FinShiksha - Calculating Annualized Standard Deviation from Stock Prices

[@kaustavsen]

9 years later, still serving students across the globe. Thanks a ton for the on point video.

[@jeenthomas8147]

Thanks a million for this precise, to the point video. You're a life saver!

[@nunolopes3910]

That was really hellpfull, finally understood something that has been bugging me for a while. Thank you.

[@KennTollens]

Thank you for the very simple step by step explanation to get volatility. I am trying to get the Sharpe Ratio and everyone else glosses over how to get this number.

[@laurentiu375]

Useful, correct but too long. There's a shortcut - i.e. you can avoid the variance calculation by simply multiplying the Daily Standard Deviation with the square root of time period. So in above example just take the StDev of 2.2% and multiply it with SQRT of 250 and you'll get the same result for annualized StDev of 35%. Similarly to get the annualized StDev from a monthly series, you compute the monthly StDev as shown in the video and then multiply it with Sqrt of 12 (as there are 12 months in an year, doh). Cheers!

[@TheAnde2lumut]

Very Helpful, simple and comprehensive! Thank you!

[@TradingHeroes]

Very useful and to the point. Thanks!

[@keysky22]

Thank you,bro !!! solving my project problem!!!

[@FinShiksha]

Thanks Yuwen. Glad to be of help.

[@managedbywhitney]

1:28 start here: standard deviation moving into daily variance

[@ramamoorthyh]

Thank you for the explanation.

[@FinShiksha]

Glad it was helpful!

[@gasher2000]

Excellent. To the point and accurate.

[@joshuaofomaja]

Thank you! Just straight to the point

[@gorandamchevski]

Great video, thank you!

[@MrDbigballs]

Thank you! Really clear and straight forward.

[@FinShiksha]

Thanks Drew!!

[@manute7891]

Brilliant!! you just made my day.

[@FinShiksha]

Glad to be of help

[@zes3813]

wr

[@afca-thfc1661]

Thank you sir

[@andreaboul3]

if you were calculating 2 year annualized volatility, when performing the calculations would you simply double the number of trading days from 250 to 500?

[@soulfulll]

You're a lifesaver.

[@FinShiksha]

:)

[@santamkafle3569]

thank you. this helped me to do my research

[@heshanpalliyaguruge6333]

Thank you so much sir

[@Atinderbal]

Thanks for the video. I have a question! If the given data is monthly for 2 year, what will be annualized variance? is it Square of Monthly Standard Deviation X Number of months (12) X Number of years (2)? Also any video tutorials to Annualizing Monthly Returns?

[@saintrroy]

That is incorrect. Check your reasoning. You already have the month standard deviation from the data, so whether that was derived from 1 year data or 2 years or n years is irrelevant. If the data is monthly, then you can now figure out how to deduce the annualized standard deviation, and therefore the annualized variance. What you wrote above is wrong.

[@gokuvegeta9500]

​@@saintrroy Thank you 👍

[@javierloa9197]

If you wanted to annualize the average daily return, would you raise by 365 (like most books) or use 252 trading days?

[@solipsist100]

can the same be done for monthly returns?

[@kwcykelvin]

thanks

[@nathancortez92]

thank you very much!!!

[@atmanirbhartrader9725]

hi, video is good but I've a doubt. where does this 35% fit in 1 standard deviation and 2 standard deviations, i.e 68.3 and 95% respectively. how to interpret that.

[@firstnamelastname890]

Sir I have annual return of 10 years calculated and it's standard deviation, I want to calculate pearson correlation coefficient, so can I just divide the std deviation by 10 and use it as x score for each year to calculate correlation?

[@kim_yeorum]

thanks alot ! :)))))) chukreya bayya

[@hasanejaz7547]

PLEASE HOW DO we calculate average annual return from monthly returns?/

[@meerathomas4734]

If the data is of more than 1 year say for example 5 years what should be done??!!.....someone please explain...

[@FinShiksha]

If it is daily data, it does not matter. It will give you daily standard deviation. You can them annualize it.

[@1ddenis1]

Hey, if i calculate daily volatility using like 10 years of data....when i get standard deviation is this average daily volatility over period of 10 years then ? If i get for example 5%, do I say that the average daily volatily over period of 10 years is 5 % or do i just say it without "average", just that daily volatility over period of 10 years is 5%. I want to calcualte average daily volatility and i don't know if thats the same. Thanks

[@chelseaf.5561]

Does this work if I only have data for about one month? I have stock prices from Februaury 2nd-March 27th of the same year, so is it okay for me to find the standard deviation the same way you did in the video?

[@googlerreviewer4368]

yea but you would multiple by the number of tradings days of your data instead of the 250 days.

[@nityekhoobchandani8201]

Using the formula for returns my value doesn't comes in percent

[@anonnochowdhury12h44]

Seriously 12years ago!!!

[@lurtsify]

This is incorrect. It creates a rough approximation of the annualized standard deviation. If a rough approximation is sufficient, you can simply take the daily value and multiple it by the square root of 250. There is no need to convert to variance, multiply, and convert back. The correct method would notice that annual returns are a product (multiplication) of the daily returns rather than a summation of them.

[@hoanglan4714]

hello thanks for your comment, yup agree that we can just take the daily SD multiplied by sqrt(250) to find the annualized SD. I'm curious what would be the method to find the exact annualized SD? And also to find annual returns, I'm not sure it would be the product of daily returns (let's say daily return is 1.5% every trading day, so the product will be (1.5%)^250 which is super small). I think we find the geometric mean of daily return and multiply by 250, or (1+mean daily return)^250 - 1. I'm not sure which one is more used in practice though

[@FinShiksha]

Hi Lurtsify, thanks for your comment. Well this annualizing is obviously an approximation using daily data, and is useful if you do not have multiple years of data available, which is an issue with most stocks and indices. If you have around 30-40 years of annual data, you can use that directly. All these calculations also assume that the returns of stocks are not on continuous compounding. It that is your assumption, the calculation will change further. But if you do not have data available for multiple years, this is a correct method for annualizing it.

[@lurtsify]

FinShiksha I figured it should be pointed out because in my experience actually using this data on projects related to CAPM and redefining the efficient frontier the difference in the computed standard deviation could be as large as 15 to 25% of the value. When looking at efficient frontiers a difference in the deviation of 15 to 25% is fairly significant. I appreciate the effort to educate students though. This is certainly much better than what most students will get from their education.

[@lurtsify]

Hoang Lan Your return formula is using 1.5%, but that is the same as using .015 in your calculations. In order to correct compounding you would need to use 1.015. Test the value for 1.015^250 and you'll find it is much much larger. When I started looking at the formulas I was actually quite confused and had to find a piece by a CFA charterholder that gave the entire formula. When looking at the formulas actually in use at successful hedge funds, they will all use my more complicated math. Ironically, a few of them will post the rough estimate method, which is faster by hand but once you program the equation into a computer you can rapidly apply it to thousands of scenarios so the advantage of the simple method is substantially reduced unless the amount of data being processed was so enormous it could actually substantially slow down the processors, but if that was the case you'd need to switch over to using data lakes (see new concept from GE) to enhance the speed of analysis.

[@PRIYANKASINGH-ij7vz]

Return kese calculate krte h

[@SweetAlexMore]

Then what does the 25% represent??? What does it mean? Is that a good thing?

[@DavidVas0032]

It was actually 35% and what it means is essentially how much the annualized return will vary from it's average annualized return. So something with a small Standard Deviation (SD) and Annualized Return (AR) of 10% will generally have a value really close to 10% every year. And take for example something with a SD of 50% and a AR of 15%, then it could be expected (about 66% chance) that that years AR will range from -35% to +65%. But this type of consideration should only be taken heavily when investing over short periods of time i.e. ranging from a few months to a few years. If you are investing for long periods of time like 15+ years, then the return you will get will be just about the same as the average AR so it is nowhere near as risky if you are investing in the long run.

[@sanjayjoshius]

Itne fast bataya kuch samja nahi.. Bhai plse ek naya step by step Wala video banao

[@vinodaxisful]

volatility scales with the square root of time and not with time. The calculation is wrong.

[@FinShiksha]

That is why the calculation has been done on variance, and then the overall square root has been taken. You could also directly multiple the standard deviation with root of 250. You will get the same answer.

[@joywithneetuofficial]

I did not understand anything sir