Options with 2 underlyings

[Tsotne]

Thanks Andy I'll try it.

 

[financeguy]

Because the dealers will only send those runs to trading desks and counter-party they are doing business with. Normally, you ask someone to put you on the list and they will send you via Bloomberg emails with a no copy, no forward lock so another party can't receive them.
If you want to get data for your study, the best way is to get something that is on the exchange.

And to make matters worse, there aren't even runs for multi-asset products generally. Those usually exist for generic vanillas, which even so as Andy points out would be impossible for you to discover a track record for. Brokers will generally keep price history for first generation exotic options that trade day to day as well - but these aren't standardized or anything. The only way for a price history to exist on a multi-asset product is if you were a real client who asked for a real or level price on it every day for a period of time, which no one ever does anyway.

 

[Tsotne]

And to make matters worse, there aren't even runs for multi-asset products generally. Those usually exist for generic vanillas, which even so as Andy points out would be impossible for you to discover a track record for. Brokers will generally keep price history for first generation exotic options that trade day to day as well - but these aren't standardized or anything. The only way for a price history to exist on a multi-asset product is if you were a real client who asked for a real or level price on it every day for a period of time, which no one ever does anyway.

So I cannot get the date for multi-underlying options? It doesn't matter whatever the underlying is. Agricultural products, equity(preferably), etc. anything...Even hypothetical, just for the educational purpose. One example needed to check the theory written...

 

[Tsotne]

The only way for a price history to exist on a multi-asset product is if you were a real client who asked for a real or level price on it every day for a period of time, which no one ever does anyway.

So how is the historical prices recorded then? It should exists somewhere. I'm going to register on Bloomberg and hope I'll get something to cling at least.

 

[Tsotne]

Just got reply from the Bloomberg team saying the following:

LOG: 4/27/11 16:17:07 Bloomberg Feedback Team
Hello: We are sorry but web does not provide such data. Bloomberg terminal does
and you would need access through your university if they lease. We provide no
other research service. Sincerely, Bloomberg Website Feedback Team

I wonder if is a secret to get option quotes unless you are an actual client...

 

[financeguy]

So how is the historical prices recorded then? It should exists somewhere. I'm going to register on Bloomberg and hope I'll get something to cling at least.

Just got reply from the Bloomberg team saying the following:

I wonder if is a secret to get option quotes unless you are an actual client...

As one of the handful of people in the market who decides what the prices of these products are, I can assure you that no historical data exists. I don't even have a record. One reason is simply that I don't price the same multi-asset option every day. The structuring team sends out (well at least at one point they did) at-the-money-spot worst-of options on certain underlyings in certain directions of certain tenors to a few clients periodically, but even those are only indicative prices and are certainly not immediately dealable. The point is, these products are not, and will not anytime soon, be publicly traded. A client calls up the sales desk and asks for a price, the salesperson asks the trader, the trader makes a price, the salesperson tells the client, the client deals, the trade is booked, an email confirmation is sent, back offices confirm and make sure the appropriate cash flows happen and appropriate credit and risk limits are adhered to. No one else will find out about this trade, sorry! That is why you won't be able to find out historical pricing. It doesn't exist.

 

[iHateVariance]

Tsotne, as most of this stuff is not liqd as financeguy said i think it makes sense as a theoretical exersise to focus on the dynamics as implied by ur dependence specification than actually matching some price? it would very important for u to understand how these dynamics will ultimately translate into pnl if u have to hold on to that risk for 1y+

 

[Tsotne]

Thanks you all for replies. I understand I cannot get the history and will have no data to compare. The only way is to construct a hypothetical basket prefferably with dependent industries since we are evaluating with copulas and use the theory to price such option. (which I have already done) I cannot check whether I'm getting reasonable answer or not. Thanks

 

[Mr Doe]

Try looking up US Dollar Index, bet you can find option data (and see dollar plunging :D Timber!).

 

[Tsotne]

I decided...hypothetical bivariate option with underlying - C and JPM. Strongly dependent...some more industries also under consideration. With 3 underlying also to be considered...Good examples for copula to estimate dependences.

 

[myampol]

Well, you're in luck -- the Nasdaq OMX just introduced a new product (barely 2 weeks ago) which is an option on the relative performance of an individual stock (e.g., AAPL) vs. a broad market ETF (e.g., SPY, which tracks the SPX index.) It trades on the PHLX, the former Philadelphia Stock Exchange, now owned by Nasdaq.

The product is called "Nasdaq OMX Alpha Index Options", and is currently being phased-in according to the schedule which can be obtained at http://www.nasdaqtrader.com/Micro.aspx?id=Alpha .

Here is some more information:
http://www.nasdaqtrader.com/content/phlx/alphafactsheet.pdf
http://www.nasdaqtrader.com/content/phlx/AlphaFAQs.pdf

Options on AAPL vs. SPY and on GOOG vs. SPY were launched last month; the tickers for the respective indexes are AVSPY and UVSPY.

http://finance.yahoo.com/q?s=^AVSPY
https://indexes.nasdaqomx.com/Data.aspx?IndexSymbol=AVSPY
http://finance.yahoo.com/q?s=^UVSPY
https://indexes.nasdaqomx.com/Data.aspx?IndexSymbol=UVSPY

For AAPL, which was the first contract to be listed on 4/18, several hundred calls and puts appear to have traded, but Bloomberg is (strangely) restricting me from seeing details about those trades.

For GOOG, 4 puts traded on the day of listing, and another 5 puts traded this past Monday.

Each day this week, one additional set of options is being launched:

Monday 5/2 GE vs. SPY (LVSPY) -- 3 calls traded Monday; 10 puts traded Tuesday ; nothing since
Tuesday 5/3 WMT vs. SPY (WVSPY) -- still no trades yet
Wednesday 5/4 INTC vs. SPY (JVSPY) -- 5 puts traded today
Thursday 5/5 MRK vs. SPY (NVSPY) -- still no trades yet
Friday 5/6 IBM vs. SPY (IVSPY)

So, other than on the first product traded (AAPL vs. SPY), there has been very little activity.

The product was designed by Prof. Robert Whaley and Jacob Sagi of Vanderbilt University.
A brief interview (transcript) with Whaley is here:
http://social.nasdaqomx.com/articledetail.cfm?EntryID=697&CategoryID=52

Their paper "Trading Relative Performance with Alpha Indexes" which describes these products is here: http://papers.ssrn.com/sol3/papers.cfm?abstract_id=1692738

[Whaley is also the inventor of the VIX index, which many of you (hopefully) know to be a variance swap which corresponds to the volatilities of options on the SPX index traded on the CBOE, a rival exchange. Whaley did that work while he was at Duke University.]

 

[Tsotne]

Thank you @myampol for such help. The sources you provided are really helpful and I'll use them sometime. For now I need rainbow options data. Historical quotes which I was unable to locate. It is hard to obtain OTC tarded product quotes generally. So I'm writing a completely theoretical paper with no ability to compare to actual results whether I have calculated a reasonable and fair price of an option in question or not. Thanks again for your help. I still have month and a half. I'll be searching for the data meanwhile.

 

[Tsotne]

Theories about it.

 

[ndirish]

Hi all. I've just read through this thread and have found it very helpful. I am just starting out doing a semester long research project on pricing Multi-Asset options for my MSFE class. The objective of my research is to produce numerical schemes for the pricing of these options using for example finite difference methods. My first question was if there is any good historical data which I can use for comparison; so I think I will use myampol's suggestions including the NASDAQ OMX Alpha Index options. I will also read into Tsotne's bivariate theories posted above. My question is what are the most popular methods in practice and theory currently in use for pricing such options? It would probably be must useful if my study included methods used in real world practice. I am thinking of doing a compare/contrast among each of them and hoping to offer suggestions for improvement. Also there are a number of these option types out there (rainbow options, basket options, FX options, best/worst options
and spread options) so I am not sure if I should focus on pricing one specific type or hope the methods apply to each type. I'd also like to compare prices generated from each model with actual historical prices (such as possibly the suggested NASDAQ OMX Alpha Index options). If anyone has ideas/suggestions/direction to help me along, it would be much appreciated. Thanks in advance.