Edwards expected CAPE to be around 10 by now, given the moves in bond yields, and admits he was guilty of a “forecasting error of epic proportions.” But his Ice Age thesis has played out as predicted in Europe, and he has also been correct to predict that stocks would look ever cheaper relative to bonds in the U.S. For now, his judgment is clear: “In my Ice Age view of the world, Robert Shiller is dead wrong. In my view, US equity valuations are a QE-fueled bubble waiting to burst.”
Now the question is whether this is really so different from the Shiller view. His model plainly suggests that stocks will do badly over the next 10 years, and that bonds will do even worse. This was the way Shiller put it in a research piece for Barclays Plc in October, (which can be found on SSRN here):
In summary, investors expect a certain return in equities as compensation for investing in a riskier asset class, and as interest rates have declined, the relative expected return for equities has increased dramatically. We believe this may quantitatively help to explain investors current preference for equities over bonds, and as such the quick recoveries we are observing (with the exception of the UK), whilst still in the midst of a pandemic. In the US in particular, we are once again observing stretched valuations and high CAPE ratios compared to history.
Bond arithmetic may help to show that Edwards and Shiller aren’t as far apart as they appear.
When yields are this low, moving to a higher yield involves serious losses. To get from the current 10-year yield of 1% back to the 3% that 10-year Treasuries were offering as recently as two years ago, the Treasury price would have to drop by two-thirds. (If yields were a more normal 4%, then a two-percentage-point increase would require a fall in the bond price of only one-third.) At this point, bonds offer low income, little upside, and risk of massive downside.
Maybe it isn’t that big an act of apostasy for someone who remains dubious about the future for stocks to predict that they should still do better than bonds.
Read entire article