UST Yield Curve Forecasting - Bond Structure Testing

Question

I have a project in mind that I am working on, but have little idea where to start. I am a relative newcomer to python (about 1 years exp.) and limited knowledge of quant finance.

What I would like to do is simulate possible yield curves going out 1-3 years. To do this, I planned on scraping the US Treasury website for historical data, and deriving historical daily changes in rates and correlations between terms. Then use Monte Carlo simulation to forecast future curves. - I have built a version of this in excel and crystal ball, but would like to code it in python....

On top of this, I would also like to compute how different bond structures perform during those simulations... bullets, vs short term callable bonds, etc.

I am thinking i would need to use PyMC for the simulation... Any ideas on how to effectively price the various bond structures?

Is this whole thing a worthless idea?

Answer 1

You might be able to forecast interest rates using the yield curve itself. I am writing this on the fly so idk where interest rates are at right now but say if the one year US treasury is at 1% then the expected rate for the 2-yr should be 2% ( since you can gain 1% for one year and at maturity purchase another one year treasury and gain another 1%)... But as you may have notice this is not the case. The one year treasury can be at 1% and the 2-year may be at 2.50% , thus implying that the markets expect an increase in interest rates in the short term since the rate is higher than the "expected" rate of 2%... I hope it helps