What kind of math is idea for financial modeling/finding patterns

I'd start by working on my understanding of what random noise is in the financial markets.

>>8189875
>technical data

Sound financial practice would not rely completely on technical data to predict the market.

Econometrics is retarded

ODEs, PDEs, stochastic processes, probability, functional analysis, numerical methods

>>8189948
In what sense? The mundane definition of the noise is variations in how people value the stocks, but how can I learn to quantify this mathematically?
>>8189952
I agree. It helps tremendously to cross-reference what the theoretical trend is doing with what is realistically happening.
Now, which classes are ideal for mathematically imagining the data of stock market trends? What classes are ideal for trying to use functions to "guess" the noise and general trends? Does it exist?

>>8189958
/thread

File: 777.jpg (78KB, 640x426px) Image search: [Google]

78KB, 640x426px

>>8189962
Characteristics of the probability distribution underlying the fluctuations, such as deviation depending on time, distribution of jump-sizes. E.g. compare with
https://en.wikipedia.org/wiki/L%C3%A9vy_flight

The Brownian notion has the advantage of having a calculus attached to it, etc.

These subjects (>>8189960) are a good answer. All the topics you listed seem to share tools needed to understand stochastic processes (e.g. topology smells a bit like measure theory and differential geometry smells a bit like analysis)

PS: Then again, I'm of the opinion that coming from another knowledge base is good to come up and do something that hasn't been done before.

PPS (assuming it's you again), I find it good that you make these threads. Math threads with an agenda and motivation.
How about you make some educational threads, where you present your notes publically, be explanatory, raise questions in context, etc.

>>8189958
>>8189969
>implying this is econometrics
>>8189960
Thanks.
>>8189979
Thanks for the link. I'm doing a stochastic class in the fall.
I'm not sure how useful my notes would be. What subject or set of ideas do you imagine the notes would be on?

>>8189962
I guess my point is that the random noise that you want to replace in Black-Scholes is already a mathematical representation of the unpredictable volatility of market valuations.

File: stird.png (338KB, 1237x867px) Image search: [Google]

338KB, 1237x867px

>>8189992
>What subject or set of ideas do you imagine the notes would be on?
huh?

The subject would be your learning project. The content is exactly the insights you gained that were not part of your uni curriculum. The stuff that makes you see the light.

Lay out your long term goal here, plz.

>>8190022
I think I understood, intuitively, that the Black-Scholes theorem's random noise is computed from an assumption that it is "unpredictable volatility."
I'm challenging that very definition: the volatility is predictable.

>>8190034
I think what you are working on is interesting. You should take advantage of the fact that many, many people have already looked at the same thing and you should spend more time researching the literature. This will save you a lot of time in the long run and will also give you an idea of the math you will need to pursue your ideas.

>>8190024
oh ok. My thinking went in about 5 different ways when you made that request.
I'd like to work towards abolishing most or all random variables. It would be hard to post my notes right now because it's a lot of abstract thinking. I guess the core idea though is that every market acts like a living organism and is similarly unique in the way it will respond to its circumstances. Also similarly to any living organism, a market is limited in its range of responses to a very fine degree based on its characteristics. By analyzing past behavior of a particular market, one can determine its "personality," and predict how it will respond to future stimuli. By transposing these ideas into mathematical ideas, one can logically deduce a market personality based on its past responses, as well as how that personality will mutate as new variables are introduced and it is exposed to new environments.
Again, it's a lot of abstract thinking right now. I'm seeking out and learning the math that would be useful in understanding the potential range of responses of a market.

File: maxresdefault[1].jpg (91KB, 1280x720px) Image search: [Google]

91KB, 1280x720px

>>8189992
Explain to me how it is not econometrics

>Econometrics is the application of mathematics, statistical methods, and computer science to economic data

>>8190075
Because econometrics, in its current form, assumes reliance on empirical data. The focus here is to create mathematically-based theoretical models.

File: memes everywhere.jpg (75KB, 640x426px) Image search: [Google]

75KB, 640x426px

>>8190048
But a market isn't a closed system, you wouldn't be able to include a Brexit happening in a year, and the protagonists and players of the market and their ways and goals change.
The Brownian term is essentially already the one simplest to understand, and there are a million models and extension of the finance math already, I'm sure.
You're idea don't sound extremely fruitful to me, family.
Besides, what do you do with the models? If it's a profit, you'll need to keep on getting your feet into the system, and to just get jobs and thus access to this world, you'll have to do internships and the certified programs. Just sayin'

>>8190142

Not gonna happen, determinism.

>>8190189
>But a market isn't a closed system
Of course not.
>you wouldn't be able to include a Brexit happening in a year
"Brexit" never happened, and in all likelyhood it won't because noboy would risk initiating it.
> there are a million models and extension of the finance math already
Of course there are.

>>8190233
>implies determinism is false
>makes a deterministic statement
Wew lad

>>8189875
Fuck off we're full.