Asshat wormie
2023 Asshat Award Winner
- 16,820
- 30,964
You will get on average 3.5 given infinite number of rolls. If you roll your dice five times, you are not getting an average of 3.5.
Investing General Discussion
- Thread starter Furious
- Start date
Gurgeh
Silver Baronet of the Realm
- 4,327
- 11,813
If you roll a dice once, you have an average of 3.5. If you roll and infinite number of dice, your average is still 3.5, but the likelyhood of getting 3.5 is 1.
- 1
Asshat wormie
2023 Asshat Award Winner
- 16,820
- 30,964
Gurgeh
Silver Baronet of the Realm
- 4,327
- 11,813
Because you know the roll you're going to get before investing ? If you're investing on a single 6 faced dice roll, on average you're getting 3.5. I mean, I hate to have to link a wikipedia page, but in that case of high school mathematics, I kinda have to.
Discrete uniform distribution - Wikipedia
I
Asshat wormie
2023 Asshat Award Winner
- 16,820
- 30,964
The average of the rolls converges to expected value given infinite number of rolls (strong law of large numbers). If I roll 5 times and get 4, 5, 4, 6 and 2 my average sure as fuck isn't 3.5.Because you know the roll you're going to get before investing ? If you're investing on a single 6 faced dice roll, on average you're getting 3.5. I mean, I hate to have to link a wikipedia page, but in that case of high school mathematics, I kinda have to.
That being said, it's true that using "expected value" could be less confusing, but, if you click on "mean" on the wikipedia page, you'll notice that it lead you to the "expected value" page, and mean and average are synonyms (mean vs. median vs. average on Vocabulary.com).Discrete uniform distribution - Wikipedia
en.wikipedia.org
I
You're confusing the distribution mean with the actual average of your sample from that distribution. If I roll a die once and get a 1 my average is 1, not 3.5. You need a very large number of rolls for your sample mean to converge satisfactorily towards the distribution mean.
Last edited:
- 1
Gurgeh
Silver Baronet of the Realm
- 4,327
- 11,813
If we're discussing finance strategy, I'm guessing we're talking ex-ante not ex-post. If I consider a 6 faced dice roll and I have to invest on it, I can't know what roll I will get, and the average, or mean, or you're right more accurately the expected value is 3.5. It's true that it is somewhat ambiguous, but it should not be in the context, and even the wikipedia page is making that "mistake" of using indifferently mean, average and expected value.
Gravel
Mr. Poopybutthole
- 36,352
- 115,436
"Average" and "mean" measure the central tendency of a set of numbers and have no meaning if you haven't sampled your distribution. The mean of zero samples is undefined. The mean of n > 0 samples is the sum of your observations over n and not the mean of the distribution. (Yes, it's fine to talk about the mean of a distribution when you have no observations, but then you're implicitly talking about a theoretical situation where n = infinity.)
And if we're going for full pedantry points here we need to be clear that we're talking about the arithmetic mean and not the weighted, geometric, or harmonic means or some other measure of central tendency.
Last edited:
Gravel
Mr. Poopybutthole
- 36,352
- 115,436
The problem here is Gurgeh sounds like someone who just took a finance class and is getting things mixed up. I'm assuming he's likely talking about portfolio theory which says that once you invest in around 15 companies, you eliminate market risk. But then he's trying to draw an equivalency to that with an index.
And it's making him look fucktarded.
And it's making him look fucktarded.
Gurgeh
Silver Baronet of the Realm
- 4,327
- 11,813
Expected value - Wikipedia
You're the ones being fucktarded, especialy given the context of the discussion.
Frenzied Wombat
Potato del Grande
- 14,730
- 31,802
You don't invest in individual stocks these days unless you are a professional investor or you just want a few independant stocks to round out your portfolio of ETF's and mutual funds. Buying a bunch of random stocks to try and emulate an ETF just so you can maybe save a 0.3% expense ratio is retarded.
That is true. FOR THE DISTRIBUTION. Your observations/investments are NOT the distribution. They are samples FROM the distribution and have their own importantly different mean.
Gurgeh
Silver Baronet of the Realm
- 4,327
- 11,813
They are distributions, you moron, if they were sample you'd know the outcome. When you invest you don't know the outcome, obvously.
Asshat wormie
2023 Asshat Award Winner
- 16,820
- 30,964
And this is the heart of your confusion. Your sample is made at the time you commit to the investment, not the time when you discover what the outcome actually was/will be. There is an information lag, but for all useful purposes the outcome is fixed when your money goes in.
If you bet on a single die roll that happens next week, are you saying the mean of that roll after it's executed will be 3.5? It's true that the expected value of the roll is 3.5, but the ex-post mean, which is the only thing that matters, cannot be 3.5 by definition.
Last edited:
Gurgeh
Silver Baronet of the Realm
- 4,327
- 11,813
So, if I have to invest on a 6 faced dice or a 20 faced dice, the amount I'm ready to pay for that investment doesn't depend on the expected value ? The results of the rolls are irrevelant to what i'm willing to bet. By nature a stock is a random variable, the moment I'm investing on it, I'm investing on something that has a mean and a variance that I might somewhat know, but I certainly won't know what it's going to be worth in 1 year.
If I'm picking a random stock from the dow jones (with the proper probability) the expected value of that strategy is the same as the expected value of the dow jones. You might argue that it is not the same once I actually picked one at random, but the strategy of picking a stock at random has the same expected value as the dow jones. The variance is not the same, but if you invest twice a month for a few years this isn't even an issue.
You're right to say that, it might be a lot of trouble for a very small and long term gain, but :
1) There is not 1, 2 or more intermediate between you and your property (the shares), or worse some synthetic might not have any shares backing them.
2) You aren't bound to a single intermediate to lower your fees.
3) You can make your own "index", for example I've been taking more industrial and less financial than the "index" I'm picking from (CAC40), so you can have a somewhat mixed strategies easily.
EFT have their use, as I said, I'm not going to buy shares from a foreign market directly, if I want to buy a chuck of korea, I'm getting an ETF obviously. But I find it a bit concerning that so many people believe that EFTs are the only way to invest, especialy for people investing on regular basis, as the service they provide is extremely limited (see the forbes article I linked, that state that playing at random outperformed any other strategy... ok, ex-post)
If you want to invest on the dow jones, getting two dozens of stocks from the DJ is going to be good enough, on average (or expected value) you'll get the same, and it seems that you're even getting a lower variance ...
Sanrith Descartes
Veteran of a thousand threadban wars
- 41,473
- 107,527
Listen, If you want to throw darts at a board and invest on the outcome then go ahead. Others might prefer to look at cash flows, balance sheets and the like. For me personally, your method means there is a percentage chance you will actually throw your money at a company like Blue Apron, General Electric, Kraft Heinz, or JC Penny. I prefer to rule out companies that are basically shit.
As for the question of low cost ETFs and how they make money, look at the AUM of the SPY and then multiply by it 4 basis points to see how much they take in a year for an algo to make decisions on matching the SP500 and a couple of dudes to sign documents. Hint... it is a fuck ton of money.
As for the question of low cost ETFs and how they make money, look at the AUM of the SPY and then multiply by it 4 basis points to see how much they take in a year for an algo to make decisions on matching the SP500 and a couple of dudes to sign documents. Hint... it is a fuck ton of money.