In thinking about asset allocation between stocks and bonds, it seems that a 50/50 split between stocks and intermediate term treasuries is pretty solid and safe, if not a bit conservative. It is has been shown that a 2x levered 50/50 split has better risk adjusted returns than 100% equity. In thinking about the cheapest most efficient way to create this I have looked at 5 and 10 year treasury futures for the bond component.
You could invest 90% to 95% of your funds in P123 equity models and use the remaining 5% to 10% to buy the treasury futures on margin, in the equivalent notional value as your equity. Treasury futures have a very low margin requirement, especially for intermediate durations. It seems like the implied cost to borrow (i.e. cost of leverage on treasury futures) is pretty close to LIBOR.
So essentially you could create a 50/50 equity-treasury portfolio, leveraged 2x less the annual LIBOR %. This would be much cheaper, than even IB margin rates. What if anything am I missing?
I do not know much about this so I will keep it short and as a true question. Bond futures, bond funds and bonds are all different animals aren’t they? Especially, bond futures seem significantly different to me in that there is NO COUPON PAYMENT. With no coupon payment doesn’t it become a zero-sum game with participants betting on the future price? The coupon payment would make it a non-zero-sum game where everyone could split the coupon payment. So my question would be: isn’t it like betting High (rising bond prices) or Low (declining bond prices) on a fair roulette table with no 0 or 00? This may be a poorly informed question.
Less of a concern to me would be that they are derivatives and carry default risk. I am thinking of XIV here—I never would have believed that was possible. The US Treasury will never default no matter how bad it gets, I think. They will, instead, just print money (probably). But the Treasury bond will last longer than any of its derivatives I would think. It is hard to image how it could be the other way—people making good on the derivative payment for a bond that has defaulted.
Anyway, sounds good to the extent that bond futures mirror the behavior of bonds. And it still could be a good play–even if futures do not mirror bonds–but that determination would take more knowledge than I possess. It needs to be considered for what it is—having bond in the name does not make it a bond.
The coupon payment is accounted for in the futures price. There should be no difference between buying a fully cash secured bond future and buying the actual bond on on a rolling basis.
I hedged my port with TLT for year or so, beeing long 130% Stocks and then adding 50% of TLT. It backtested very well and I had success with it, but I unwound the TLT Trade:
I researched long term Bond / Stocks correlations and the correlation change, e.g. it used to be the case that Bonds went up, when stocks went down, but that regime changes quite a bit.
Though I have to say it was the perfect trade the last 15 Years or so, but since I can not backtest longer I rather do not play around with somthing I do not understand.
Andreas