To assess whether stocks are cheap or expensive, we explored P/Es and decided that overall, the S&P is trading at about average levels. However, the playing field shifts with interest rates. Right now, interest rates are being artificially held down by the quantitative easing practices of the Federal Reserve. Part of Ben Bernanke’s plan (or plot, some might say) is to compel investors to buy “risk-on” assets, such as equities, as interest rates hover near zero and the dollar loses value.
Conclusion: the Fed’s Zero Interest Rate Policy (ZIRP) should support higher multiples going forward.
The P/E 10 is used to ‘smooth out’ shorter-term (year–to-year) fluctuations in total S&P 500 corporate profits. It does not use ‘adjusted’ earnings (which eliminate non-recurring items).
This is how the P/E 10 looks right now compared with where it has been since 2003. The historical middle quintile for P/E 10 ratios has been 14.3x – 17.3x.
At a glance, today’s 20.36x doesn’t look cheap, except when compared with the pre-recession period.
Computer screens don’t take into account when outlier data points need to be thrown out.
In 2009 an unprecedented spike in P/E resulted from the enormous write-offs that occurred in the financial sector. That year’s trailing P/E multiple did not represent investor enthusiasm. It registered high because earnings fell dramatically (massive, one time non-recurring losses greatly reduced “earnings”). As such, the 2009 number was totally not meaningful (NMF).
Published P/E 10 figures include that 123.7 P/E in their calculations. While technically correct, this turns all smoothed data into apples-to-oranges comparisons.
Clearly P/E 10s of 25x – 27.5x were flashing warning signals from 2004 through 2007. That was a message that should have triggered caution.
Here is the chart of Trailing 12-Month P/Es – this shows single-year P/Es without the 2009 result tainting subsequent years.
Trailing 12-month P/Es in that pre-recession period got into nosebleed territory at 34x – 46x. By comparison, the current trailing P/E is not expensive.
The key take-aways from the charts:
- Today’s S&P 500 P/E falls squarely in the mid-range of historical levels
- Market-wide Single-digit P/Es are extremely rare events
- P/Es above 22x typically highlight excessive valuation
- Year 2000 and 2007 bubbles were historically significant SELL signals
- The 2009 period P/E was misleading and not-meaningful
A key piece of information is missing.
Price/Earnings ratios are not set in a vacuum. Equities compete with bonds, CDs, real estate and all other asset classes when investors decide where to place their money.
To know where P/Es ‘should be’ you need to compare them to interest rates. Higher available coupon rates dictate lower expected market P/Es. This makes sense. In the early 1980s, you could buy fully-insured bank CDs paying 10% -15%. 30-year treasury bonds locked in non-callable 13.5% coupons.
When fixed income pays well…why take stock market risk? That’s why the P/E 10 hit a post-1950 low of 6.8x in 1982.
Unless you believe we live in deflationary times, today’s all-time low, ZIRP-induced, interest rates offer no competition for stocks. Bonds offer guaranteed losses on an after-tax, inflation-adjusted basis.
That makes a ‘normal’ P/E 10 look like an incredible bargain.
Dr. Paul Price, Dec. 17, 2012