The Divergence In Opinions
I often find the comments section of Seeking Alpha articles as informative as or even more informative than the article itself. It is the reason I believe it is my duty as an Author to respond to all comments in a meaningful way. I find it useful to sometimes use an article to more comprehensively respond to comments. I had intended to move on from the matter of P/Es to the matter of yields in this Part 2, but the following extracts from the comments section of Part 1 need to be addressed. In this instance, an article is required, with tables included, to do justice to the subject matter. Concluding part of a comment from Chuck Carnevale:
So to conclude, I do not believe that a 15 P/E ratio is over simplistic as a valuation reference. However, I would agree that it is not an indicator of the future return you can expect from investing in a given stock. That will be a function of earnings growth (or lack thereof) plus dividends if any. In other words, I believe your article is arguing with yourself but not me. In truth, we agree more than we disagree. Nevertheless, I do appreciate your reference to my article, my only goal here is to make sure that the record is clear.
Clearly, I have not convinced Chuck that P/Es are not useful as a valuation reference (the underlining in the quote is mine). Exchange of comments between David Van Knapp and this Author:
David Van Knapp - Changes in EPS are the company's contribution to price changes. Changes in P/E are the market's contribution. Both are important, and looking only at one always gives an incomplete picture.
Author - I essentially agree with what you say, but I take a different view on PE. I regard PE merely as a "result" of a calculation and not a causative factor. The two causative factors are EPS, and not PE, but share price. The EPS and changes in EPS are influenced by the company management, e.g., increasing sales and thus profits, and outside influences such as a change in tax rates. As you say, the share price is influenced by EPS, but also by many other factors. I intend to cover these matters and the very important matter of dividends in subsequent parts. There are some semantics in play here whether "other influences" affect the share price, the price people are prepared to pay in different stages of the economy, or if these "other influences" affect the PE with a resultant effect on share prices. I might yet agree with you that it is PE that is affected, but at the moment I lean more towards blaming share prices, they being a more tangible and direct measure.
David Van Knapp - Hi Robert, Good discussion. I look forward to your follow up article. As you deciphered, I see PE as a "cause" of price, because I see PE as reflecting investor sentiment about a stock or the market as a whole. Over time, as I am sure you know, the PE for a particular company can change drastically. I often use JNJ to demonstrate how a company's price may not at all reflect how it is growing earnings. JNJ's dead-money era was like that. Earnings grew very well, but PE sank simultaneously, resulting in flat price for a decade.
What I take from the comments by these DGI thought leaders
Part 1 of this series essentially looked at whether historical average P/Es and/or discrete past periods of high and low P/E levels were any guide to the appropriateness of today's P/E levels. My findings were firstly, that they were not useful; and secondly, P/Es at any period in time are not a suitable guide to the returns that might be achieved. There appears to be no disagreement from either of the DGI authors on my second point. In fact one has only to read their various articles to know their decisions on whether or not to invest in a given company has little to do with P/Es and very much to do with in-depth fundamental analysis. There also appears to be no stated disagreement to the findings that past P/E levels are not a useful guide in the current day situation. But, the illuminating comment for me was from David Van Knapp, "As you deciphered, I see PE as a "cause" of price, because I see PE as reflecting investor sentiment about a stock or the market as a whole." I know I will have difficulty winning this argument whether price drives P/E, or vice versa, because David is not alone in this view. It is apparent from the many articles written on the subject that this view is widely held among investors, and investors determine price. But I believe that is a mistaken view and below I provide an evidence-based argument to prove my point.
Now this question of "investor sentiment" is one worth exploring because it is a significant factor in the market's pricing of shares on a day-to-day basis. In theory, "Mr. market" takes various inputs such as the bond rate, a factor for the higher risk for stocks, a stock's beta, reported earnings and dividends, and guidance on future earnings, and comes up with an appropriate price to pay for a share in any particular company. On that basis, share prices should be stable in the absence of a change in any of these inputs. But share prices can fluctuate widely, over the course of days and weeks and months, in the absence of any significant change in those inputs. These fluctuations are, I presume, what David Van Knapp is referring to as changes in "investor sentiment". I can understand why David might prefer a P/E, rather than share price, for monitoring changes in "investor sentiment". Take a company with a share price of $100 and EPS of $5 and a company with a share price of $50 and an EPS of $2.50. Comparing the $100 share price to the $50 share price, or say an increase in share price of $10 and $5 respectively, is fairly meaningless. But in those examples, each has a P/E of 20 prior to the change and a P/E of 22 after the change, allowing a more direct comparison of changes. This use of P/Es may facilitate comparison even if the companies have different P/Es at the outset.
Evidence-based argument that P/Es do not drive share price
I have collated further data for the dividend aristocrats included in Part 1 to support my contention that P/Es do not drive share prices. P/Es are merely a result of dividing current share price by current reported EPS, rather than share price being determined by multiplying EPS by a P/E somehow determined by the market. The EPS I use in TABLE 1 below and throughout this article is the trailing 12-month actual EPS reported quarterly on NASDAQ and by the respective companies in their 10-K and 10-Q SEC filings. The share prices I use are the closing prices as reported on NASDAQ. The effective date for the TTM EPS to be updated in PE calculations is the date of each respective company's quarterly conference call.
TABLE 1 above shows share prices, EPS and calculated P/E at August 3, 2015, and 52 weeks later at 29 July, 2016, for each of Johnson & Johnson (NYSE:JNJ), Coca-Cola (NYSE:KO), The Procter & Gamble Co. (NYSE:PG), and McCormick (NYSE:MKC). Also shown are the percentage changes in share price, EPS and P/E for the 52-week period for each of the companies. There is considerable variability among the companies as regards to share price growth and changes in EPS and P/E. It is possible to analyze these changes in share price into: changes occurring when changes in EPS and forward guidance are given at quarter end; and changes that take place at other times. TABLE 2 below provides an analysis of the changes in share price and P/E occurring when changes in EPS and forward guidance are given at quarter end.
I refer again to my quote above of David Van Knapp's comment, "Changes in EPS are the company's contribution to price changes. Changes in P/E are the market's contribution." In nine of the 12 instances of EPS changes highlighted above, it can be seen that disclosure by the four companies of EPS changes and forward guidance, at quarterly conference calls, did not result in any significant share price change. The share prices stayed much the same. Rather than change the share price, the EPS changes caused a change in the P/E ratio. I believe this supports my argument that published P/Es are merely the arithmetical result of dividing share price by reported EPS. Even in the three instances where there were significant changes in share price after EPS announcements, there were also significant changes to P/E. Clearly, if EPS causes changes in share price, and I am sure it does, it is not the EPS reported at quarterly conference calls. I believe the EPS that does drive share prices is the aggregate of investor sentiment about the future profitability of the company. This EPS is quite dynamic and is impacted by external events such as oil price shocks and Brexit. In the case of Brexit, there was a great deal of concern about the impact of foreign currency exchange effects on profits for US companies with international operations. There was also a perception of heightened general market risk associated with uncertainty. My point is, we cannot easily separate nor readily identify the EPS the market is factoring into its thinking from movements in share price due to general market risk sentiment, and specific risks, such as potential interest rate increases. Therefore, P/Es based on quarterly reported EPS data are fairly useless as a guide to investment. Not only does reported EPS quickly become out of date, but also as illustrated in TABLE 2 above, changes in reported EPS are not dynamically driving share prices. That raises the question of when, and by what cause, did the bulk of the changes in share price, per TABLE 1 above, occur.
Further Analysis of the Changes in Share Price per TABLE 1
The period covered in TABLE 1 was from August 3, 2015, to 29 July 2016. To facilitate meaningful analysis, this is now broken down into three discrete periods, the first being pre-oil shock, the second being the oil shock period and aftermath, and the third being the post-oil shock period. What I am terming the pre-oil shock period runs from August 3, 2015, to November 9, 2015, as per TABLE 3.1 below:
Pre-Oil Shock Period (August 3, 2015, to November 9, 2015)
I will comment on TABLE 3.1 by company, but bring to attention that all four companies had falls in share price of up to 7% to 10% during this period, but all finished the period within 1.8% of the opening price. On the other hand, performance as measured by P/E was all over the place with ending P/Es for JNJ, KO and MKC more than 8% above beginning P/Es while PG's P/E was 9.5% lower.
Johnson & Johnson
JNJ's P/E increased from 17.6 at beginning to 19.3 at the end of the period, making it appear much more expensive. However, a purchase at $100.02 on August 3, 2015, would have given a dividend yield of 3.0%, and a purchase at $100.24 on November 9, 2015, would have given a dividend yield of ~2.98%. It can be seen there was ample opportunity throughout the period to buy below $100 per share, and even down to $91.00 which would have provided a dividend yield of ~3.3%.
As for JNJ, P/E increased; in KO's case, from 24.3 at the beginning to 26.5 at end. But share price and dividend yield remained unchanged at $46.54 and 3.18% at the beginning and end. A closing low during the period of $37.99 gave the opportunity to buy at a ~3.5% yield.
The Procter & Gamble Co.
During the period, share price fell by 1.3% from $76.40 to $75.40 while P/E fell by 9.5% from 31.3 to 28.3 due to EPS increasing from $2.44 to $2.66. Here again, we see the P/E apparently changing as a result of a change in EPS rather than P/E and increased EPS acting to increase the share price. Dividend yield at the beginning and end was 3.47% and 3.52%, respectively. A share purchase at the closing low during the period of $68.42 would have provided a dividend yield of 3.88%.
MKC was virtually the mirror opposite to PG, with EPS falling by 5.8% from $3.27 to $3.08 and share price increasing by 1.8% from $82.04 to $83.55, resulting in P/E increasing by 8.1% from 25.1 to 27.1. Once again, we see EPS mainly changing P/E, rather than lower EPS multiplied by P/E, resulting in a lower share price. Dividend yields based on beginning and ending share prices were 1.95% and 1.92%, respectively. Dividend yield obtainable at the low share price for the period of $76.02 was 2.1%.
Oil Shock Period and Aftermath (November 9, 2015, to April 21, 2016)
We have seen from TABLE 3.1 that share prices for all four companies at November 9, 2015, had changed very little from August 3, 2015. But share prices had fluctuated in between, providing opportunities to invest at lower prices (high-single-digit percentages lower). TABLE 3.2 above covers the period of the "oil shock", which was reported on at the time per this headline and an extract from the February 11, 2016 edition of MarketWatch:
Dow closes at 2-year low, dogged by global market turmoil
The Dow Industrials and S&P 500 rang up their fifth losing day in a row Thursday, falling amid a global rout led by tumbling oil prices and losses in financial stocks.
It can be seen that share prices for the four companies did fall back on the 5th February, but by February 18, 2016, they had completely recovered from the effects of this global market turmoil. During the period November 9, 2015, to April 21, 2016, except for MKC, there was almost no opportunity to buy shares at a price significantly lower than the share price at the beginning of the period. This was so, even during the "oil shock" period. I believe the aftermath of the early recovery from the so-called "oil shock" global financial turmoil is that investors now have less fear of the likelihood of an event that will be both damaging and long term in its effect. Maybe financial commentators have "cried wolf" once too often. It is also worth noting that three of the four companies had dividend increases in this period. Based on annualized dividend rates, KO's dividend yield increased from 3.18% to 3.21%, the dividend increase more than offsetting the share price increase in the period. For PG and MKC, the dividend increase was not sufficient to completely offset share price increase. PG's dividend yield fell from 3.52% to 3.31%, and KO's dividend yield fell from 1.92% to 1.88%.
Post Oil Shock Period (April 21, 2016, to July 29, 2016)
TABLE 3.3 shows in the three month period to July 29, 2016, JNJ's, PG's, and MKC's share prices showed strong gains while KO's share price was virtually unchanged from the beginning of the period. Once again, with share prices on the increase, there was almost no opportunity during the period to purchase any of the shares at a price below the beginning levels. And again, there is clear evidence that changes in TTM EPS are not being reflected as corresponding changes in share price. If share price was a function of P/E multiplied by EPS, then changes in EPS should result in corresponding changes in share price. I can only conclude the market sets share prices by means other than multiplying reported EPS by a P/E the market considers appropriate. The market determines the market price, and P/E is merely the result of dividing market price by reported EPS.
Summary And Conclusions
Investors are showing less fear of a surprise knockout punch
Based on TABLES 3.1 to 3.3 above, share prices over the last 12 months, at least for the four dividend aristocrats included, have progressed from higher to lower volatility, and from overall minimal growth to a higher growth rate which is showing signs of flattening out towards the end. I believe yield-hungry DGI investors have survived so many financially crises, extravagantly reported in the media, they are becoming less fearful. In February this year, dividend aristocrat share prices recovered within a fortnight from the impact of falling oil prices, widely reported on as a period of global market turmoil. As can be seen from TABLE 3.3 above, investors brushed off the so-called Brexit crisis, with share prices recovering within a matter of days of the result being reported on June 24, 2016. Like the boxer who, after a series of onslaughts, realizes his opponent does not possess a knockout punch, investors are seemingly becoming less fearful of a surprise event from which share prices will not quickly recover.
The belief P/Es are used in investment decisions is more perception than fact
The best way I can describe what I believe is happening here is to tell a story of my experience with lifts, and more particularly the positioning of the buttons for the various floors. I once worked in a building with a large bank of lifts, including just the one lift going to the basement, which was rarely used. The regular lifts had two vertical rows of floor numbers, with the odd numbers on the left and the even numbers on the right. Because this lift going to the basement had an extra floor on the control panel, the even numbers were on the left and the odd numbers were on the right. Before I ever used the basement lift, I would have been certain I looked at the control panel and selected my floor number. After some time, I came to use the basement lift for the first time. It was quite a surprise to me to find the button I was looking at as I pressed my floor number was not my floor number, but where my floor number was in the regular lifts. I subsequently watched and observed many others make the same mistake. I think something similar is happening with the perception that P/Es are used in the decision to actually invest, when in fact that actual decision is determined by other factors.
As mentioned in Part 1, black swan events cannot be predicted, but it is possible to consider what events, brought on by a black swan event, might result in a drastic direct or consequential impact on the share market and on individual companies. I do believe an event that causes interest rates to increase significantly will have an adverse impact on share prices. For those companies that are highly geared, there will be a double whammy in the form of higher interest rates depressing EPS, with a flow on to share prices, together with higher interest rates depressing the price of shares generally. Low interest rates have continued for a long time now, and I do not see how this will change any time soon. And that is why I believe any change will likely arise from a black swan event. It will happen, but one could be a very long time out of the market waiting for it to happen. Or it could happen in the very short term. We need to prepare by understanding what actions we can take now to mitigate against adverse consequences. In Part 3 of this series, I will discuss the nature of various "yields", because just as "Oils Ain't Oils", "Yields Ain't Yields". Understanding the differences can lead to better investment decisions and will also contribute to decisions on actions to mitigate against the consequences of any event that leads to higher interest rates while not necessarily forsaking the higher yields available for shares versus bonds.
Disclaimer: The opinions in this document are for informational and educational purposes only and should not be construed as a recommendation to buy or sell the stocks mentioned or to solicit transactions or clients. Past performance of the companies discussed may not continue and the companies may not achieve the earnings growth as predicted. The information in this document is believed to be accurate, but under no circumstances should a person act upon the information contained within. I do not recommend that anyone act upon any investment information without first consulting an investment advisor as to the suitability of such investments for their specific situation.
Disclosure: I/we have no positions in any stocks mentioned, and no plans to initiate any positions within the next 72 hours.
I wrote this article myself, and it expresses my own opinions. I am not receiving compensation for it (other than from Seeking Alpha). I have no business relationship with any company whose stock is mentioned in this article.