Pure Arbitrage...Still in Existence?

[bvam1]

MR.NBBO

"The world wasn't invented yesterday. If you've thought of it.....we've got it covered already."

_______________________________________

I really like that statement! However, when it comes to arb. techniques, success is determine by efficiency rather than mere knowledge alone. Everyone may know about it, but only a few can be successfully by doing it in the best, efficient way possible?

Furthermore, if your statement holds true, then it would be quite hard to make money in this business, right? After all, you guys got it covered already! Backtesting with six years of historical data (since inception of QQQQ) shows a rather different view!!! People can make all the assumptions they want, but facts do not lie!!!
________________________________________

And Landboy,

If you check Mr.NBBO's post, you'll see that NAV is not calculated once per day, but many times aday. But you're right about one thing though, Official NAV is calculated once a day! Arb. trading experience and your work experience are two uncorrelated things!

 

[winter]

Quote from bvam1:

MR.NBBO

"The world wasn't invented yesterday. If you've thought of it.....we've got it covered already."

_______________________________________

I really like that statement! However, when it comes to arb. techniques, success is determine by efficiency rather than mere knowledge alone. Everyone may know about it, but only a few can be successfully by doing it in the best, efficient way possible?

Furthermore, if your statement holds true, then it would be quite hard to make money in this business, right? After all, you guys got it covered already! Backtesting with six years of historical data (since inception of QQQQ) shows a rather different view!!! People can make all the assumptions they want, but facts do not lie!!!

I think you are confusing two issues, I believe that MR.NBBO was referring to arb techniques, as in any arb technique you can think of has already been covered - that doesn't mean you can't make money "in this business" but for you to make money arb'ing means you have to compete against big players who are already doing whatever you are considering doing and likely have lower costs and other advantages over you.

I have no idea if this is relevant but since I havent seen it thrown out here I thought I would, could the timing on when dividends are paid on QQQQ vs the individual stocks explain some of the premium?

 

[landboy]

No I'm still right, just misunderstood the original post. If you want pure arbitrage like what the big institionals do, they settle with the bank and "create new units." A "dirty NAV" is not the same as a NAV, so in essence I have been the only one using the term correctly. I charge a few hundred dollars striking a NAV per day, if you want to try and calculate it yourself, by all means go ahead, I get paid either way.

Quote from Sparohok:

Well, you were wrong before and you're wrong now. AMEX datafeed contains NAV estimates (Intraday Indicative Values) that are updated every 15 seconds throughout the day.

http://www.amex.com/?href=/etf/Glossary/Gloss.htm

BTW, another cost related to ETF arbitrage is the creation unit cost. It is on the order of one or two cents per share but that's enough to cut into arbitrage margins.

Martin

 

[LoosenUp]

Pardon my ignorance, but do we have to also account for the costs of slippage or the bid/ask spreads? NAV are calculated on the last transacted price but you must bid up to get the stocks you want. Maybe NAV calculation is much more complex than that.

 

[Sparohok]

Quote from landboy:

If you want pure arbitrage like what the big institionals do, they settle with the bank and "create new units."

Arbitrageurs replicate shares of the ETF throughout the trading day and then trade creation units at the close in order to zero out their inventory. I'm quite certain they don't actually rely on the IIV datafeed, instead calculating the fair price themselves based on their own cost models. Nonetheless, the NAV is effectively published every 15 seconds in the IIV. If your end of day NAV is so much better that it's the only price worth being anointed with those three precious letters, well, hey, more power to you. It sounds like a nice job.

Martin

 

[landboy]

It sure pays the bills for what is often a "regulatory hurdle."

Quote from Sparohok:

Arbitrageurs replicate shares of the ETF throughout the trading day and then trade creation units at the close in order to zero out their inventory. I'm quite certain they don't actually rely on the IIV datafeed, instead calculating the fair price themselves based on their own cost models. Nonetheless, the NAV is effectively published every 15 seconds in the IIV. If your end of day NAV is so much better that it's the only price worth being anointed with those three precious letters, well, hey, more power to you. It sounds like a nice job.

Martin

 

[stock777]

Wow, a few hundred a day to do NAV's? I would think a computer would be doing that for 2 cents a week.

You work for Barclays, or did they need to go freelance cause no one there could figure out how to do the massive complex calculation.

 

[bvam1]

Finally, I just found out the answer to the mystery!

The Nasdaq, along with other data vendors, posts NAV in a sort of confusing way. Since, NAVs are calculated at the end of the day and posted tomorrow, it lags behind by a whole day.

So, the NAV on 7/13/05 at 38.32 as posted by the Nasdaq is, or should I say, was the NAV for the close of 7/12/05 trading day.

Nevertheless, on 8/3/2001 the premium was .82. On several occasions since '01 - 05, the premiums/discounts were somewhere in the .2-.3 range. Normally, the prem/disc. falls in the range of -.05 to +.05.

I would dare to venture to say that the break even point for institutional arb. is normally .02 or .03 prem/disc.

 

[landboy]

Things like corporate actions, cash settlements, income distribution, financial statements etc. take a bit of effort, but in reality you're right it's not too difficult.

Think of it as doing the accounting for a mutual fund, only that it is close ended, so there is little capstock activity. With SOX and the like, we're all making pretty good money these days. Plus remember, I do more than one fund a day, otherwise (like you can imagine) I'd be sitting on my hands all day

Quote from stock777:

Wow, a few hundred a day to do NAV's? I would think a computer would be doing that for 2 cents a week.

You work for Barclays, or did they need to go freelance cause no one there could figure out how to do the massive complex calculation.

 

[ig0r]

Quote from bvam1:

Finally, I just found out the answer to the mystery!

The Nasdaq, along with other data vendors, posts NAV in a sort of confusing way. Since, NAVs are calculated at the end of the day and posted tomorrow, it lags behind by a whole day.

So, the NAV on 7/13/05 at 38.32 as posted by the Nasdaq is, or should I say, was the NAV for the close of 7/12/05 trading day.

Nevertheless, on 8/3/2001 the premium was .82. On several occasions since '01 - 05, the premiums/discounts were somewhere in the .2-.3 range. Normally, the prem/disc. falls in the range of -.05 to +.05.

I would dare to venture to say that the break even point for institutional arb. is normally .02 or .03 prem/disc.

I believe someone already mentioned this but I'll hit on it again - the biggest problem I have with this kind of arb is the fact that you need to lift the offer on 100 stocks aftering selling your Q hedge (leaning on the 100 stocks instead of the Qs would definitely not remedy the situation). Assuming that most of them are trading at the bid, you essentially lose a penny on each one of those. Even a seemingly big discount like 0.75 would quickly get eaten up this way. This is ignoring trading costs, of course. The same above applies for a premium to NAV. I personally would not even be comfortable putting this arb on alone unless the premium/discount is greater than like 1.25 (very possible that the market will move away from you)