duplicate

mysticman · Nov 25, 2005

Quote from alassio:
I wonder whether the variant trying to profit from component/index call skew differences behaves the same in this szenario: If you sell OTM component calls (not ATM straddles) and hedge with OTM index calls where the IV differences are big enough to show significant skew differences the total position should be for a credit. The hope is to keep part of that credit in most szenarios.
I assumed the worst case a strong move of all components against you, where you are properly hedged with the index calls. But maybe the worst case szenario instead is a single component with a strong rally and the index mostly unchanged. However, in this case the premiums received from the other components should still mostly cover the loss from the runaway stock.
Has anybody already simulated such szenarios?

Because the difference between the component IV and the index IV is substantial, these ideas are worth taking a look at. Is the reverse dispersion really viable? What are the worst case scenarios that have occured recently and what are their causes? Perhaps a portfolio of short component IV would spproximate index IV such that no index hedge would be needed? These are questions that need to be answered. Some of us will be modeling dispersion in Excel at another site.

mysticman · Nov 26, 2005

With regard to the reverse dispersion strategy, one option authority says that it is generally not profitable to buy the index put options and sell the stock put options because of the volatility skew since 1987:

"The false argument goes like this: since stock options trade for higher implied volatilities than index options, one can sell the stock options and buy the index options (in the proper ratio) and profit. Do you see the fallacy in this argument? It misses the point that the combination of stock options behaves just like the index.
That is, even though the stock options individually have higher implied volatilities, the combined effect of selling all of them reduces the net position to the equivalent of the index position. Hence there is no âedgeâ unless the index options happen to be cheaper â an unlikely occurrence as far as puts are concerned.
On a rare occasion when the volatility skew in index options is especially cheap, it might be the case that
out-of-the-money calls are cheaper than their stock option counterparts and therefore the index calls could be bought and the equity calls sold. Even so, the reason such a trade is possibly feasible is not that the stock options individually have higher implied volatilities than the index options do."
--- McMillan:eek:

dantes · Nov 26, 2005

Quote from Profitaker:

Yes, I'm running a live model started last Friday (18th Nov)...and this was a snapshot of the portfolio 5 minutes ago.
<TABLE border=2>

<TR><TD> Stock 18 Nov </TD> <TD> 25-Nov </TD> <TD> % Chg </TD> <TD> Strike </TD> <TD> Qty </TD> <TD> Prem </TD> <TD> Value Â£ </TD> <TD> 25-Nov </TD> <TD> P / L </td></TR>
<TR><TD> BP 644 </TD> <TD> 658.5 </TD> <TD> 2.3% </TD> <TD> 650 </TD> <TD> -16 </TD> <TD> 14 </TD> <TD> 2240 </TD> <TD> 19 </TD> <TD> -800 </td></TR>
<TR><TD> Shell "A" 1814 </TD> <TD> 1833 </TD> <TD> 1.0% </TD> <TD> 1800 </TD> <TD> -2 </TD> <TD> 50 </TD> <TD> 1000 </TD> <TD> 58 </TD> <TD> -160 </td></TR>
<TR><TD> RBS 1683 </TD> <TD> 1690 </TD> <TD> 0.4% </TD> <TD> 1700 </TD> <TD> -12 </TD> <TD> 27.5 </TD> <TD> 3300 </TD> <TD> 26 </TD> <TD> 180 </td></TR>
<TR><TD> Barc 612 </TD> <TD> 604 </TD> <TD> -1.3% </TD> <TD> 600 </TD> <TD> -12 </TD> <TD> 23 </TD> <TD> 2760 </TD> <TD> 16.5 </TD> <TD> 780 </td></TR>
<TR><TD> Glaxo 1497 </TD> <TD> 1441 </TD> <TD> -3.7% </TD> <TD> 1500 </TD> <TD> -4 </TD> <TD> 35 </TD> <TD> 1400 </TD> <TD> 10.5 </TD> <TD> 980 </td></TR>
<TR><TD> Astraz 2615 </TD> <TD> 2662 </TD> <TD> 1.8% </TD> <TD> 2600 </TD> <TD> -4 </TD> <TD> 78 </TD> <TD> 3120 </TD> <TD> 136.5 </TD> <TD> -2340 </td></TR>
<TR><TD> Vod 128 </TD> <TD> 128 </TD> <TD> 0.0% </TD> <TD> 120 </TD> <TD> -20 </TD> <TD> 8.5 </TD> <TD> 1700 </TD> <TD> 9 </TD> <TD> -100 </td></TR>
<TR><TD> UKX 5499 </TD> <TD> 5520.80 </TD> <TD> 0.4% </TD> <TD> 5525 </TD> <TD> 12 </TD> <TD> 55.5 </TD> <TD> -6660 </TD> <TD> 54.5 </TD> <TD> -120 </td></TR>
<TR><TD> ...... </TD> <TD> ...... </TD> <TD> ...... </TD> <TD> ...... </TD> <TD> ...... </TD> <TD> Total </TD> <TD> Â£8860 </TD> <TD> Total </TD> <TD> Â£-1580 </td></TR>
</TABLE>

All Call options, Dec05 expiry. The ratios probably need tweeking further, but this is the basis of what I hope to run in December.

Mystic, get back to you soon.

I would think you need to at least hedge the individual deltas if you looking to be short the dispersion. But maybe I am missing your intention here?

Profitaker · Nov 26, 2005

Quote from dantes:

I would think you need to at least hedge the individual deltas if you looking to be short the dispersion. But maybe I am missing your intention here?

They are hedged, UKX is the component index, of which I'm long.

dantes · Nov 26, 2005

Yes I see that you are long the index option but this means that you have exposure to correlation between your basket of stocks and the index. I would think you were looking to only have exposure to the correlation between vols?

mysticman · Nov 26, 2005

Quote from dantes:
Yes I see that you are long the index option but this means that you have exposure to correlation between your basket of stocks and the index.

You asked whether he was hedged and he is. But the fact he is hedged does not mean he is has correlation risk. It is the partial replication that gives him that. But what other alternative would you propose?

I would think you were looking to only have exposure to the correlation between vols?

I'm not sure what you are talking about here (what "correlation between vols" means) or how you would go about achieving that. As I said before there is a problem with the use of this term so we must clearly state its meaning when we use it. Could you say more?

dantes · Nov 26, 2005

The whole point of a dispersion trade is to capture the difference in vols between the components and the index. If you go short the component volatility and long the index volatility you are hoping that your gamma on the index will make you more money then what you loose on the components. This happens when the realized correlation between the individual stocks are high. If the correlation between indivdual stocks are low (they move all around the place) then the index realized volatility will be low (because on average things have stood still) and you will not gain anything on your index gamma but loose big on your short components gammas.

However if you don't delta hedge each option position in the dispersion there is not really any point of talking about gamma effects. You have much bigger risks to worry about.

I am sure this is an interesting trade but it is not a dispersion as the term normally is used.