“History does not repeat itself, but it does rhyme” (Mark Twain)
I am not inclined to put much weight on the behaviour of historical equity prices to predict future trends but I must admit that a 1.000 batting average over half a century deserves attention. From Charles Schwab’s Liz Ann Sonders:
(…) there is a strong (well, perfect since 1962) historical tendency for the stock market to give back a decent amount in a typically-first half corrective phase. The good news caveat is that there has been an equally strong/perfect historical tendency for subsequent major rallies.
Let’s verify what the Rule of 20 was saying in each of those presidential mid-term years. As a reminder, the Rule of 20 says that fair P/E is 20 minus inflation. When the Rule of 20 P/E (P/E on trailing EPS + inflation) is at or above 20, the odds are that “something” will eventually happen that will correct this inherently unstable overvaluation as the chart below illustrates (see also Understanding The Rule Of 20 Equity Valuation Barometer). .
The chart below plots the Rule of 20 P/E reading for every mid-term years since 1962 before and after each correction, 20 being the border line between cheap (undervalued)and expensive (overvalued) equities.
Taking all 13 mid-term years, the Rule of 20 P/E averaged 21.0 before the correction and 17.9 after. In 10 of the 13 years, the Rule of 20 P/E was above 19.2 before the correction (22.2 average). Evidently, the Rule of 20 P/E is very sensitive to changes in inflation. At the 20 level, a 0.5% variation in inflation will change the fair value of the S&P 500 Index by 2.5%. Inflation rose in 7 of the 13 mid-term corrections and declined in 2. In 1982, the positive effect of the 2.5% decline in inflation was more than offset by a 9.4% drop in earnings (recession). Significantly, earnings rose in 7 of the 13 corrections (+10.5% on average), indicating that earnings gains do not necessarily prevent meaningful corrections although they can help offset the effect of rising inflation.
Several important observations from the above:
- Equity markets have a “natural” tendency to self-correct overvaluation defined as 20 or higher on the Rule of 20 scale. Ten of the 13 mid-term corrections occurred after the Rule of 20 P/E exceeded 19.2.
- Three corrections occurred when the Rule of 20 P/E was well below 20: in 1978 (15.5x, –14%), in recessionary 1982 (17.8x, –14%), and in 1986 (17.1x, –9%). In both 1978 and 1986, the correction was very short and milder than the 21% drop experienced in the 10 corrections from high valuation levels.
- Specific economic or financial catalysts are not necessary ingredients for markets to correct.
- Rising earnings do not prevent markets from correcting.
- Rising inflation has been present in 7 of the past 13 mid-term corrections. Since 4 corrections lasted less than 3 months, we can argue that 7 of the 9 lengthier corrections witnessed rising inflation rates. Correlation is not necessarily causation but rising inflation rates are seldom positive for equity valuation. When inflation fell, it was during a recession (1982) and during the short-lived post-crisis panic of 2010.
In all, the risk/reward approach to equity investing using the Rule of 20 is very much validated by the mid-term bust-boom pattern.
It is rather futile and dangerous to seek and wait for specific catalysts when the risk/reward equation becomes unfavourable. The Rule of 20 P/E fluctuating around the “20” fair value level, it is easy to calculate potential return vs potential risk and adapt one’s investment strategy to one’s own risk tolerance level.
This is a mid-term election year and equities have yet to perform their usual correction, unless the 6% late January decline counts as a mid-term bust. In any event, the Rule of 20 P/E is currently 19.1x. Inflation has yet to show any definite upward trend even though Fed officials vow to bring it up towards their 2.0% target, and even beyond as per the official FOMC March 2014 statement:
(…) the Fed’s official policy statement included a new line noting that officials expect to keep rates lower than normal even after inflation and employment return to their longer-run trends.
Let’s now look at the “boom” part since
The good news caveat is that there has been an equally strong/perfect historical tendency for subsequent major rallies.
Knowing this, you may just as well decide not to bother with the “bust” risk. However, you might want to consider the following facts that Ms. Sonders omitted:
- Even though the average boom is +32%, following an average bust of –19% it produces a net gain of only 7% over the entire bust-boom period. Many may decide to avoid the aggravation and just wait for the hurricane to pass.
- Four of the periods (30%) ended with a net loss, ranging from –1.6% in 1962 to –15.7% in 2002. In the 1990 and 1994 corrections, the net gain was only 3.2% and 4.6% respectively.
- In four other periods, 1966, 1970, 1982 and 2006, the boom was fuelled by sharp declines inflation, something central bankers around the world are currently fighting against.
- That leaves three years to examine:
- in 1986, the 40% gain during the 12-month boom period to September 1987 was entirely lost during the next 2 months when the crash deflated the well overvalued Rule of 20 P/E from 23.1 to 17.4;
- in 1998, the 38% gain during the 12-month (internet) boom period to August 1999 was pure irrational exuberance as the Rule of 20 P/E rose from 23.2 to 29.5;
- In 2010, the 31% gain during the 12-month boom period to June 2011 was supported by the strong 20.8% jump in trailing earnings which more than offset a 2.4% increase in inflation.
To conclude, the mid-term bust risk is significant, dangerous and unforeseeable. Current market valuations are certainly high enough to make investors very nervous and trigger-happy. The above analysis demonstrates that betting on the bust carries much better odds than betting on the subsequent boom.