Covariance matrix shrinkage: Ledoit and Wolf (2004)

Sample covariance matrix applications in portfolio optimisation are often criticised for the excessive noise that such matrices contain which results in "estimation error maximisation" and unrealistic optimal portfolios. Covariance matrix shrinkage as proposed by Ledoit and Wolf (2004) is one of the most commonly used and of the most elegant remedies for this issue. Today we are investigating how to shrink a real-world asset return covariance matrix, discuss the concepts behind the procedure, and how an application of shrunk versus non-shrunk matrix impacts portfolio management.
Don't forget to subscribe to NEDL and give this video a thumbs up for more videos in Investment Managemet!
Please consider supporting NEDL on Patreon: / nedleducation

[@NEDLeducation]

You can find the spreadsheets for this video and some additional materials here: drive.google.com/drive/folders/1sP40IW0p0w5IETCgo464uhDFfdyR6rh7
Please consider supporting NEDL on Patreon: www.patreon.com/NEDLeducation

[@scotthicks1795]

Maybe a segment on stock-based compensation? Restricted Stock Units (RSU), Performance Stock Units (PSU), and Employee Stock Purchase Plan (ESPP) seem to use a variety of methods. Not sure how cohesive it would be, but it does show up in the notes in earnings reports, so probably good to know.

[@Trivitu]

Awesome video as always!!! Is it better to use this method to estimate the covariance matrix and use it as the initial covariance matrix in the DCC GARCH framework? Also, if possible, coud you do a video showing how to estimate a Flexible DCC GARCH (Billio, Caporin and Gobbo, 2006)?

[@ashaykakde583]

Hello Sava... thanks for sharing these videos.. can you also share a video on pricing of Range Accrual Swaps... Thanks a lot...

[@jasonwayne4647]

Hello Sava, Is there a spreadsheet to go with the video - cant see one on the drive

[@davidghope1]

Is there a spreadsheet to go with the video - cant see one on the drive

[@Bystander_88]

would you follow the same procedure with daily prices/returns?

[@drek273]

can it be used for a handful of stocks like 3 maybe?

[@jasonwayne4647]

legendary

[@PanagiotisPampakas]

I believe there is something wrong in your spreadshit, because Delta is defined as the frobonious norm, which means that you need to devide your delta estimation by the number of assets.