Incorporating Quantitative Methods To Your Trading

C

cdery6

There is an extensive number of posts regarding quantitative trading, but what if you are a nonquantitative trader wishing to incorporate ever increasingly sophisticated data analysis to study the past and understand how to address risk management issues?

I have intermediate skills in Python, SQL, R, and I have been spending months trying to look for the right approach and I feel like I’m just not making significant progress. It was about time I post this online.

I have tried to build infrastructures in AWS using APIs and scrapers happy to connect to Power BI to interactively visualize various scenarios, use open-source exploratory data analysis tools, but this means managing a process. Isn’t there a simpler way to go? Maintaining such a system has been very time-consuming and distracts me from investing. I think I will be using a more complex system down the line (I advise small hedge funds), but I need something that I can work with NOW.

Examples of studies I would conduct:

I would like to know how often a particular event occurs

I would like to know which companies and when have gone through a specific event

I would like to know return and volatility metrics as a result of an event being triggered

Find the optimal holding period to maximize profits
 
cdery6 said:
There is an extensive number of posts regarding quantitative trading, but what if you are a nonquantitative trader wishing to incorporate ever increasingly sophisticated data analysis to study the past and understand how to address risk management issues?

I have intermediate skills in Python, SQL, R, and I have been spending months trying to look for the right approach and I feel like I’m just not making significant progress. It was about time I post this online.

I have tried to build infrastructures in AWS using APIs and scrapers happy to connect to Power BI to interactively visualize various scenarios, use open-source exploratory data analysis tools, but this means managing a process. Isn’t there a simpler way to go? Maintaining such a system has been very time-consuming and distracts me from investing. I think I will be using a more complex system down the line (I advise small hedge funds), but I need something that I can work with NOW.

Examples of studies I would conduct:

I would like to know how often a particular event occurs

I would like to know which companies and when have gone through a specific event

I would like to know return and volatility metrics as a result of an event being triggered

Find the optimal holding period to maximize profits
More...
%%
I would not worry about it;
even an average or above plan gets goofed if to many people do that...…………………………….
Most profitable traders/inverstors study price ……...
200 day moving average helps;
Market still going up,mostly except on SPXS, DXD...………….
Stock Traders Almanac helps;OCT tends to be up+ when its not=short term hint.:caution::caution::caution::caution::caution::caution::caution:
 
murray t turtle said:
%%
I would not worry about it;
even an average or above plan gets goofed if to many people do that...…………………………….
Most profitable traders/inverstors study price ……...
200 day moving average helps;
Market still going up,mostly except on SPXS, DXD...………….
Stock Traders Almanac helps;OCT tends to be up+ when its not=short term hint.:caution::caution::caution::caution::caution::caution::caution:
More...

I will definitely study price but it needs to fit my criteria which requires various data points about companies(Fundamentals, market data, stock data, etc.)
 
panzerman said:
Why do you think Quantopian is going under? Steve Cohen is tired of funding all the retail quant systems that just don't work or add enough alpha.
More...

If efficient, I am happyy to pay a fee to access those systems. Seems like there are other systems with a different bussiness model surviving.
 
cdery6 said:
There is an extensive number of posts regarding quantitative trading, but what if you are a nonquantitative trader wishing to incorporate ever increasingly sophisticated data analysis to study the past and understand how to address risk management issues?

I have intermediate skills in Python, SQL, R, and I have been spending months trying to look for the right approach and I feel like I’m just not making significant progress. It was about time I post this online.

I have tried to build infrastructures in AWS using APIs and scrapers happy to connect to Power BI to interactively visualize various scenarios, use open-source exploratory data analysis tools, but this means managing a process. Isn’t there a simpler way to go? Maintaining such a system has been very time-consuming and distracts me from investing. I think I will be using a more complex system down the line (I advise small hedge funds), but I need something that I can work with NOW.

Examples of studies I would conduct:

I would like to know how often a particular event occurs

I would like to know which companies and when have gone through a specific event

I would like to know return and volatility metrics as a result of an event being triggered

Find the optimal holding period to maximize profits
More...

You certainly don't need AWS / PowerBI / Python / SQL / R for all of that, they are actually often a huge distraction and you're absolutely on a right track spotting that early on and asking those questions.

It is likely that there are specialized data providers offering this data and you just need to hook them up with a backtesting tool. Check AmiBroker and some of data providers it supports. When I was doing discretionary trading I would often use a scan for AmiBroker, where I would quickly type in an extreme event I am seeing unfolding as a formula and will have it got thru whole US stocks universe over last 50 years to see how many times it occurred in a past and what followed. There were like 20 columns in it. This day close measured in some sort of average or recent range, next day open, next day high/low, same, all normalized to that one unit of measurement to give an idea on best / worst case scenarios. Then 5d, 10d, 20d and lots of extra stats to help size up probabilities, optimal bet size and profit/loss target. This is just one example.

Val
 
cdery6 said:
There is an extensive number of posts regarding quantitative trading, but what if you are a nonquantitative trader wishing to incorporate ever increasingly sophisticated data analysis to study the past and understand how to address risk management issues?

I have intermediate skills in Python, SQL, R, and I have been spending months trying to look for the right approach and I feel like I’m just not making significant progress. It was about time I post this online.

I have tried to build infrastructures in AWS using APIs and scrapers happy to connect to Power BI to interactively visualize various scenarios, use open-source exploratory data analysis tools, but this means managing a process. Isn’t there a simpler way to go? Maintaining such a system has been very time-consuming and distracts me from investing. I think I will be using a more complex system down the line (I advise small hedge funds), but I need something that I can work with NOW.

Examples of studies I would conduct:

I would like to know how often a particular event occurs

I would like to know which companies and when have gone through a specific event

I would like to know return and volatility metrics as a result of an event being triggered

Find the optimal holding period to maximize profits
More...

I like your questions. I'm actually on the same path. If you want to team up to share, just drop me a private message. Cheers
 
You must log in or register to reply here.
Back
Top