1
$\begingroup$

I used Principe component analysis on yield curve data this was the result

   PC1          PC2         PC3         PC4         PC5         PC6
1 MO  0.06343397  0.539948899 -0.01295235  0.48033733 -0.39805007 -0.32974832
2 MO  0.08366276  0.488494834  0.08271174  0.15205995 -0.14343924  0.09932244
3 MO  0.10556314  0.433790733  0.10996224  0.13652605  0.57093458  0.41559098
6 MO  0.16617630  0.382170715 -0.07112618 -0.43921427  0.36857147 -0.07944536
1 YR  0.25867698  0.226049937 -0.36725405 -0.57548373 -0.20590395 -0.30838927
2 YR  0.34789322 -0.037574785 -0.38928757  0.03715630 -0.23639526  0.48581201
3 YR  0.37198594 -0.138197458 -0.33817669  0.16876088 -0.09745012  0.28721603
5 YR  0.37913405 -0.161349395 -0.13956382  0.22980595  0.20514848 -0.19157065
7 YR  0.37754795 -0.147318409  0.02674855  0.22557977  0.26041966 -0.29500317
10 YR 0.37383343 -0.107436709  0.21950977  0.05548224  0.11804235 -0.30672779
20 YR 0.32959443 -0.068533577  0.42923007 -0.10170653 -0.08934241  0.01546754
30 YR 0.29871813  0.003783446  0.56783802 -0.23757826 -0.34579482  0.26592482

I would like to know how to find which parts of the yield curve are cheapest and most expensive using PCA

$\endgroup$
  • 1
    $\begingroup$ A good tutorial from Tororono Dominion's research tdsresearch.com/currency-rates/… explains how you might use principal components to make a rich-cheap grid. $\endgroup$ – Dimitri Vulis Oct 6 at 2:58
  • 1
    $\begingroup$ Thanks for the above. I quite like the section on macro correlations & PC's which is something more technical literature tends to omit. $\endgroup$ – oronimbus Oct 6 at 10:28

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Browse other questions tagged or ask your own question.