Specifically, I am not only looking for the spread between ten years and short term notes (although that would be a great start) but I would like to use the breakdown of the various yields at various maturities to predict bond prices.
The yield curve is a useful tool for macroeconomic forecasts (for example an inverted yield curve is supposed to mean trouble for banks and high debt companies and possibly the economy) and it may also be useful to help predict bond ETF prices.
Theoretically, the data is in the system already because we can derive the yield curve by comparing the yield on the various treasury ETFs. Is there a way to get at this data? How?
Okay, here is what I’m after. I am trying to create a system that picks government bond ETFs based on the expected total return over the next month or so.
There are three variables that affect the total return of a Treasury bond: the interest rate at the time of purchase, the contango or roll yield (that is, the difference between the current interest rate and the interest rate at maturity) and the change in interest rates over the next month. Of these three the roll yield is the least appreciated by investors and the change in interest rates is the most overrated. (Rising rates means a lot when it happens but it doesn’t happen too often and it is partially hedged by the stock market holdings–the Fed does not like to raise rates unless the stock market is rising).
I like to use treasuries as a hedge. The beauty of switching from one treasury ETF to another is that it will automatically move to short term treasuries when the fed raises the short term rates and inverts the yield curve. Knowing that the US government would like to raise short term rates sometime over the next decade or so I would prefer that my system adjusts automatically to it. (My guess is that it is very hard for the Fed to raise rates while other countries are lowering rates because of a too strong US dollar)
The question is how to implement such a system. I am puzzling over the math to calculate the expected interest rate at maturity given the current yield curve and the market value at resale. Is anyone else interested in this?
Hi Chipper. I was looking at bond systems since fall 2014, yours and Tom’s for the same purpose like you mentioned: hedge to stock portfolios. Since Jan-Feb 2015 TLT’s crazy swings I started to dig the same way you described above. I’d prefer bond rotation based on the foundation like yields curve slope.
[quote]
Since Jan-Feb 2015 TLT’s crazy swings…
[/quote]Hi Konstantin,
For what it’s worth in 2015 so far the maximum drawdowns are as follows:
TLT -10%
SPY -5%
60% SPY / 40% TLT only -3.6%. So treasuries are still a hedge for stocks.
60% SPY / 40% The Bond Hedge -2.4%.
Alert: Price drop on The Bond Hedge R2G.
My original plan was to charge a little more and limit subscribers so that there would be enough liquidity for everyone. I have since come to the conclusion that when more people trade ETFs (up to a point of course) it can actually benefit everyone as it can cause market makers to step in to narrow the spread. So I am lowering the price and expect to take more subs.
[quote]
I’d prefer bond rotation based on the foundation like yields curve slope.
[/quote]I am working on this. It’s a bit complicated to program and still a long way from completion.
Thanks for the links. The best systems (by far) adjust allocations to stocks and bonds based on market conditions. I cannot wait for P123 to allow books that do this.