Python for Finance: Historical Volatility & Risk-Return Ratios

[@marjanulislam4553]

One of the great Chanel I discovered for Quant. Clearly explained

[@anuragbisht1200]

great work ! this channel is valuable than gold+platinum combined.

[@amanmanamanman]

I think you need to multiply by np.sqrt(TRADING_DAYS) while computing the sharpe ratio and sortino ratio because volatility for a day = volatility for the rolling window/sqrt(trading days)

[@davidlia1102]

Well explained!

[@QuantApplicantMattKulis]

great stuff thanks

[@camkrik5812]

Great channel! Much appreciated. Just wondering about your way of calculating log returns. The formula I know is log(pct_change+1). Your calcs don't seem to tally to the cumulative return with np.exp(log_returns['CBA.AX'].sum())-1

[@konturgestalter]

nice one

[@QuantPy]

Thanks, do you have any requests for future videos?

[@konturgestalter]

@@QuantPy definitely the direction you have now. many only focus on basics like what is vola. I ld love to see more indepth quant analysis of portfolios for sure

[@srcheekychappy]

I only wish I had your skills.

[@_el_yeyo]

Is there a specific place where you would recommend learning Python? Books or courses?

[@QuantPy]

Set yourself a project, and then I recommend using KZbin / googling / stack exchange to solve all your problems. Best way to get going is to set yourself a goal of building something. Good luck

[@user-hk1wd1uv9p]

Hi, I am in Brisbane. I remembered that you said you did your MFM in UQ. I am not sure whether you are in Brisbane, lol! I am currently studying PhD in finance at UQ focusing on quant finance research! What a coincidence!

[@avinashdas8272]

Can I ask a question - the way your initial dataset was structured, it was easy to calculate log returns and then calculate the std dev for each column since your stocks are in different columns. What if all stocks were in one column with the date index repeating from start time to end time for each stock? And also if the number of days for which you have data for each stock isn't always the same - implying you can't use a rolling window to calculate std dev? I actually do have such a dataset and unable to handle that aspect of it. I tried using pivot() to stack the unique tickers into different columns, but i'm getting an error. Any help would be appreciated

[@aarondelarosa3146]

What's the best platform to run python? What platform are you using?

[@QuantApplicantMattKulis]

jupyter and vs code... replit too all great places to begin

[@dennissawyers9916]

Is this relevant to options as well to stocks?

[@alexanderfiner7552]

is .AX only for Australian stock? in yahoo finance

[@QuantPy]

Yes, you'll have to go to Yahoo Finance and use the serach funciton to find how tickers are represented in the market you are interested in!

[@DeejayGabin]

Why did you divide your risk-free rate by 252? Is it for annualized rate? I mean, for example, US treasury bond rates or T-Bill are already annualized

[@sharangkulkarni1759]

they call me chicken little, they call me bubble boy, I go long on stock which shows highest positive values on anti sortini, this is my sortini but for positive volitily, I buy them i see them down, its just like climbing steadily over mountain. ahahaha!

[@sharangkulkarni1759]

they call me chicken little , they call be bubble buster, i short sell the stock which shows highest negative values in sortino all the time, i short them instantly i see them up. its just means falling like slow river, slow waterfall ahahahaha !!!!