Down goes the Dow

Gruntz wrote:

Ghost of Igloi wrote:

?????

Looks like Igy lost yet another argument. LOL!

Hey Detector Dude, you’ve gone from witchcraft back immature computer games. It all fits your crazy persona.

I don’t recall Tokyo Joe, but I do the boiler room calls with a hot stock tip. You were “broker of the day” and get this hot tip from “small cap connection,” or some other shop. Often out of Florida.

Ghost of Igloi wrote:

Gruntz wrote:

Looks like Igy lost yet another argument. LOL!

Hey Detector Dude, you’ve gone from witchcraft back immature computer games. It all fits your crazy persona.

I have no idea what that means. Are you drunk posting again?

Racket wrote:If you want to up your game and actually understand how certain financial instruments behave then let me know. Otherwise, enjoy tilting at windmills.

Racket,

the ability to use a bunch of math terms in a handful of sentences doesn't demonstrate that you know HOW to do something. Don't tell me about mathematical theory, or what you THINK others can do, give us a demonstration of your chops. Or of somebody's chops.

Here are two softball challenges:

- using information available up until, say, mid-February 2020, show that it would have been possible to make a confident prediction of market value on 23 March

- using the same information, and no more, show that it would have been possible to make a confident prediction of market value on 8 June

When you can show me that, I will accept that you have understood the math terms you've been throwing around.

BTW - I'm not sure if you think it bothers me to call what I've presented as "cute." I used the term "naive" to be clear about what I had done. Which, by the way, did not include any linear regression. It was way "cuter" than that; all I did was look at the whole set of daily changes (as %) and generated percentile rank, looking forward and backward over various time intervals. It was an admittedly ham-fisted analysis, and I'm quite sure I presented it as such, with no attempt to give it a sheen of theoretical rigour.

Racket wrote:Someone has told you, apparently (probably some hack in OR), that "probability" doesn't exist or is subjective in some way, shape, or form. This is incorrect. Probability was rigorously axiomatized by Kolmogorov into a consistent space aptly called a probability space (Lp spaces come into play here but I'm only familiar with them in terms of some of their applications to Hilbert spaces). The space is measurable through the probability measure.

You're getting trapped by the difference between the mathematical definition of probability and the practical notion, or what probability means in real terms to real people who need to make decisions under uncertainty.

Math does not define the real world, but often does a good job, or even a very, very good job of describing the real world. There are important cases where math (and physics) falls short. And this thread revolves entirely around one of them. Many others exist in nature.

the idiot wrote:There are important cases where math (and physics) falls short. And this thread revolves entirely around one of them. Many others exist in nature.

I might add that my professional career revolves around the practical application of math and physics in an area of the natural sciences, and using that mainly for the purpose of assessing the probabilities of specific damaging events, and advising clients who pay for advice to help make decisions in the use of resources (time, effort, money).

That said, I am not "expert" in matters of finance, I am, in this arena, admittedly an idiot... and maybe the difference between me and some others is that I know this and admit it, loudly and often. Not that I am calling anyone else an idiot... :-)

European markets were flashing red when I started my day, but have all swung hard the other way. I'd like to "predict" a great day for the US markets today (and futures are way up to support that opinion), but I don't want to jinx things... Yesterday's call couldn't have been more wrong... :-)

the idiot wrote:

I'd like to "predict" a great day for the US markets today (and futures are way up to support that opinion), but I don't want to jinx things... Yesterday's call couldn't have been more wrong... :-)

If you’re wrong again, you risk joining Igy in the Contrarian Indicators Club. ?

PRD wrote:

Gonna wait a couple days to see how things are going and then probably dump more money into an energy ETF. VDE, XLE, and IXC are all lower at present than they were at the bottom of 2009 and the lowest they've been since about 2004. They also come with a dividend yield of over 5%. Seems like a pretty safe bet unless I am completely missing something. Can't really see big oil going bankrupt any time soon even with the slow trend towards green energy.

Now this looks like decent way of playing the fundamentals in my book.

That being said, the evil is in the details. I would rather invest in individual companies than in ETFs, checking their gearing ratios (which would help you understand their exposure to financial risk), and their individual P/E, ideally comparing its evolution to average WTI for the year, in order to get a rough idea of their capacity to stay profitable in a downgraded price environment. I'm pretty sure you'll find it is worth the time gathering these pieces of info and adding it to the individual P/E ratio or yield in your decision making.

the idiot wrote:

Racket wrote:If you want to up your game and actually understand how certain financial instruments behave then let me know. Otherwise, enjoy tilting at windmills.

- using information available up until, say, mid-February 2020, show that it would have been possible to make a confident prediction of market value on 23 March

- using the same information, and no more, show that it would have been possible to make a confident prediction of market value on 8 June

GBM - choose a value for expected return and a value for volatility. Run Monte Carlo simulations (10,000 should be trivial in R). Result is a log-normal distribution of likely end points for the selected time frame. Bias is introduced when selecting aforementioned variables. If you selected variables carefully then congrats, you win. Common problems in the model include the fact that volatility changes day to day and that stock prices are occasionally non-continuous (and GBM produces a path that is continuous everywhere). Quants will then build models that cover many likely scenarios or scenarios they need to cover to protect risk exposure. For instance, what if volatility goes way up, or way down, or our expected return suddenly needs to change. No one has every claimed you could predict everyday with 100% accuracy, because as I've pointed out (let's see.... 10th time? 11th perhaps?) these are stochastic processes.

Again, this process that you claim is impossible is done everyday by very intelligent. I can literally find job openings for it on Indeed or LinkedIn. Perhaps even a book or two on Amazon, or a college course at probably a dozen or more institutions.

lol this is some next level teleological horsesh!t if I've ever seen it.

the idiot wrote:

the idiot wrote:There are important cases where math (and physics) falls short. And this thread revolves entirely around one of them. Many others exist in nature.

I might add that my professional career revolves around the practical application of math and physics in an area of the natural sciences, and using that mainly for the purpose of assessing the probabilities of specific damaging events, and advising clients who pay for advice to help make decisions in the use of resources (time, effort, money).

That said, I am not "expert" in matters of finance, I am, in this arena, admittedly an idiot... and maybe the difference between me and some others is that I know this and admit it, loudly and often. Not that I am calling anyone else an idiot... :-)

So you're an actuary which is literally just real-world applications of Bayesian probability, which means my suspicions were correct several pages ago : you're trapped trying to argue yourself in circles about frequentism vs Bayesianism.

A cute debate, but hardly applicable.

the idiot wrote:

Which, by the way, did not include any linear regression. It was way "cuter" than that; all I did was look at the whole set of daily changes (as %) and generated percentile rank, looking forward and backward over various time intervals. It was an admittedly ham-fisted analysis, and I'm quite sure I presented it as such, with no attempt to give it a sheen of theoretical rigour.

Yes, this would be the naive Bayesian model. The "sophisticated way" to do this would be to construct a rather large network but I doubt it could really be done for equity prices unless you have some sort of doomsday computer (any lurking quants, please feel free to correct me).

So again, if you want to know more about how it's actually done in the real world, then let me know!

Racket wrote:No one has every claimed you could predict everyday with 100% accuracy, ...

No indeed. And yet, you still have not produced something more accurate than the "cute" scratch work I've done that you keep criticizing. Please do try, using one or both of the scenarios I gave you. Until you do that, or at least attempt to DO something, rather than TALK about what you think is possible, you're just blowing smoke.

Racket wrote:... if you want to know more about how it's actually done in the real world, then let me know!

I think I've asked you to show me how it's done more than once...

Racket wrote:So you're an actuary...

Um, no. You're not much of a guesser. Nor a forecaster, based on evidence so far...

the idiot wrote:

Racket wrote:No one has every claimed you could predict everyday with 100% accuracy, ...

No indeed. And yet, you still have not produced something more accurate than the "cute" scratch work I've done that you keep criticizing. Please do try, using one or both of the scenarios I gave you. Until you do that, or at least attempt to DO something, rather than TALK about what you think is possible, you're just blowing smoke.

I'm sorry, am I talking to brick wall? Did you not see the part where I literally described how GBM works and how such models are produced? And the part about the quants, did you not see that either? Do I need to increase font size to 24 maybe?

You built a wood shack and claimed it's indestructible, and you won't hear otherwise unless I fully construct Fort Knox? Is that how this is going? Can you not just see that steel and concrete might be a little better than scotch tape and twigs? Because that's what's happening here.

the idiot wrote:

Racket wrote:... if you want to know more about how it's actually done in the real world, then let me know!

I think I've asked you to show me how it's done more than once...

Here's a great, legendary guide, in pdf :

Why don't you educate yourself and then you can talk sh!t

Racket wrote:I'm sorry, am I talking to brick wall?

Did you not see the part where I literally described how GBM works and how such models are produced?

I was wondering much the same thing...

Racket, don't tell me HOW you THINK something is done (by others). SHOW me (yourself). If you're incapable of doing it yourself, show me what someone else has done, similar to the challenges I've presented.

You can talk to me all day about monte carlo simulations (I run them all the time), and blow smoke about all the fancy math terminology you like (I happen to have P=NP, which I don't happen to believe, and a reference to Fermat's last theorem tattooed on my forearm; long personal story). I know you're SMART, but are you CAPABLE of using those smarts to DO what you say is possible?