Posts  / #POST-252475
REDDIT

Value of Long Bonds

B
Oct 4, 2026 · 02:43

I'm beginning to notice a lot of fear in the US bond market, and I suspect it might be a good opportunity. Be greedy when others are fearful. I see fear around bonds. Yields are spiking and people are increasingly worried about some kind of bond crisis. I want to share my thoughts. The first thing to consider is the expected rate of return of US stocks, which obviously requires making some assumptions.

Expected return = \[(1 + EPS growth) × (Ending P/E ÷ Starting P/E)\^(1 ÷ Years) − 1\] + Dividend yield

If the S&P 500 continues to grow EPS at its historical average of about 6.5% per year, maintains roughly its current 1.05% dividend yield, and the P/E stays around its current 26x TTM, then the implied return over the next 10 years is about 7.5% per year. That changes very quickly if valuations normalize. If EPS still grows 6.5% per year and the dividend stays the same, but the P/E falls from 26x to its historical median of about 16x, the expected return falls to about 2.5% per year over the decade. If the P/E falls to 13x, which we saw around 2011, the expected return is only about 0.4% per year. If inflation averages around 3.5%, which is roughly its long-term historical average, then both of those scenarios would result in negative real returns.

Also keep in mind that all of this assumes EPS continues growing at 6.5% per year for 10 straight years. Markets historically don't work that smoothly. We've had major earnings declines during periods like 1929-1932, 2000-2002, and 2007-2009. We also haven't really had a prolonged period of earnings weakness in quite a while. The 1970s are a little different because nominal EPS actually grew, but inflation was high and the P/E multiple collapsed. So even strong nominal earnings growth did not translate into strong real returns. If instead we assume 0% EPS growth over the next 10 years due to deteriorating economy, keep the same dividend yield, and the P/E returns from 26x to its historical median of 16x, the expected annualized return is about -3.7%.

Now compare all of this to the current 10-year Treasury yield of around 5.25%. If EPS continues growing at 6.5% with no recession in the moderate future and the P/E stays around 25-26x, then stocks will outperform bonds. But many more scenarios result in bonds outperforming. If the P/E returns closer to historical levels, or earnings growth is weaker than normal, then treasuries can easily outperform stocks over the next decade.

What I find interesting is that you don't even need yields to fall for the bond investment to work. If you buy a 10-year treasury around 5.25% and hold it to maturity, you already know approximately what your nominal return will be. If inflation falls, economic growth weakens, or the Fed eventually cuts rates, yields would likely fall and bonds would also appreciate in price. The risk is inflation rises further (currently 3.4%) and persists for the decade eating away at the 5.25% yield.

I personally prefer 5-10 yield bond maturity because they currently offer most of the yield available on 30-year bonds without taking nearly as much duration risk. You can lock in roughly 5%+ yields for a meaningful period, but you are less exposed if long-term rates rise further.