[Boards: 3 / a / aco / adv / an / asp / b / bant / biz / c / can / cgl / ck / cm / co / cock / d / diy / e / fa / fap / fit / fitlit / g / gd / gif / h / hc / his / hm / hr / i / ic / int / jp / k / lgbt / lit / m / mlp / mlpol / mo / mtv / mu / n / news / o / out / outsoc / p / po / pol / qa / qst / r / r9k / s / s4s / sci / soc / sp / spa / t / tg / toy / trash / trv / tv / u / v / vg / vint / vip / vp / vr / w / wg / wsg / wsr / x / y ] [Search | | Home]

How are Markov Chains used in economics? Don't give me

This is a blue board which means that it's for everybody (Safe For Work content only). If you see any adult content, please report it.

Thread replies: 29
Thread images: 12

How are Markov Chains used in economics?

Don't give me a simple example. Give me actual, REAL calculations.
>>
>>981697
Why does that diagram look like the female reproductive system?
>>
>>981697
one application is using HMMs for pattern recognition/financial time series forecasting
>>
>>981697
>tfw I'm about to get banned
>>
File: aee4e4d2.png (17KB, 200x200px) Image search: [iqdb] [SauceNao] [Google]
aee4e4d2.png
17KB, 200x200px
>>981697

They're useful for describing transitional relationships between market regimes and are implicitly useful for anticipating the next regime change.

This is extremely useful if you have several strategies that tend to ebb&flow in profitability relative to each other and you want to model long term future performance. Like a more sophisticated Monte Carlo simulation.
>>
>>981787
What are some finance related jobs where I'll be doing this sort of math stuff ALL DAY LONG?
>>
>>981697
lol dude isnt that a vagina
>>
File: Saddam-southpark.png (47KB, 200x150px) Image search: [iqdb] [SauceNao] [Google]
Saddam-southpark.png
47KB, 200x150px
>>981801

Like... none.
You code it once into the software and it's done.
We have computers now, get with it. If you want to math all day then get into programming.
>>
>>982086
I'm studying applied math which is basically a combination of programming and pure math. Tell me about things for jobs for in the future?
>>
>>982089

Jobs are for chumps.

If you can code, and you're into math: get into algorithmic trading. If you're good at it, you don't need a job, you'll just need a bankroll.
>>
>>982103
I'm becoming an algotrader.
>>
>>982110

The force is with you, but you are not a Jedi yet.
>>
>>982373
How do I become like you?
>>
>>981697

The are mostly called regime switching models in academic finance jargon. Check out this paper and those that refer to it:

http://www.nber.org/papers/w17182
>>
>>982751

Just do it.
>>
>>982751
get a job as a researcher/algo trader

you'll have a much higher probability of success and will likely learn a lot more if you're initially learning from successful people at a profitable prop firm or fund than if you just start out on your own
>>
>>982973
Fuck yeah I'll do this. What else does it require next to a degree in Applied Math? Should I minor in Finance, or Mathematical Finance or something? How to increase my chances of getting hired at such a firm?
>>
>>983435

>Degree
>Getting Hired

You're not ready.
This isn't mechanical engineering.

This is something that you have to sit down and figure out for yourself.

Those that can do it aren't teaching it, they're getting stupid-rich without leaving the house.
>>
File: advice1.jpg (75KB, 400x398px) Image search: [iqdb] [SauceNao] [Google]
advice1.jpg
75KB, 400x398px
>>984443
hi, its you again
i read the book you recommended and got some inspiration for new algos, thx
>>
>>984447

Happy to help.

Pardo's chapters on Optimization are especially important.

Check out this whitepaper, it's a great example of practical regime switching. Even if the strategy it's demonstrating isn't particularly profitable.

> daveab com / mr_swing/MR_Swing_NAAIM_2010.pdf
>>
>>984454
i don't have enough progress to incorporate regime switching, most things i developed have too high risk or don't beat the market well enough to be worth the hassle
at least i always have new ideas and techniques to experiment with
shit's been going on for years...
>>
>>984454

why are these algos limiting themselves to local data?

it's like these people can't handle more than two dimensions.
>>
>>984443
I know you're the rain man who was in a thread I made about a week ago dealing with learning algorithmic trading as a hobby.

I've read a bit of The Evaluation and Optimization of Trading Strategy and it is a great book, don't get me wrong; however, he seems to stress a lot about Walk-Forward Analysis and it seems like half of the book is just attempting to hammer WFA into my head. Although I haven't gotten to the chapter regarding Optimization yet, so I could just be looking through rose-colored glasses.

I just wanted to ask you if this book list: (https://www.quantstart.com/articles/Quantitative-Finance-Reading-List) is anything special, or if I don't even need to understand how a quant thinks or any of that bullshit; that I just need to know how to implement and optimize strategy.

I appreciate your help though, since your comments have inspired me to grow more interested in algo.
>>
>>984454
How did you git gud? Online PDFs or did you go to college for something?
>>
>>984465
What if the other regime was cash? Might take some of that risk away. Try running a moving average through your performance curve and only trading your risky strategy when it's performing above the average.

>>984485
Practicality. Reducing the number of variables you're optimizing to <=2 keeps the compute time reasonable for long backtests and also carries an implicitly higher probability of selecting the optimum outcome due to a much smaller optimization space.

Sounds like you've got an idea, try it out!

>>984491
WFA is extremely important and it's why I tell newbs to read that book.

I've read about half the titles in the Quant/HFT list. I noticed one or two are the same books I've already suggested to you. Read everything you can find on the topic, you never know what might inspire you, but once you've got the basics the rest is just practice.

>>984494
If you keep doing anything long enough, eventually you'll stop sucking at it. ~ Andy Weir
>>
>>984575
>Sounds like you've got an idea, try it out!

no motivation.

I develop industrial regression subsystems, and thought the skills might be applicable.

but tbqh the market is too slow and boring imo to actually start developing something and scrape all the apis together.
>>
>>984575
You could write a brief summation of your advice and recommendations in the literature and post it here >>984686 for us anons who were not in any of the other threads.
>>
File: Uterus-2015135.png (27KB, 500x375px) Image search: [iqdb] [SauceNao] [Google]
Uterus-2015135.png
27KB, 500x375px
>>981697
Looks like a uterus/vaginal channel to me.
>>
>>984698
I like the idea.
Thread posts: 29
Thread images: 12


[Boards: 3 / a / aco / adv / an / asp / b / bant / biz / c / can / cgl / ck / cm / co / cock / d / diy / e / fa / fap / fit / fitlit / g / gd / gif / h / hc / his / hm / hr / i / ic / int / jp / k / lgbt / lit / m / mlp / mlpol / mo / mtv / mu / n / news / o / out / outsoc / p / po / pol / qa / qst / r / r9k / s / s4s / sci / soc / sp / spa / t / tg / toy / trash / trv / tv / u / v / vg / vint / vip / vp / vr / w / wg / wsg / wsr / x / y] [Search | Top | Home]
Please support this website by donating Bitcoins to 16mKtbZiwW52BLkibtCr8jUg2KVUMTxVQ5
If a post contains copyrighted or illegal content, please click on that post's [Report] button and fill out a post removal request
All trademarks and copyrights on this page are owned by their respective parties. Images uploaded are the responsibility of the Poster. Comments are owned by the Poster.
This is a 4chan archive - all of the content originated from that site. This means that 4Archive shows an archive of their content. If you need information for a Poster - contact them.