- My Forums
- Tiger Rant
- LSU Recruiting
- SEC Rant
- Saints Talk
- Pelicans Talk
- More Sports Board
- Winter Olympics
- Fantasy Sports
- Golf Board
- Soccer Board
- O-T Lounge
- Tech Board
- Home/Garden Board
- Outdoor Board
- Health/Fitness Board
- Movie/TV Board
- Book Board
- Music Board
- Political Talk
- Money Talk
- Fark Board
- Gaming Board
- Travel Board
- Food/Drink Board
- Ticket Exchange
- TD Help Board
Customize My Forums- View All Forums
- Show Left Links
- Topic Sort Options
- Trending Topics
- Recent Topics
- Active Topics
Started By
Message
re: Stocks...
Posted on 5/2/10 at 11:14 pm to RedStickBR
Posted on 5/2/10 at 11:14 pm to RedStickBR
I emailed this guy, because I know he is somewhat accessible, and he is the de-facto leading expert on this stuff (I quoted something from him earlier about negative betas). We'll see if I get an answer. 
ETA: I'm just looking for someone to explain to me how the correlation doesn't become binomial when you divide it by standard deviation.
ETA: I'm just looking for someone to explain to me how the correlation doesn't become binomial when you divide it by standard deviation.
This post was edited on 5/2/10 at 11:15 pm
1
Posted on 5/2/10 at 11:43 pm to kfizzle85
Wouldn't standard deviation be a positive number regardless though? For example, if Stock A is expected to return 20% with a standard deviation of 10 percentage points, it could either return 20% plus 10% or 20% minus 10%, but the standard deviation is still positive 10. It wouldn't be +10% if the stock gained 30% nor would it be -10% if the stock gained only 10%. Rather, the standard deviation is just 10 percentage points.
For Beta coefficients, it would be <-1.5, -1.5, -1.0, -.5, 0, .5, 1.0, 1.5, >1.5, although it wouldn't be a normal bell curve since there are a ton more stocks who trade with positive Betas than negative Betas. For ease of explanation though, let's assume 0 is the top of the curve. The first standard deviation would be .5, the second 1.0, the third 1.5, etc. The standard deviation would be a positive figure, but when applying it against zero, it would lead to some negative results , i.e. a standard deviation .5 below 0 would be -.5, but the standard deviation is still .5.
ETA: Or wait, maybe that was your point. Since SD is always positive and a correlation can be positive or negative, you'd get some negatives when you divide correlation by SD.
For Beta coefficients, it would be <-1.5, -1.5, -1.0, -.5, 0, .5, 1.0, 1.5, >1.5, although it wouldn't be a normal bell curve since there are a ton more stocks who trade with positive Betas than negative Betas. For ease of explanation though, let's assume 0 is the top of the curve. The first standard deviation would be .5, the second 1.0, the third 1.5, etc. The standard deviation would be a positive figure, but when applying it against zero, it would lead to some negative results , i.e. a standard deviation .5 below 0 would be -.5, but the standard deviation is still .5.
ETA: Or wait, maybe that was your point. Since SD is always positive and a correlation can be positive or negative, you'd get some negatives when you divide correlation by SD.
This post was edited on 5/2/10 at 11:48 pm
Posted on 5/3/10 at 12:05 am to RedStickBR
Honestly, I'm not sure, so I'll just explain it how I'm thinking of it, which may turn out to be backwards.
1) correlation = measure how two assets (in the case of beta, a stock and the market portfolio) move together. A correlation of -1 means they move exactly oppposite, 0 no correlation whatsoever, +1 perfectly correlated.
2) So if stock A has a correlation of 1 with the market, it moves exactly how the market moves. A 5% up move in the market = a 5% up move in the stock, and vice versa. It tells you the direction of the move.
3) covariance takes correlation and makes it a unitless measure.
4) beta takes covariance, the unitless measure of correlation, and divides it by the standard deviation of the market, transforming the covariance into units of standard deviation.
5) standard deviation is movement around a mean (as you explained just now), it gives no indication of direction, its a range of outcomes.
6) Now you've taken the direction (correlation/covariance) and combined it with standard devation (% change, in either direction, binomial distribution).
I am thinking that I might be thinking about it backwards. Instead of taking a correlation coefficent and making it directionless by dividing by SD, beta takes the directionless SD and gives it direction by dividing with the covariance?
1) correlation = measure how two assets (in the case of beta, a stock and the market portfolio) move together. A correlation of -1 means they move exactly oppposite, 0 no correlation whatsoever, +1 perfectly correlated.
2) So if stock A has a correlation of 1 with the market, it moves exactly how the market moves. A 5% up move in the market = a 5% up move in the stock, and vice versa. It tells you the direction of the move.
3) covariance takes correlation and makes it a unitless measure.
4) beta takes covariance, the unitless measure of correlation, and divides it by the standard deviation of the market, transforming the covariance into units of standard deviation.
5) standard deviation is movement around a mean (as you explained just now), it gives no indication of direction, its a range of outcomes.
6) Now you've taken the direction (correlation/covariance) and combined it with standard devation (% change, in either direction, binomial distribution).
I am thinking that I might be thinking about it backwards. Instead of taking a correlation coefficent and making it directionless by dividing by SD, beta takes the directionless SD and gives it direction by dividing with the covariance?
Posted on 5/3/10 at 12:58 am to kfizzle85
I think you're wrong on Step 3. Correlation takes covariance and makes it a unitless measure (i.e. between -1 and 1). Likewise, SD takes variance (which is a squared value) and makes it a number that is not squared.
Since correlation of -1 equals a security that moves in the opposite direction of the market, and since it is relatively rare that a security moves in the opposite direction of the market, it will thus be relatively rare that you end up with a security that has a Beta w/ a negative value after you divide correlation by SD.
Since SD is positive and most correlations will be either 0 or 1, when you take a correlation (remember most have value of 0 or 1) and divide that value by SD, you get a Beta value of either 0 or greater than 0 for most cases. Only rarely do you see a negative correlation being divided by a positive SD and thus see a negative Beta.
Since correlation can have a value of 0, it therefore must be the numerator, because if you were to divide by 0, you would get nothing.
So Beta must equal correlation / SD. Since SD is always positive and since correlation is usually either 0 or positive, most stocks will have a Beta that is either 0 or positive.
Thus, Beta accounts for both volatility (SD) and direction (correlation).
Since correlation of -1 equals a security that moves in the opposite direction of the market, and since it is relatively rare that a security moves in the opposite direction of the market, it will thus be relatively rare that you end up with a security that has a Beta w/ a negative value after you divide correlation by SD.
Since SD is positive and most correlations will be either 0 or 1, when you take a correlation (remember most have value of 0 or 1) and divide that value by SD, you get a Beta value of either 0 or greater than 0 for most cases. Only rarely do you see a negative correlation being divided by a positive SD and thus see a negative Beta.
Since correlation can have a value of 0, it therefore must be the numerator, because if you were to divide by 0, you would get nothing.
So Beta must equal correlation / SD. Since SD is always positive and since correlation is usually either 0 or positive, most stocks will have a Beta that is either 0 or positive.
Thus, Beta accounts for both volatility (SD) and direction (correlation).
This post was edited on 5/3/10 at 1:06 am
Posted on 5/3/10 at 1:11 am to RedStickBR
Alright, I don't know if my correlation / SD formula is correct, but this explanation hits the nail on the head:
LINK
Thus, you get your direction from the correlation and your volatility from Beta, and since correlation is a function of Beta, Beta actually gives you both direction and volatility!
quote:
There is a simple formula between beta and volatility:
That is, beta is a combination of volatility and correlation. For example, if one stock has low volatility and high correlation, and the other stock has low correlation and high volatility, beta can decide which is more "risky".
This also leads to an inequality (since |r| is not greater than one):
In other words, beta sets a floor on volatility. For example, if market volatility is 10%, any stock (or fund) with a beta of 1 must have volatility of at least 10%.
Another way of distinguishing between beta and correlation is to think about direction and magnitude. If the market is always up 10% and a stock is always up 20%, the correlation is one (correlation measures direction, not magnitude). However, beta takes into account both direction and magnitude, so in the same example the beta would be 2 (the stock is up twice as much as the market).
LINK
Thus, you get your direction from the correlation and your volatility from Beta, and since correlation is a function of Beta, Beta actually gives you both direction and volatility!
Posted on 5/3/10 at 1:12 am to RedStickBR
I'll go with that.
Posted on 5/3/10 at 1:16 am to RedStickBR
Thus, the correct understanding of Beta can be seen in View 3:
And it should be noted that very rarely will stocks have a negative Beta value (only 72 of the thousands of stocks that trade on the Amex, Nasdaq, and NYSE).
*** Sorry to everyone who's had to read this shite.
*** I am one happy camper right now now that this is settled!
*** We need to be looking for small-caps with Betas close to 0 (or if ballsy, with negative Beta values since we're expecting a downturn by the Dow). But choosing stocks with Betas close to 0 would be the safest bet because even if the Dow doesn't turn down, we'd still be fine since our picks don't follow the market anyway.
*** Also, I should stop saying Dow and instead should say S&P500 as when used in Beta, 1 exemplifies the S&P500 from what I've read, and not the Dow.
quote:
Any positive Beta value moves in same direction as market. Beta value of 2 moves in same direction as market, but twice as much. Beta value of .5 moves in same direction as market, but half as much. Beta value of -2 moves in opposite direction as market, but twice as much. Beta value of -.5 moves in opposite direction as market, but half as much.
And it should be noted that very rarely will stocks have a negative Beta value (only 72 of the thousands of stocks that trade on the Amex, Nasdaq, and NYSE).
*** Sorry to everyone who's had to read this shite.
*** I am one happy camper right now now that this is settled!
*** We need to be looking for small-caps with Betas close to 0 (or if ballsy, with negative Beta values since we're expecting a downturn by the Dow). But choosing stocks with Betas close to 0 would be the safest bet because even if the Dow doesn't turn down, we'd still be fine since our picks don't follow the market anyway.
*** Also, I should stop saying Dow and instead should say S&P500 as when used in Beta, 1 exemplifies the S&P500 from what I've read, and not the Dow.
This post was edited on 5/3/10 at 1:18 am
Posted on 5/3/10 at 1:20 am to RedStickBR
It can be anything, but most sites use the s&p 500 to represent the "market portfolio" (even though it doesn't). I'm annoyed I was wrong, but I am okay with it because it was a math question that I couldn't answer. I know the beta formula by memory, I was just thinking about it backwards. I suck at teh maths.
Posted on 5/3/10 at 1:25 am to kfizzle85
I was wrong, too. I don't know what the hell I was thinking saying that anything away from 1 does not move with market and anything at 1 does move with market. In reality, a Beta of 10 still moves with market, but just with 10 times the magnitude. If I'd have picked stocks with high Betas, I'd have been screwed if the market took a downturn!
I think we are now the leading authorities on Beta. Our explanations are now better than either of the Professors that emailed you and THF, and any of the crap we've seen on the internet, apart from Wikipedia. Amazing that Wiki was the most accurate source
I think we are now the leading authorities on Beta. Our explanations are now better than either of the Professors that emailed you and THF, and any of the crap we've seen on the internet, apart from Wikipedia. Amazing that Wiki was the most accurate source
This post was edited on 5/3/10 at 1:26 am
Posted on 5/3/10 at 1:32 am to RedStickBR
Yeah that was a pretty thorough dissection/construction, I don't think I'll ever have any comprehension issues with beta ever again. I wish I could think through shite like that all the time. Surprisingly easy to communicate too, considering it is a message board and the topic was jargon-filled and a really specific question. See that old people? We communicate just fine, its you guys who lack the ability to use technology, not us (/rant).
Posted on 5/3/10 at 1:34 am to RedStickBR
quote:
If I'd have picked stocks with high Betas, I'd have been screwed if the market took a downturn!
Yeah there is no question about that. Higher beta definitely equals higher risk and vice versa, and that's readily observable (utilities with betas less than 1, tech companies and financials with beta over 2). I usually just think of it like that and rarely in terms of its individual components like we just went at lengths to break down, but I'm really glad we went through all that in all seriousness.
Posted on 5/3/10 at 1:48 am to kfizzle85
Yeah, I am, too. I appreciate you walking me through it at first so I could come to some general understanding of it. Plugging correlation into the equation is really what brought it around full circle for me.
I did a scan and came up with this:
# of stocks w/ Betas below 0: 72
# of stocks w/ Betas equal to 0: none
# of stocks w/ Betas between 0 and .49: 518
# of stocks w/ Betas between .5 and .99: 1,282
# of stocks w/ Betas between 1 and 1.49: 1,424
# of stocks w/ Betas between 1.5 and 1.99: 819
# of stocks w/ Betas greater than 2: 745
Thus, the large majority of stocks trade with Betas greater than or equal to 1, meaning most move in the general direction of the market but with greater volatilities. This makes sense as rarely does the S&P see a 20% gain, but very frequently do individual stocks see such gains.
I think investors in this economic climate would be best suited investing in the 518 companies with Betas between 0 and .49.
*** Results include stocks trading on either Amex, Nasdaq, or NYSE.
Now all I need is for a Beta question to appear on my Income Tax II or Family Law exams.
I did a scan and came up with this:
# of stocks w/ Betas below 0: 72
# of stocks w/ Betas equal to 0: none
# of stocks w/ Betas between 0 and .49: 518
# of stocks w/ Betas between .5 and .99: 1,282
# of stocks w/ Betas between 1 and 1.49: 1,424
# of stocks w/ Betas between 1.5 and 1.99: 819
# of stocks w/ Betas greater than 2: 745
Thus, the large majority of stocks trade with Betas greater than or equal to 1, meaning most move in the general direction of the market but with greater volatilities. This makes sense as rarely does the S&P see a 20% gain, but very frequently do individual stocks see such gains.
I think investors in this economic climate would be best suited investing in the 518 companies with Betas between 0 and .49.
*** Results include stocks trading on either Amex, Nasdaq, or NYSE.
Now all I need is for a Beta question to appear on my Income Tax II or Family Law exams.
This post was edited on 5/3/10 at 1:50 am
Posted on 5/3/10 at 2:28 am to RedStickBR
How do you think the chinese banks raising rates will effect the chinese stocks? I also read something about the auto industry being over supplied etc.
Posted on 5/3/10 at 8:44 am to RedStickBR
quote:
SORL should have a good report, yes. They're going to blow away their year-over-year Q1 earnings and should start being priced for their fiscal year 2010 yearly earnings estimate, which is 80 cents. I don't really see anything that could go wrong there.
They are expected to make .14 per share this Q. So you expect them to be better than that?
Posted on 5/3/10 at 9:18 am to wegotdatwood
I think you'll see .16 or .17.
Hey, would you mind posting the links to the articles you read re: China?
Hey, would you mind posting the links to the articles you read re: China?
Posted on 5/3/10 at 10:04 am to RedStickBR
LINK
This wasn't the exact one. The one I read was on google finance earlier but I couldn't find that one.
This wasn't the exact one. The one I read was on google finance earlier but I couldn't find that one.
Posted on 5/3/10 at 10:28 am to wegotdatwood
You think it's a good time to get in on lpih? It's down to about 2.50.
Posted on 5/3/10 at 10:37 am to wegotdatwood
Yep, I think it's putting in a bottom today. Warrants should be about exhausted now.
Posted on 5/3/10 at 11:13 am to RedStickBR
So it's good that the warrants are about exhausted? Do you currently have any lpih?
Posted on 5/3/10 at 11:14 am to RedStickBR
CLRT's definitely havin a good day. Anyone familiar with the drug company market? How long will it usually climb?
Page 17 of 26
Popular
Back to top
Follow TigerDroppings for LSU Football News