Kbg wrote: ↑Fri Aug 07, 2020 8:17 am
The great thing about bonds is that they are incredibly predictable and by this I mean throw a scenario at them and you know what is going to happen because the math is all there to tell you what will happen.
Without a doubt the best prediction for a bond is it's interest rate. It's not even a prediction, it's fact (unless it defaults). Now of course a PP rolls out of a LTT bond after X number of years so all that bond math begins to matter a lot.
LTTs are going to stay volatile and Tyler's convexity graphs are phenomenally good in allowing one to understand how all that works in a very intuitive way. Dialing down duration, dials down volatility so if you are looking for the punch of LTTs you aren't going to get it unless you dial up your allocation.
Bonds and cash all pretty much suck right now. Nothing we can do about that, it is what it is. We also know pretty much exactly what we are going to get from those assets, right now, in terms of return. There's really no guess work about this at all.
So for me the question of LTTs is do I still want their volatility in the portfolio or not? Over the short term, LTTs are an excellent diversifyer of stocks due to negative correlation. If we have a sustained rise in interest rates then they become positively correlated and we hope our gold pulls us through it.
The PP needs the volatility, but my question is, how much volatility? So what I'm getting at is that at interest rates bear 0, LTTs become extremely volatile, very much more than at 5% or 10%. What were rates when HB designed the PP? It matters, because the PP needs volatile assets to get the rebalancing benefit. But does the PP machine need the
extreme volatility of LTTs at near 0 rates, and is it worth the extreme risk?
For example, looking at Tyler's convexity chart, it would appear that you could get about the same volatility using 1) 20-year treasuries at around 8% or, 2) 10-year treasuries at near-0 rates (don't check my math as this is an approximation). So the volatility is about the same, but the difference is that in the with the 10-year bonds you are reducing the losses if/when interest rates rise.
I noticed that some have suggested using 50% ITT instead of 25/25 Cash+LTT. I wouldn't do that or even propose it as I think it breaks the workings of the PP machine.
For different Interest rate environments:
High: Cash, TLT, IAU, VTI
Mid: Cash, TLH, IAU, VTI
Low: Cash, IEF, IAU, VTI
Of course you can backtest all of these, but the results would be totally meaningless considering the decreasing interest rates of the last 40 years.