|
Post by Red Shirted Ensign on Jun 21, 2015 19:51:36 GMT -8
insert code here
Balls in your court Taylor....
|
|
|
Post by archibaldtuttle on Jun 21, 2015 19:57:35 GMT -8
Fascinating. Well, now we'll see how much that will cost them!
|
|
|
Post by jmolloy on Jun 21, 2015 20:14:53 GMT -8
Eddy Cue @cue 19m19 minutes ago #AppleMusic will pay artist for streaming, even during customer’s free trial period Eddy Cue @cue 18m18 minutes ago We hear you @taylorswift13 and indie artists. Love, Apple Excellent news. Now next time respond faster and then we don't get shit like this on the BBC: www.bbc.com/news/entertainment-arts-33216778(yes I know the BBC is slightly biased against Apple.) (a smidgeon of sarcasm there)
|
|
4aapl
Moderator
Posts: 3,598
|
Post by 4aapl on Jun 21, 2015 20:23:10 GMT -8
Understood. How about something during the first 90 days....even ten percent of the full cut...? Come on, credit where credit is due. Apple is relying on these artists to make them money. Apple decided it wanted to offer it for free for 90 days. Why should the musicians sacrifice their cut to promote Apple's new service? It's not like Apple doesn't have the money. Ya know, while I respect that Apple now chose to change this, I hope they worked it in to everything. I mean, if Apple is paying out quite a bit more for streaming in the long run, then I do think that artists should have shared the lack of income in the trial period. Something about having your cake and eating it too, or having the best of both worlds. Or like braking or turning on ice/slippery snow. Pick one. Apple apparently found a way to make the numbers work. But was it the right move, or just one of those things that they could afford since they have the money in the bank, even though they really shouldn't have?
|
|
|
Post by jmolloy on Jun 21, 2015 20:47:11 GMT -8
Ya know, while I respect that Apple now chose to change this, I hope they worked it in to everything. I mean, if Apple is paying out quite a bit more for streaming in the long run, then I do think that artists should have shared the lack of income in the trial period. Something about having your cake and eating it too, or having the best of both worlds. Or like braking or turning on ice/slippery snow. Pick one. Apple apparently found a way to make the numbers work. But was it the right move, or just one of those things that they could afford since they have the money in the bank, even though they really shouldn't have? Hmmm. I assume you have never made music to live with. Fairly certain that gtrplyr does, as do I. Here's the thing. Your boss decides to work for free, but only on the condition that everyone under him does so, save it's not a condition, it's an agreement that you were contractually signed to despite the fact you'd never sign that contract. Apple wants to use the first 90 days as a free trial. But they are not prepared to pay for it themselves, they'll get the artists and rights holders to do that for them Seriously? From a shareholder point of view I can grasp what you are saying. BUT FFS this is not what musicians agreed to and it starts the whole Apple Music thing off on the wrong foot. Spotify put music up that they had no agreement in place to use. just upload it all - basically Napster. Apple just tried to pull the same stunt. They are better than this and have learned the hard way that they were wrong.
|
|
|
Post by Red Shirted Ensign on Jun 21, 2015 20:51:08 GMT -8
Apple will pay at different rates for trial period...and Eddie Cue called Taylor Swift...
Love when it all makes sense and comes together....
|
|
|
Post by Red Shirted Ensign on Jun 21, 2015 20:52:42 GMT -8
|
|
mark
fire starter
Posts: 1,544
|
Post by mark on Jun 21, 2015 21:16:26 GMT -8
I bought some in the money jan 2016 calls. I won't give the exact strike but they weren't deep in the money. Close to current price. Thanks Nagrani! Out of curiosity, what's your worry about sharing the specific strike? I know in giving real-time trading data, sharing the specific strike can reveal more than you'd prefer. At one time I had a script that grabbed CBOE quotes every 15 minutes, and if I wanted I could have found the approximate size of trades that Gregg or a couple other larger investors on TMF were making, when their trades were likely a couple times larger than mine. But it never really mattered much to me, and though it's sometimes fun to see the specs on someone who shares all the data including quantity, it really doesn't change things. If we both make 80% off of an options trade, it really doesn't matter if you make $80k while I only made $8k off of it. Some people are worried about a "seek and destroy" algorithm reading through our posts. Really the only worry is the slim possibility of everyone jumping into the same strike and boosting it's OI enough to make it a magnet point, but frankly I don't think we have the numbers for that. At the same time, I can understand that, and for the last few years I generally try to make my trades down a strike or two from large OI. I usually purposely pick the large OIs when possible. The essential reason is the larger the OI, the better the liquidity (most of the time). And the better the liquidity, the quicker I can execute once I make a decision.
|
|
Deleted
Deleted Member
Posts: 0
|
Post by Deleted on Jun 21, 2015 21:37:34 GMT -8
If Taylor Swift really had the artists' back, she go after the labels that take so much of the cut. THAT's the 800lb gorilla, not a three month trial period. She's grandstanding, pure and simple.
|
|
|
Post by BillH on Jun 21, 2015 21:40:12 GMT -8
Props to Eddy and Apple for their rapid response. Maybe you music guys could straighten me out on something. Had I been following this from the start my inclination would have been to ask if the artist's bitch wasn't really with the labels who suggested and/or agreed with the deal. Don't the labels pay the artist's? Aren't the musicians contract's with the labels? If someone's going to eat the expense shouldn't it be the labels as a promotion expense or Apple and the labels? I must be missing something here or one of you guys would have brought this up already. Thanks for your help.
Sign me,
Confused
|
|
|
Post by BillH on Jun 21, 2015 21:41:37 GMT -8
If Taylor Swift really had the artists' back, she go after the labels that take so much of the cut. THAT's the 800lb gorilla, not a three month trial period. She's grandstanding, pure and simple. Damn, talk about not waiting long enough for someone to step up. I was still typing when you posted this.
|
|
JDSoCal
Member
Aspiring oligarch
Posts: 4,181
|
Post by JDSoCal on Jun 21, 2015 22:05:57 GMT -8
If Taylor Swift really had the artists' back, she go after the labels that take so much of the cut. THAT's the 800lb gorilla, not a three month trial period. She's grandstanding, pure and simple. Egg-zactly. Which made me wonder if she was paid by the labels as a proxy for this fight. I'm just glad Apple batted this one down quickly, since we have no need for Pyrrhic victories on a break-even or loss leader service.
|
|
|
Post by mace on Jun 21, 2015 23:04:55 GMT -8
Privacy, size of my trade and I don't want people following me blindly - don't want the guilt if my trade doesn't pan out. 1000 contracts !!!!! Over a million dollar worth, bravo.
|
|
|
Post by mace on Jun 21, 2015 23:37:29 GMT -8
Taylor Swift just shoots out of her mouth for her own personal gain... hate it when celebrities talk about noble goals when deep down is for their own self-interest. If you want out, then out, stop talking for other people. And quote a noble goal without doing due diligence.
|
|
4aapl
Moderator
Posts: 3,598
|
Post by 4aapl on Jun 22, 2015 8:22:22 GMT -8
Here's the thing. Your boss decides to work for free, but only on the condition that everyone under him does so, save it's not a condition, it's an agreement that you were contractually signed to despite the fact you'd never sign that contract. Apple wants to use the first 90 days as a free trial. But they are not prepared to pay for it themselves, they'll get the artists and rights holders to do that for them Seriously? If they were paying them the same as other services in the end, sure you'd be right. And that was my initial feeling, that obviously the artists should be paid during this time. But then there's this thing where they are paying more than other streaming services. So the question comes to be, what's the breakeven timeframe, and was this part of the negotiations. To me, it seems that it's a tradeoff, where the artists would make a little less upfront, but then be better off 1-1.5 years off. 5 years later, they'd make significantly more with Apple's service than others. EDIT (ok, last edit. One of the articles said Apple's US payout was going to be 71.5%, as opposed to spotify saying they pay out 70%. If the difference really is so little, then no I don't think Apple could justify not paying during a free trial unless it was mutually agreed on.) If I was deciding between 2 stocks that were exactly the same, one with a dividend of 3% and the other with 4%, but the 4% one was going to miss their first quarterly dividend due to a one time payment to buy out a company and get more customers by doing such, I'd be buying the company with the 4% payout if I was looking at the longterm. What I would hate to see is this being part of the negotiation, a trade-off that was made by both sides in order to give a higher payout rate, and then for one side to come back later and try to change it via public outcry. But again, it's all part of the trade-offs in a contract, and it's been renegotiated now so it doesn't matter. Nice to see green today! (second weekday of summer break, and the kids are already climbing the trees and barking like dogs. It's going to be a fun summer)
|
|
bud777
fire starter
Posts: 1,352
|
Post by bud777 on Dec 22, 2020 9:43:13 GMT -8
I think I was one of the first ones to talk about algorithms here so I wanted to elaborate on the kinds of algorithms I was talking about. When I read "Dark Pools", I learned about new algorithmic work that was based on more than the behavior of the stock price and volume. For those who don't spend a lot of time with statistics, let me give a little background. One of the most general statistical tools is called the General Linear Model or GLM. GLM allows you to calculate the ability of a formula to predict a variable or variables. The variables can be continuous like we see in regression, but they can also include state variables like "home or away". The GLM calculates the overall variance of the dependent variable ( the one we want to predict) and then shows us how much each of the variables contributes to that variance. So, if you wanted to explore how much age, wealth, race, and education contributed to political attitudes, you could set up an equation ( or model) like P = aA +bW + cR + dE where A=Age, W = wealth, R = race and E = education, and the model would calculate the coefficients (a,b,c,d) to give the best fit. The quality of the fit is called R squared. R squared equals 1 is a perfect fit. The art of using something like this is to find the right set of variables. Normally an analyst or experimenter would use their knowledge of the underlying system to develop the model, but a few years ago things changed. Keep in mind that there is nothing here that limits the number of variables. A human would probably stop at 5 or 6, but a computer could have an unlimited number. You could have the computer just try variables at random and keep whatever works best. Now think of the variables as a string of beads or maybe as a set of amino acids on a strand of DNA. While a random set of variables might eventually lead to a better predictor, maybe there is a better way to choose the new combinations. Given the analogy to DNA, people developed a way to generate the combinations by imitating the mechanisms of genetics. This gives a new candidate string that is different from the previous one, but not that different, a controlled mutation that can evolve. Once this is in place, it is just necessary to turn it loose and supply it with data for the variables. Data can be the normal financial data, the number of FUD articles or the number of times cars are mentioned on AFB. No one really knows what variables are being used, not even the owners of the algorithm. All that matters is the R squared. According to "Dark Pools" there is solid evidence that these work. The algorithms develop their models, make their predictions, and place their bets untouched by human hands. We are just along for the ride. I hope this wasn't too pedantic, but I wanted to distinguish between algorithms guided by analysts and thus subject to FUD and bias and what is really going on. If you haven't read "Dark Pools" I recommend it. It is a fascinating account of how computers changed Wall Street. I posted this in 2015. I am not sure if "cars" on AFB were ever part of an HFT algorithm, or if the algorithms would be smart enough to distinguish between cars as an indicator of euphoria and "cars" as a discussion of future products. I would imagine that any influence that the variable might have would be overwhelmed by the general behavior of the market and analysts' opinions, but it is all guesswork. The predictive equations are a model and as we all know....All models are wrong, some are useful.
|
|
4aapl
Moderator
Posts: 3,598
|
Post by 4aapl on Dec 22, 2020 11:32:54 GMT -8
I think I was one of the first ones to talk about algorithms here so I wanted to elaborate on the kinds of algorithms I was talking about. When I read "Dark Pools", I learned about new algorithmic work that was based on more than the behavior of the stock price and volume. For those who don't spend a lot of time with statistics, let me give a little background. One of the most general statistical tools is called the General Linear Model or GLM. GLM allows you to calculate the ability of a formula to predict a variable or variables. The variables can be continuous like we see in regression, but they can also include state variables like "home or away". The GLM calculates the overall variance of the dependent variable ( the one we want to predict) and then shows us how much each of the variables contributes to that variance. So, if you wanted to explore how much age, wealth, race, and education contributed to political attitudes, you could set up an equation ( or model) like P = aA +bW + cR + dE where A=Age, W = wealth, R = race and E = education, and the model would calculate the coefficients (a,b,c,d) to give the best fit. The quality of the fit is called R squared. R squared equals 1 is a perfect fit. The art of using something like this is to find the right set of variables. Normally an analyst or experimenter would use their knowledge of the underlying system to develop the model, but a few years ago things changed. Keep in mind that there is nothing here that limits the number of variables. A human would probably stop at 5 or 6, but a computer could have an unlimited number. You could have the computer just try variables at random and keep whatever works best. Now think of the variables as a string of beads or maybe as a set of amino acids on a strand of DNA. While a random set of variables might eventually lead to a better predictor, maybe there is a better way to choose the new combinations. Given the analogy to DNA, people developed a way to generate the combinations by imitating the mechanisms of genetics. This gives a new candidate string that is different from the previous one, but not that different, a controlled mutation that can evolve. Once this is in place, it is just necessary to turn it loose and supply it with data for the variables. Data can be the normal financial data, the number of FUD articles or the number of times cars are mentioned on AFB. No one really knows what variables are being used, not even the owners of the algorithm. All that matters is the R squared. According to "Dark Pools" there is solid evidence that these work. The algorithms develop their models, make their predictions, and place their bets untouched by human hands. We are just along for the ride. I hope this wasn't too pedantic, but I wanted to distinguish between algorithms guided by analysts and thus subject to FUD and bias and what is really going on. If you haven't read "Dark Pools" I recommend it. It is a fascinating account of how computers changed Wall Street. I posted this in 2015. I am not sure if "cars" on AFB were ever part of an HFT algorithm, or if the algorithms would be smart enough to distinguish between cars as an indicator of euphoria and "cars" as a discussion of future products. I would imagine that any influence that the variable might have would be overwhelmed by the general behavior of the market and analysts' opinions, but it is all guesswork. The predictive equations are a model and as we all know....All models are wrong, some are useful. Oh no! Now the weekly threads are all out of order! This is an interesting one to bop back to. Unfortunately our local library doesn't have "Dark Pools", but it does have "Flash Boys" so I added that to my list. Isn't part of this all about data mining vs analysis, where just searching for something in the statistics is mining, but analyzing once you have reason is justified. It's more of an issue with limited data, or data across timeframes that change a bit. Often in trying to find safe withdrawal rates on 30+ year timeframes this sort of thing is hit. It wasn't until 20-25 years ago that more people started being able to easily trade individual stocks, whereas before then maybe you were in a mutual fund. And then there's mutual funds vs ETFs. And that's not even getting into the massive changes in monetary policy, in many ways. My math teacher in HS, while doing the logic class, would point out the difference between "If it rained outside, the ground is wet" vs "The ground is wet, so it rained outside". Likewise, it's sometimes pointed out that those that floss (no schedule mentioned) live longer. But, do they live longer from flossing (plausible), or is it the type of people that floss that also do other things that extend their lifetime? A problem with statistics and the market is that people are trying to beat out their competitors, trying to find a way to be a few minutes, hours, days or more in front of the trend. It's speculative, with the easiest digestible data to me being the slow but steady march forward on a MWSF rally, going from being tied to the keynote, to stepping forward by days, to then being reset. The same thing happens with earnings, where it steps earlier and earlier as long as things go well, but gets reset when there is a stumbling point. The supercomputer algos make it so "they" don't have to try to be days in front, as just a few minutes, seconds, or even microseconds can still make a tradable difference. Compare that to those of us that have to think about something a bit, and maybe drag our feet. And that's still fast compared to some of those out there that still get their daily news through newspapers, radio, or TV. It may seem out of place in this day and age, but I still see the "day after" move, especially in the first 30 minutes of trading. I'd be interested to read about those dark pools. People and groups are still looking for an edge. Only a few years ago I knew people in town working for a group that dealt with arbitrage, I think especially related to cross-currency advantages. Apparently in the area they were in at that time, there was still room for human operators to take advantage of this. There's always going to be something, as stocks in a company go from undervalued to overvalued and back. Emotions take hold, even as the likes of "The Unemotional Investor" tried to find ways to avoid it. In reading historical pieces, such as about various bubbles and manias, or certain operators like Gould and Jesse Livermore that might try a variety of schemes that would even move up to cornering a market, it does seem to be such a different time now. And yet if you start looking as certain sub-cultures, especially in small market cap stocks, you can still see it vividly. A fluff piece on a penny stock? Investment newsletters of various quality. Tiny mining companies that are so intertwined with each other that it's hard to make sense of them. Sometimes the liquidity is so low that the stock can really hop, like LODE jumping up with a 100% daily gain from $1 up to $2 for a couple hours, likely as someone wanted in with a larger position after the company finally turned a profit again. It still happens with the big companies. It's just that it's not the same crazy percentages, and so a little harder to peg the needle. And harder to point to just one thing that changed it, whereas the other side of the coin is that it would be harder to get in trouble with the SEC even if you were a big fish, since you could blend in with all of the other trading. Thanks for reminding of this, along with a good book to take a look at. EDIT: There's a review on seeking alpha seekingalpha.com/article/719521-book-review-dark-pools-by-scott-patterson Between it and the comments, it reminds me that it all depends on your investment mentality, and that behavioral science is still important. If getting into a position for the longer term, Dark Pools don't matter much. OTOH, throw in some psychology and learning from other fields, and the algos could find a way to juice a stock a little more on the upside, or sink it a little more on the downside. If you don't have sitting limit orders, or get too freaked out and buy/sell, it doesn't matter much. But for those looking at shorter term trading, it is one more thing out there. And it will probably be trying to go against you, whereas they can still make mistakes and give opportunities, like being on the other side of the flash crash by buying when the chips were down. I know I nearly did, and that was while being onsite at a job that didn't want you using the internet, and before I had an iPhone.
|
|