You'll often hear people talk about P/E ratios for an index being far above or below their mean, with the implied suggestion that it is misvalued. I was thinking about this concept over the weekend and found the argument unsatisfying: in particular, I felt that a rational investor cannot look at P/E ratios in isolation but must consider other potential sources of return when allocating his capital.
The most natural asset to compare P/E with are sovereign bonds, as a liquid, often held investment and as a useful 'risk-free' measure (for a given value of risk-free, of course).
My hypothesis is that a rational market should eventually bring P/E 'yield' in line with sovereign yield, with a risk-adjusted premium. Simply put, if AAPL is going to return your capital in 10 years whereas treasuries are going to take 50 years, in the absence of a market or AAPL meltdown I would expect money managers to plump for AAPL.
Below is a chart of 10Y treasury yield vs the historical P/E of the S&P 500 Index from Jan 1970 to Dec 2007, along with a non-linear line of best fit. As can be seen there is indeed a significant relationship between yield and P/E, as my hypothesis predicts.
The most natural asset to compare P/E with are sovereign bonds, as a liquid, often held investment and as a useful 'risk-free' measure (for a given value of risk-free, of course).
My hypothesis is that a rational market should eventually bring P/E 'yield' in line with sovereign yield, with a risk-adjusted premium. Simply put, if AAPL is going to return your capital in 10 years whereas treasuries are going to take 50 years, in the absence of a market or AAPL meltdown I would expect money managers to plump for AAPL.
Below is a chart of 10Y treasury yield vs the historical P/E of the S&P 500 Index from Jan 1970 to Dec 2007, along with a non-linear line of best fit. As can be seen there is indeed a significant relationship between yield and P/E, as my hypothesis predicts.
The relationship however gets particularly interesting when we add in the data post 2008.
The data points in the red oval are from late 2008 and 2009 when the SPX plummeted. An interpretation of the fact they stayed so high for so long was that - despite the speed of the crash - it still wasn't fast enough to match the wipeout in earnings during that period.
Since the crash, the SPX has recovered nicely and is currently near its pre-2008 highs. The period from 2010 to present is highlighted in the green oval above. As we can see, despite the fact that the P/E for the index is above its long term mean, it is still undervalued relative to its historical relationship with treasury yields.
There are of course a variety of explanations for why this may be the case. Lingering fears over the situation in Europe, the US fiscal deficit, stubbornly high unemployment and a permanent reduction in growth potential of the Western world (in the face of competition from China and other EM) are the most obvious. However, all of the above have occurred in the last 40 years at some point or other, and the relationship has held. It is my view that SPX will return closer to its predicted P/E ratio.
That ratio, utilising the equation from the graph with all current data, is about 25 - compared to a current reading of 18. In other words, SPX has a potential upside of roughly 40% from here. Even if that is too ambitious a reading, I prefer being long SPX than short at these valuations.
Note: The equation of best fit for both differ in form, but choosing an exponential form for the first graph would result in nearly the same R^2.
No comments:
Post a Comment