Suddenly, everybody is focusing on inflation, not only because the Fed is fixated on it, but also because inflation’s impact on equity valuations is getting more widely recognized. Today’s WSJ carries this from James Mackintosh:
Inflation’s Low. That’s Not Enough To Justify Expensive Stock Markets Hollywood isn’t likely to make “Federal Reserve: The Movie” any time soon. The problem for the scriptwriters would be the absence of a villain for Ms. Yellen to fight, now that inflation seems to have all but disappeared.
(…) [a] question looms large for investors: even if inflation has been conquered, does that justify high stock prices?
The question might seem dumb. Low inflation means low interest rates and so higher prices for equities. As projected future profits are worth more in today’s money, companies spend less to service their debts and shares become more attractive relative to low-yielding cash and bonds. What’s not to like?
History confirms the simple view, in that stock valuations have on average been much higher when inflation is below 4% than when it is above. Even better, the S&P 500 on average rose about 8% in the year following inflation coming in below 4%, against just 2% gains for faster inflation.
Unfortunately, averages conceal a lot, and in this case they hide the truth. The truth is that what matters most to stock prices isn’t where inflation stands, but where it will stand in future compared with what is currently priced in. Investors like low and stable inflation, but some of the best times to buy stocks have been when inflation is very high, and about to plummet, as in 1979. Equally, some of the worst times to buy stocks have been when inflation seems under control, but is about to take off, as for example at the end of 1936.
Even worse, the average hides massive variation. Data put together by Yale Prof Robert Shiller for the S&P and U.S. inflation back to 1871 show that investing when inflation was between 1% and 2%, as it currently is, offered a 1-year gain in the S&P averaging 8.6%—with dividends on top. Not bad, you might think as you dial your broker. But the range was huge, from a whopping rise of 41% to a loss of 35%—again depending on whether inflation subsequently rose or fell.
This might seem like ancient history, but in August 2007, considered by many to be the start of the financial crisis, inflation was below 2% and the stock market was booming.
Perhaps most relevant is late 1965. Inflation had been below 2% for seven years, stocks were on a roll and Beatlemania was at its height in America. Investors seemed to agree with John Lennon as he sang “I Feel Fine,” and stock valuations hit their highest since 1929 on the widely used Shiller P/E ratio, which smooths the cycle by comparing price to 10 years of earnings. It would be another 30 years before U.S. shares were again so highly valued. (…)
Mackintosh nails it when he writes “what matters most to stock prices isn’t where inflation stands, but where it will stand in future compared with what is currently priced in.”
This is the Rule of 20, without any forecasting: fair P/E is 20 minus inflation. As inflation fluctuates, so does “fair P/E”. Superimpose fluctuating investors psychology and you get varying discounts/premiums over “fair P/E” and you get
- some of the best times to buy stocks when inflation is very high and falling or
- some of the worst times to buy stocks when inflation seems under control and takes off.
In August 2007, inflation was indeed 2.0%, down from 4.7% in September 2005. The S&P 500 Index closed August at 1474, up 20% in 2 years. The trailing P/E was 16.4 and the Rule of 20 P/E was 18.3, both still in ok range. But CPI spiked in subsequent months reaching +3.5% YoY in October. The trailing P/E had reached 17.8 but the Rule of 20 P/E had spiked to 21.3 when the S&P 500 peaked on October 31 at 1575. Inflation peaked in January 2008 but earnings were then collapsing from their June 2007 peak.
Fluctuations in the Rule of 20 P/E (black line below) are pretty regular between 15 and 23, always providing an objective assessment of the valuation risk/reward ratio. (The yellow line on the chart plots the Rule of 20 fair value combining the Rule of 20 P/E with trailing earnings since earnings trends occasionally offset inflation trends.)
Mackintosh is right to mention the mention the mid-1960’s as the current period could turn out similarly with low and stable inflation requiring a hands off Fed while profits advanced regularly (look at the yellow line between 1961 and 1966). The S&P 500 peaked in January 1966 when the Rule of 20 P/E touched 20 (actual P/E 17.7), right when inflation troughed at 1.9% and profits stalled. The S&P 500 lost 17% in 9 months, bottoming out at the inflation high point of 3.8% in October 1966. It subsequently rose 26% in the following 12 months. The 100 basis points drop in inflation more than offset the 4% decline in profits.
The problem is that inflation reading is getting pretty difficult.
But the advance is debatable as shown yesterday. Now the Fed and the inflation hawks could continue to claim that recent lowflation is due to transitory factors, From The Daily Shot:
The core CPI would rise above 2% if one removes what the Fed considers to be “transitory” factors such as mobile service, medical care, vehicles, etc. This is a dicey approach because what’s viewed as a temporary effect can shift from sector to sector. Regularly removing these factors could introduce a measurement bias. Nonetheless, here it is.
But BloombergBriefs has another viewpoint:
If the inflation data reaccelerate over the next few months, a significant component of the pickup may merely reflect supply-chain disruptions resulting from hurricanes Harvey and Irma, thereby masking the persistent weakness. The risk is that policy makers will gain an unjustified sense of confidence in an inflation pickup and potentially proceed with rate normalization at a pace faster than they would otherwise deem appropriate.
For this reason, BI advocates using some type of filtering when evaluating the next several inflation reports. At the very least, analysts should strip out the components obviously distorted by the hurricanes, such as gasoline and motel/hotels; but more appropriately, they should pay closer attention to “trimmed-mean” metrics of inflation, which more robustly filter out distortions.
This pattern may provide some comfort to analysts that the deceleration is leveling out, but substantial contributions from categories such as non-school lodging away from home, which rose more than 5% in the month, suggested that hurricane impacts may have influenced the core as well. The unrounded core was 0.2485%, of which BI estimates that about five basis points was related to hotel/motel costs.
Here’s the Cleveland Fed’s wrap up on inflation trends:
The debate will surely intensify as recent transitory deflators meet with upcoming transitory reflators.
Meanwhile, in the background:
Central Banks Edge Away From Easy Money The Fed, ECB and BOE all appear to be on a similar path toward less accommodative monetary policy
And in the black ground: