I'm not sure I follow you, the owner of ETF's always have the option of exchanging their ETF shares for the shares of the underlying following the terms spelled out in the prospectus. If at any time the ETF price deviates enough from the underlying it would a simple matter for somebody to arb the two and reap the profit (buy the ETF shares and sell the underlying or vice-versa). Commission costs would make something like that prohibitive for a retail trader but I am sure there are institutional trading programs that do exactly that.
I did end up doing a Google search for "EWJ prospectus" and skimming through it, found that: - Barclays indeed does not intervene in the price - Barclays will redeem EWJ shares for shares of the underlying, but there is a minimum of 240,000 shares and a $2,400 fee. - They do this based on the price at the close of the US market. SO, there is indeed a separate trading action for EWJ that is related to Toyota and the other components, but is nevertheless unrelated in terms of the fluctuations between the open and close of the US market (but perhaps may be related to perceptions concerning how the US trading today might affect business for Japanese companies, in this case). BTW, Barclays calls this "tracking error" and calls the price at the close the "NAV". The prospectus contains stats on the tracking error in the past, and in this case, there have been two days per year of greater than 2.5% error, and two days per year of more than -2.5% error. All in all, I think that the "possibility" of shares being exchanged is enough to keep the shares roughly equivalent to the underlying. I doubt that anyone ever shows up at Barclays with several million dollars worth of EWJ shares. (However, market makers evidently must fork over the requisite amount of Toyota, etc. in order to "create" EWJ shares.) This also means that "retail investors" are dependent solely on this capability that vast institutions could step in, if the ETF price deviates too far. Someone who owns a standard lot of 100 shares has no recourse, it would seem.
Your statement is pretty much ridiculous. Your attitude is that of a 10 year old in a school yard. Pros eat weenies like you for lunch every day. An ETF is "not arbed"? Go make a few 1000 trades... and get an education. Until then... just STFU.
All the "execution" is the world won't matter... If you do not have ** proprietary **, real-time ETF arbitrage software... And are competing against 10 ms black boxes with your eyes and fingers. Keep on barking up the wrong tree. Over and out.
I didn't say that the purpose of this information is "arbing", you keep assuming that. The purpose of the information is understanding. If I start at the beginning with certain conclusions, then I end up in the same place I started. I can only get somewhere if I first obtain understanding of what is going on. Only then can I figure out the best way to go. PS Getting back to the topic, I found the following on/by Yahoo:
Additional information: http://www.federalreserve.gov/pubs/ifdp/2006/872/ifdp872.htm http://www.wisdomtree.com/home.asp I have other visual models I can give you if needed. Just pm me and let me know the format of choice. mathematica, Maple or Derive. All three are interactive, streaming-updatable with drill down. No, I'm not selling anything. BTW the answer to your question is both but at different intervals. The laymans answer is by trades and then underlying adjusted. (reset)
Wow, project much? Look, you were wrong when you said "One stock would never be 30% of an ETF... maybe max 3%." and you were wrong when you said it was true only for exotics. Why resort to ad hominems and profanity when someone corrects you?