The Opening Orders Thread

yes that will be helpful, but how do you get the e^(r-q)t.

which r do you use?
which q do you use?
how about the volatility of r?

its a simple enough formula but where do you get your inputs from?
 
Quote from econometrics:



its a simple enough formula but where do you get your inputs from?
More...
As said above, we dont get the r and q. We just use the FV number and blindly thrust the provider!
 
that so-called "FV" is what the college kids learn in finance 101 in the first class or so. the idea is very simple if you know the inputs. so you just trust the number provided by some sites without knowing how the numbers are calculated? don't you even know where they get the numbers from?
I guess it is covered in 2 pages in Hull's, because he simply assumes some continuouesly compounded dividend yield and some constant r. How do you really get the r and q (emphasis on the latter) ??

also assuming you really "know" the r and q, that will just give you the no-arb premium/discount from spot in a perfect market of the index. then finding the movement away from that if-constant index value will only tell you how the index might have changed, not how individual names should change. how do you map the change in index value to change in single name values? you call this a "strategy"??
 
Quote from econometrics:

that so-called "FV" is what the college kids learn in finance 101 in the first class or so. the idea is very simple if you know the inputs. so you just trust the number provided by some sites without knowing how the numbers are calculated? don't you even know where they get the numbers from?
I guess it is covered in 2 pages in Hull's, because he simply assumes some continuouesly compounded dividend yield and some constant r. How do you really get the r and q (emphasis on the latter) ??

also assuming you really "know" the r and q, that will just give you the no-arb premium/discount from spot in a perfect market of the index. then finding the movement away from that if-constant index value will only tell you how the index might have changed, not how individual names should change. how do you map the change in index value to change in single name values? you call this a "strategy"??
More...
If you want to trade the OPG, then the FV that we blindly get is perfect. No need to complicate things. You could also just use the close and pre market of SPY without the FV.
If you care about the r and q and can't help you. I understand that you need the estimates of dividends and all that but I leave that to other people.
The calculation that we make is close for most stocks, the idea is that without news they should move like the futures (assuming a beta of 1). You can also have a look at the pre bid/ask in the ECN's. But no, in theory, the calculation does not predict the price of a single issue. It's all about the probabilities, they do about most of the time or close to so it's a good estimate.
Yes, it's a strategy. You expect a stock to open at 34.75, pre bid/ask is 34.60/34.85, you buy at 34.58, sell at 37.88 and try to make a couple pennies if you get the price of the auction. It's a scalp of a couple seconds, minutes sometimes. The numbers are easy but doing it in real time is the chalenge.
 
i understand that it is not perfect, even index arb in a BB IB leaves a certain degree of imperfection, but the degree of imperfection just worries me to categorize it as a "strategy".
First, that "FV" is calculated without any understanding, that whole thing is very sensitive to r and q and you dont bother to know where they are from??

second, if every constituent moves 1:1 with the index, then the index wouldn't move by the amount (contradiction). it will give an idea of direction and its magnitude, but it wont give you an expected open price of a single name. It might give a range, but that range would be large given the errors in the index "FV" and the mapping.

Ok, fair enough, I never attempted this "strategy" and if you are really making money off it, good for you.
 
Quote from econometrics:

i understand that it is not perfect, even index arb in a BB IB leaves a certain degree of imperfection, but the degree of imperfection just worries me to categorize it as a "strategy".
First, that "FV" is calculated without any understanding, that whole thing is very sensitive to r and q and you dont bother to know where they are from??

second, if every constituent moves 1:1 with the index, then the index wouldn't move by the amount (contradiction). it will give an idea of direction and its magnitude, but it wont give you an expected open price of a single name. It might give a range, but that range would be large given the errors in the index "FV" and the mapping.

Ok, fair enough, I never attempted this "strategy" and if you are really making money off it, good for you.
More...
I do understand what the dividend yield and the risk free rate are. It is true that I dont know where they come from but I did some "testing" before using it.
There can be a contradiction and surelly the stocks dont move 1:1 with the index. It's an approximation and we enveloppe that approcimation with a range. We get to know uor stocks so we keep some, adjust some and delete some. It's an art, not a science.
I agree with you, enough said.
 
i understand that you guys use an error range, or envelope whatever hell you wanna call it. Given the errors and compounded error involved in this college 101 logic, I just don't think the range is wide enough. How i see it is that you assume the point estimate to be correct then "envelop" it using a "correct value range" out of which you think represents under/over-pricing. however, the range seems more like an error-range to me out of which does not indicate under/over-pricing (it may if the error has an opposite sign).

anyways, as you said, enough has been said, it is just a very rough guide to make a couple of pennies off the openings. but then again, it is a grandmother-can-do-it calculation, not a strategy as you guys claim it to be.
 
Quote from econometrics:

i understand that you guys use an error range, or envelope whatever hell you wanna call it. Given the errors and compounded error involved in this college 101 logic, I just don't think the range is wide enough. How i see it is that you assume the point estimate to be correct then "envelop" it using a "correct value range" out of which you think represents under/over-pricing. however, the range seems more like an error-range to me out of which does not indicate under/over-pricing (it may if the error has an opposite sign).

anyways, as you said, enough has been said, it is just a very rough guide to make a couple of pennies off the openings. but then again, it is a grandmother-can-do-it calculation, not a strategy as you guys claim it to be.
More...

how is it not a strategy? You are getting in on the opening print which is basically manipulated by the specialist based on supply and demand. You are providing liquidity for that print and you get compensated for it. The calculation just tells you where the stock should theoretically open.
 
Quote from samueldoernte:

how is it not a strategy? You are getting in on the opening print which is basically manipulated by the specialist based on supply and demand. You are providing liquidity for that print and you get compensated for it. The calculation just tells you where the stock should theoretically open.
More...

how is it a strategy? if you call something like "buy when Px > 20d-MA or nothing otherwise " a strategy, then ok there you go.

if the calculation were to tell you the "right" price to open, then you would be providing liquidity, but it does not. anyways, i got the idea of what you guys mean by "opening strategy", let's just agree to disagree. it all matters if you make money.
 
You must log in or register to reply here.
Back
Top