Pro's and Egar are definitely advocating more than 15 or 20 names. They somethjing like 40% of names(!!). Perhaps that does seem like a little much too. And in addition to favorable SV/IV ratios, Larry MacMillan, in searching for attractive straddle candidates, advocates looking for cheap volatility by looking for those options with low IV's relative to their own historical levels (say 1 yr, 2 yr, 5 yr). Don't know which is best. Anybody have any opinions?
thanks for the info , VG. If one can successfully pick "cheap" volatility candidates , why he needs to lock the position in the partial hedge and limit his potential profit ? As for the second part , unless you compare SV/IV adjusted to benchmark (VXO , VIX or whatever) those ratios won't tell you a lot.
You guys need to trade with real money: anyone who has traded equity options with real money for even one week would know the strategies described here cannot work with real money: index stock option bid/offer spreads make any arb strategy for retail wannabee traders big losers.
A naysayer without supporting data. Well let's see. Looking at the real time real money december quotes for djx, there is a spread of .05 to .10. That's $10 max. Of course you need more than one. The OEX spread is .50 ($50) and the XEO spread is .80 ($80), but you can probably split both of these. Looking at the YM the spread is 4 to 6 points. That's $20 to $30. Do you think these costs will doom the strategy? There is a concern here, but instead of the index options, you might have said something about the equity options. I'll post more on those later.
mistic , just ignore him... With new IB 75 cents per contract , total commissions are approx 0.20 % from the total and this includes all 30 longs and short on Index (i do DIA)
Vol Guy, I believe that both of McMillan's methods for finding cheap vol are valid. If possible the IV/SV ratio should be <1. The IV should also be below its 1 year historical mean. Could you post a sentence or so from the Egar reference where you are getting the number of "names" (stocks)? Maybe I missed that. I am attaching a screenshot from May 2004. It shows a beginning and very basic look at doing the strategy using the OEX. As you can see, 16 stocks would replicate 50% of the OEX, while 26 stocks would replicate 67% of the index. The cost of the straddles for the top 16 stocks are shown to the right. The value for 100 shares of the top 16 is incomplete because it ignores the delta factor.
IV Trader, Are you talking about a SV/IV ratio or just component IV compared to VIX? The question about why hedge if you can successfully pick winning straddles is a strategic question that needs to be discussed more. I think there are shades of gray. Perhaps you can share about why you have chosen to do a 100% replication of the Dow? The concerns you raise regarding risk when the index and <100% basket diverge are very real. McMillan calls this the "tracking error". That is why it would be good to also check the correlation of the components to the index as well as their volatility. I believe Egar is saying what you have understood above. It is on the ivolatility website and I now see where Vol Guy got the 40%. In addition, Egar's "How to read..." PDF talks about an Implied Index Correlation that they think is important, but I don't fully understand why yet. They say: "Implied Index Correlation defines correlation level between the actual implied volatility of the index and the implied volatility of its stock components. In other words, it is a component-averaged correlation between implied volatilities calculated from the formula of the portfolio risk ..." Then they say, "The greater Implied Index Correlation, the stronger correlation between the index implied volatility and that of its constituent stocks, and therefore the more suitable the market conditions for deploying a dispersion strategy." I'm not sure whether they are saying that a high IIC is good for the initial condition, while then the correlation can diverge, or something else. I'll try to find out.
1. if XYZ's IV was at 60 two years ago (while VIX was 30 , ratio=2) and it's current IV at 30 ( VIX at 15 , so ratio is still 2) is current IV "cheap" compare to historical ? 2. After reading all those posts I am not sure if I am doing what anybody else does ( not sure yet if its a good or a bad thing , lol) , so I don't want to confuse you. As I said before , I am selling Index straddle and buying all components straddles in certain proportion. I always was better in "reacting" rather that "predicting" (except of Vols predictions around Events) . The initial position has a slight +expectancy and the rest comes from adjustments to the long component's straddles. Earnings , pre-warnings , up/down grades and Sector rotation are the main reasons. I will look into the sources that you mentioned , maybe I can improve my existing strategy somehow , although the idea of partial hedge is not so clear yet.
Is it an advantage to use a narrow index dominated by a few companies or is it better to take a broad index? E.g. the SMI (Swiss Market Index) is dominated by 5 companies, which make 70% of the index together. If you take the biggest 10 companies, you get 90% of the index. Is it worthwhile to investigate a partial replication by the biggest 5 companies only?
Relative to the index, XYZ's IV has maintained the same ratio, so in that respect it is neither cheap nor expensive. With regard to its historical IV, it would be difficult to make an assessment based on two "snapshot" data points. Unless we know the historical period of interest and the max and min within that period, who can say? Don't worry about you are doing the same as everyone else or whether you will confuse someone. Variety is good. It looks to me like you are doing the full 100% dispersion, so that shouldn't confuse anybody since it is the standard. It also seems like you are taking in more premium than you are selling and gamma scalping the longs. What contract are you selling? There are two reasons to move toward using a partial replication of the index: 1) the cost of going long all the comps in an index can be substantial, and many of those positions will end up being losers through lack of movement. 2) Since volatility in its reversion to the mean is easier to predict than price direction, using those components whose volatility is cheap or fair and avoiding those whose IV is expensive should make the strategy more efficient and profitable. I have attached a file from the past showing an initial read of the Dow stocks. For longs I would choose among those stocks whose IV is fairly priced.