Why Portfolio Diversification is so Effective (In Theory)

[@PerfilievFinancialTraining]

Hello friends! Thank you so much for watching! I’ve only recently started on KZbin, and this is one of my first videos. I really hope you’ll find it interesting and somewhat entertaining. Please, please do subscribe to the channel - at this early stage, your support has a HUGE impact, and absolutely every person counts. I am doing this full-time now, and if you want to see how it goes, it would be great to have you on board! As always, feel free to reach out for any feedback, questions and suggestions. You can ping me on Twitter or via email in the channel description. Thank you for your help and support!

[@ashutosh567]

Wow! This video has changed my mind about diversifying stock portfolio! Amazing video! Love the simulator

[@PerfilievFinancialTraining]

Thank you so much! Glad that you found it useful! :)

[@Love_sharma_01]

I am catching up with your videos mate. Two reasons. They are fun and obviously knowledge!! Great content as usual. Keep up the good work!

[@PerfilievFinancialTraining]

Thank you so much for your support! It's really great to hear! :) I try to make the vidoes interesting, as some financial concepts can get boring very quickly.

[@BrunoSchwalbach]

To make the sim look more like reality, you’ll need to incorporate a jump diffusion model (or similar). Effectively simulating crashes where the correlations between your assets tend to 1 because real life is super fat tailed vs the GBM used above. Would be really cool to see how you model this in your code, I’m sure there are cleaner ways than how I do it 🙈

[@PerfilievFinancialTraining]

Hi Bruno! Yup, exactly! The simulation generated stocks that were Normally distributed without any jump-diffusion and hence completely independent. Therefore perfect diversification was possible. If I added a jump-diffusion model to simulate market crashes or some other 'correlation' parameter, then it would be more realistic, since we wouldn't be able to diversify all of the risk away. Thanks so much for watching the whole video - I really appreciate it! :) Glad that you enjoyed it! All the best.

[@zkkrhfhska]

obviously? this is an introduction video. gbm is just a pedagogical tool.

[@danieloda1]

Greetings from Brazil! One of the best videos about diversification i've ever seen!

[@PerfilievFinancialTraining]

Wow, thank you so much for your kind words and for watching it! Really happy to hear that you liked it! :)

[@christianjodwyer]

Good luck with this venture. It looks like great content!

[@PerfilievFinancialTraining]

Thanks so much, Christian! I'll definitely need a bit of good luck with this :) If you find anything relevant or interesting, you're also very welcome to share it with the team :D

[@andrii3254]

Nice, quite a clear explanation of the drift and volatility of stocks!

[@PerfilievFinancialTraining]

Thank you, Andrii. Glad you liked it!

[@nikolaykogut7546]

Love the simulation!! Very clear way to show what exactly diversification does to a portfolio.

[@PerfilievFinancialTraining]

Thank you! Yeah, I hope the simulation made things a bit clearer. Thank you for watching.

[@tropikk512]

Quite a good explanation of what is diversification.

[@PerfilievFinancialTraining]

Thank you - I'm really happy you found it useful!

[@serhiisadchikov2647]

Isn't your portfolio equally weighted in the simulation? As opposed to S&P 500, which is a cap-weighted index?

[@PerfilievFinancialTraining]

Yes, well spotted! :) That was another issue - S&P 500 is heavily weighted towards the big tech companies at the moment and hence isn't as diversified as an equally weighted index.

[@GrantWasHere]

This was one of the best explanations for diversification on KZbin! I particularly enjoyed you explaining the concept simply on the board before heading to the simulation too. It would be great to see more videos on your top tools used to make these. For example, if you built the simulation, then a tutorual could be beneficial. Alternatively, if it was prebuilt, the websites with simulation tools would also be interesting.

[@PerfilievFinancialTraining]

Hi Grant! Wow, thank you so much for your kind words and for watching it! Really happy to hear that you liked it! :) Thanks for the suggestion - noted. Will try to do that in the future :) Thank you!

[@shonitsinghal]

IF you find a teacher who can teach difficult things in the easy format you have hit a jackpot and I think I do. Thank for the video. And that background with a ticker on the wall is awesome. Could you do me a favour and tell me where I can buy that 🙈

[@PerfilievFinancialTraining]

Hi Shonit, thanks so much for watching the video - I really appreciate it! :) The LED strip was a bit of a DIY project. It's seven of WS2812B LED panels connected together, with Raspberry Pi running the logic behind them. And the data is from Yahoo finance via yfinance Python package.

[@shonitsinghal]

@@PerfilievFinancialTraining Always love when people don't just spend on stuff and do creative DIY project . 😇 Thanks for replying though

[@PerfilievFinancialTraining]

@@shonitsinghal No problem :) I just wasn't able to find something like this to buy, so had no other choice but to get a bit creative with this :)

[@ДарьяШевчук-у2я]

That was easy to understand! Thanks for the video

[@PerfilievFinancialTraining]

Welcome! Thank you so much for watching it! :)

[@MrCat-yj5lj]

very informative video, thank you!

[@PerfilievFinancialTraining]

Thank you for watching! Glad you liked it :)

[@applepie8058]

Nicely explained, thanks very much for making these videos (already a sub!)

[@PerfilievFinancialTraining]

Thank you so much for your support and for watching it! :) I really appreciate it! Glad you enjoyed the video ;)

[@birdxi]

1. Normal->fat-tail 2. Independent->correlated Nice video btw!

[@PerfilievFinancialTraining]

Hey Zhuo, exactly, spot on! Simulation is a purely theoretical exercise with normally distributed and uncorrelated returns. Thanks so much for watching the whole video - I really appreciate it! :) Glad that you enjoyed it!

[@ritamroychowdhury]

Great video! The simulation is beautiful! The sim is missing correlation and assuming the asset returns are independent of each other

[@PerfilievFinancialTraining]

Hi Ritam, exactly! All assets were normally distributed and uncorrelated with each other - hence perfect diversification was possible. Thanks so much for watching the whole video - I really appreciate it! :) Glad that you enjoyed it!

[@ChurykidzeS]

thanks for the video! would be great to see the simulation with a correlation parameter included!

[@PerfilievFinancialTraining]

Thank you for watching! Hopefully will get some time to add a correlation setting there...

[@zakariam11]

Thank you from Sweden

[@PerfilievFinancialTraining]

Hi Zack, you are very welcome! Thank you for watching it :)

[@Tartarus116]

Great video! I think the missing component is correlation. During crises, all correlations go towards 1.

[@PerfilievFinancialTraining]

Yes, exactly! Thank you for watching! :)

[@leonardotancredi2477]

I guess the difference in volatility is due to your model assuming that the stocks are completely independent, while in reality they're quite correlated.

[@PerfilievFinancialTraining]

Hi Leonardo, yup, exactly! The simulation generated stocks that were completely independent, and hence perfect diversification was possible. If I added a 'correlation' parameter, then we wouldn't be able to diversify all of the risk away. Thanks so much for watching the whole video - I really appreciate it! :) Glad that you enjoyed it! All the best.

[@mindingthedata4218]

Super cool simulation! What tools / programming languages did you use to make the web app?

[@PerfilievFinancialTraining]

Hey, thank you so much for watching - I really appreciate it! :) Glad you liked the simulation too. It was done using the basic web technologies - front-end is HTML/CSS and the back-end simulation engine is in JavaScript.

[@mindingthedata4218]

@@PerfilievFinancialTraining Thanks for getting back to me! I've been wanting to learn how to make web apps like this and have been trying out some solutions using Python. I also want to try out HTML and JavaScript. Any suggestions for the best way to learn these?

[@PerfilievFinancialTraining]

@@mindingthedata4218 Initially, I was hoping to make a Python back-end work on the web, but couldn't find a nice/neat way of doing this. Hence decided to go with JavaScript, which is a weird choice for financial apps :) If you know Python, then HTML/CSS/JavaScript shouldn't be difficult. Starting with some basic tutorials online will get you halfway there - once you're familiar with the basics, you can start building whatever you want by googling things, as you go along. Unless you really want to get into JavaScript, in which case you might also have a look at these two: JavaScript and JQuery: Interactive Frontend Web Development by Jon Duckett JavaScript: The Good Parts by Douglas Crockford

[@mindingthedata4218]

@@PerfilievFinancialTraining Great! Thank you so much for your advice 😄

[@PerfilievFinancialTraining]

@@mindingthedata4218 You're very welcome! Good luck! Awesome channel by the way - very interesting videos :)

[@DEVILinDISGUISE6]

Great video!!

[@PerfilievFinancialTraining]

Thank you! :)

[@artemkolenchenko5883]

thank you for the video! when you diversify, isn't there a risk that losses might overweight the gains and net net you've lost?

[@PerfilievFinancialTraining]

Yes, that's also possible. However, the idea isn't to increase the returns, but to reduce the risk. If all assets in the portfolio have a positive growth rate, what should cancel are the volatilities.

[@edbrandt8972]

Good use of "prop humor" in addition to the educational content

[@PerfilievFinancialTraining]

Haha, thank you :) I try to include something fun, if I can think of it :)

[@eduardvoll516]

Great work! Spasibo;)

[@PerfilievFinancialTraining]

Thank you for watching, Eduard! I really appreciate it! :) Pozhalusta! :)

[@nevmer470]

Were the stocks generated using a normal distribution in your sim? If so, they would not show the same risk as the S&P 500

[@PerfilievFinancialTraining]

Yes, that's exactly correct. The simulation assumes independence and normal distribution.

[@random-e1u]

Sir what video program do you use to build and edit these videos

[@harryj1081]

A raw egg that can spin that fast is what’s fascinating.

[@PerfilievFinancialTraining]

Hi Harry! Hahaha, I had to practice that spin for half an hour before recording :) Thank you for watching! :)

[@АртёмСытник-ц8ъ]

Many already mentioned it, but your simulation seems to be missing correlation. Stocks are independent and that allows to diversify all the risk.

[@PerfilievFinancialTraining]

Yes, exactly! The stocks were generated as independent variables and hence a perfect diversification was possible. Thank you!

[@valeriiamushtuk2045]

Great introduction, very funny

[@PerfilievFinancialTraining]

Thank you :)

[@StudioGamingFrance]

The missing parameters couldn't be the ponderation of the portfolio ? because SP500 is highly ponderated by the large caps tell me if i'm wrong

[@PerfilievFinancialTraining]

Nice one - well spotted! :) I was actually thinking if anyone would pick up on that too, as this is also a big factor! In my simulation, the portfolio was equally weighted, whereas the S&P 500 is heavily weighted towards big tech, which can easily influence the index performance. Thanks so much for watching the whole video - I really appreciate it! :)

[@harryj1081]

Isn’t there’s a saying about google effect on s&p 500? New entrant with inflated prices and ousted member will decline more. Wouldn’t a hedging position that consist call options on the most influential position in spx and put option on the recent inflated entrant be a perfect portfolio diversification to benefit from volatility in reality, since there’s no way to achieve extremely low volatility?

[@StudioGamingFrance]

@@PerfilievFinancialTraining I'm begining in trading and finance, I can easily affirm that your videos are highly qualitative, and very understandable for a french ! so you deserve everything and continu in this way ! nice day

[@PerfilievFinancialTraining]

@@harryj1081 Hi Harry, yeah, there are many ways to play such situations, especially when you're considering options and volatility trades.

[@PerfilievFinancialTraining]

@@StudioGamingFrance Hahaha, that is so great to hear, C K :) Thank you so much for your kind words! Really appreciate it :) Have a great day!

[@elektrolabs]

hello, the video is very informative, I have a question, is it possible to use diversification in short-term trading? Is it obligatory that all positions are directed in one direction?

[@jettoth3]

I think that the main reason why the simulation that you used under estimated the actual portfolio volatility that you would observe in the S&P 500 index, is because your simulation has some built-in assumption that the individual stocks are not highly correlated. In the real (stock market) world, the reason why true diversification is so difficult is because most stocks are surprisingly correlated, even those of very diverse industries.

[@jawwdinnzgme1777]

Hi, great videos keep up the great work. Is the parameter Beta? Since you said the S&P is more volatile than the model, I think we can use beta for the individual stocks volatility instead of a flat 16% volatility rate. This might produce more accurate results, as long as the weighted average beta of the whole portfolio = 1.0 (to mirror S&P behavior).

[@edbrandt8972]

So I ask myself....why does the Russel 2000 index seem to be more volatile than the s& p 500?

[@PerfilievFinancialTraining]

Well, it's not only the quantity that matters - there are systematic/market risks that you cannot diversify away. This risk will remain. Russel 2000 is also concentrated in small-cap stocks, whereas S&P 500 consists of more established businesses.

[@baloothecarefreebear7242]

correlation of the sigmas!

[@PerfilievFinancialTraining]

Spot on, exactly! :) Thanks so much for watching the video!

[@javierloa9197]

We are missing systematic risk - Beta. Long-termVariance + (alpha + Beta)^t * (VarianceCurrentEstimate - Long-termVariance)

[@PerfilievFinancialTraining]

Yes, exactly! Thank you for watching! :)

[@albertosantangelo6872]

First, you need to change distribution, since S&P500 index returns have Pareto tails.

[@PerfilievFinancialTraining]

Hi Alberto, yes, exactly! The simulation generated independent, normally distributed stocks that were completely independent, and hence perfect diversification was possible. And yes, the simulation certainly doesn't capture the Pareto tails of S&P 500 :) Thanks so much for watching the whole video - I really appreciate it! :)

[@kevinq6628]

hi, are you planning to do a course on algo trading? i will be the first one to buy it

[@PerfilievFinancialTraining]

Hi Kevin, thanks so much for your support. In the immediate future, unfortunately, I'm not planning something like that. Maybe will get to it sometime later. But thank you in advance! :)

[@edbrandt8972]

Google told me that volatility has increased over time since technology allows us to trade much more rapidly than in the past.

[@PerfilievFinancialTraining]

Hi Ed, yeah that sounds reasonable - improvements in technology certainly contributed to that. Thanks for watching the video!

[@btc1m654]

Market risk!

[@PerfilievFinancialTraining]

Exactly! :) Thank you for watching!

[@vajliakduke6231]

Problem is in reality, when systemic risk happens, these stock's return will not be random, they're correlated. They tend to fall together in a market crisis. Meaning: no positive risk to offset negative risk. Volatility is exacerbated during bad time. Do diversify with another asset classes! Anti-correlation asset classes.

[@PerfilievFinancialTraining]

Exactly! Thank you for watching! :)

[@alexsilvertyk8589]

isn't the model assuming that stocks are independent of each other, whereas in real life they aren't?

[@PerfilievFinancialTraining]

Yes exactly! :) The stocks were fully independent of each other.

[@WillHampson]

Correlation

[@PerfilievFinancialTraining]

Yes, exactly :)