Futures Spreading Greeks... Or just term structure greeks?

I don't understand you....

 

Not really... These aren't macro models and will tell you nothing about the Fed.

 

The title in your thread started that analogy!

 

So what could pca extract out of a yeild curve...

 

I don't know that gamma translates from options to futures in the way your describing... That or I don't understand what your saying... Typically a correlation is build in relation to the front contract... As opposed to options to their underlying

 

Bear with me. I am gonna shut up for awhile. This is a great thread and I may be clouding things.

 

I just wanted you to clarify... I appreciate everyone's input

 

Well, it could potentially tell you, for any given contract, whether it is rich/cheap vs a "PCA-fitted" strip. Among other things...

 

Principal components are by construction uncorrelated in the period analyzed as a whole, but we also have the rule of thumb that extremely steep curves are followed by reduction in the level of interest rates. How do I intuitively reconcile those contradictions?

For 2 to 30y swaps I get 2nd factor inflection at 4y point for USD (2000-2014) and between 7y and 10y for EUR (2004-2014), and 3rd factor belly point at 10y for EUR and 4y for USD (same periods). If I do PCA on ED gold bundle (2004-2014) I get both inflection and belly at 8th contract - I'd guess it's similar for EURIBOR. Correlation between changes in factors for swaps vs STIRs is meh for USD (0.6 for level and slope, 0.33 for curvature - 2004-2014 period).

How would you fundamentally interpret the difference between factors for STIR and for the whole curve?

 

Could you define belly and inflection. Pretty please. Thx