This quarter we take a deep dive into the fundamental prospects for developed world companies and how these prospects are currently being valued. There remains a mismatch between sentiment and fundamentals in our view, allowing us to continue to take a positive view of Developed Markets Equities within our recommended Tactical Asset Allocation.
Residents of the United States, please read this important information before proceeding
Please read this important information before proceeding.
Listen to the experts?
Persistent doubts continue to hobble the performance of global stock markets. In the US, earnings growth expectations are again sinking fast (Figure 1), with sales growth forecasts not far behind (Figure 2). Equity analysts are telling us that the horizon is darkening – are they right this time?
It is always worth remembering that analysts have not tended to be very good lead indicators over time. Their persistent tendency is towards excessive optimism, as demonstrated by Figure 3, but neither have they been very good at predicting recessions. If the hard economic data continues to defy shaky market sentiment, we may find that analysts soon revert to their traditionally sunnier role.
For our part, we still contend that our view of the true underlying revenue and earnings trends for the US quoted corporate sector has been obscured by the surge in the US dollar and plunge in oil prices seen over 2014 and 2015.
The difficulty here is how one disentangles the effects of currency movements on corporate revenues and earnings. How much of the hit to revenues simply results from translating overseas earnings back into that stronger dollar? What proportion is actually higher dollar prices deterring business with US companies? Hedging policies, both physical and financial, with disclosure sometimes limited, make these lines very hard to draw even with the benefit of some hindsight. We will know more as the corporate earnings roll in over the quarters of this year and next, though still not all obviously.
Our attempt to clean this up is necessarily rough around the edges – we have simply taken the 12-month trailing revenues of the US corporate sector and adjusted them for non-dollar exposure using the Fed’s broad dollar trade-weighted index1. While imperfect, the currency weightings of the broad dollar trade-weighted index approximate the international revenue exposure of the S&P 500 (with deviations of a few percentage points for major export destinations). This provides us with a rough but sensible approximation of the dollar translation effect (Figure 4). The results would support anecdotal evidence from a variety of globally exposed US corporations at recent earnings results2 – reported US corporate revenue growth has, of course, been slowed appreciably by currency translation.
Finally, we take our analysis another step by controlling for both the dollar effect and the bloodbath in commodity markets that has so decimated revenues within the commodity producing companies. As Figure 5 illustrates, the resulting revenue trends are probably best described as normal.
Many readers may wonder why we should make all these adjustments in the first place. Surely if we take out everything negative, we shouldn’t be surprised that everything looks so normal. The aim here is not so much to tell us what has happened, but what we might reasonably expect in coming quarters. If we accept that falling commodity prices tell us more about supply than demand (Figure 6), then revenues linked to oil and commodities are probably adding unnecessary noise. The same is true of the translational effect on overseas earnings, as this is an effect that, as noted above, should fall out over the course of this year.
What about profits?
The same is true for return on equity, a more holistic measure of corporate profitability, which has remained stable for both the US and World ex-US for four years now if we exclude resources (Figure 7). On the same basis, both US and World ex US earnings are still on an upward trajectory without any evidence of slowing (Figure 8) – further acquitting the US companies not engaged in commodity extraction. Importantly, Figure 9 shows the strong relationship between global output growth and return on equity. The inference here is that if growth is okay, then corporate profits are likely to follow suit. Many of those currently writing off the prospects for the US corporate sector mistakenly have the relationship the other way around.
Overall, we still expect the world’s major economies to continue muddling along, and this should continue to support sales and earnings growth for the world’s corporate sector, long after the base effects of dollar and oil fade.
What do valuations tell us?
The problem with valuation is one of context. Much of the debate around valuations assumes that history contains a precise answer to how much we should pay for a dollar, euro or pound of corporate earnings generated today. However, while history provides useful context and perspective for the investment debate, those looking for a precise playbook for today’s markets will likely be disappointed. For instance, if we compare the market’s current price to earnings ratios with the average of the last 10 years (Figure 10), we surely need to acknowledge that this period contains the most serious recession in living memory – is this useful context for a near future that probably does not contain such a record breaking decline in profits? Those looking further back will still find biases both up and down that dilute how carefully we should listen to the investment voices of the past.
Cyclically-adjusted PEs suffer from the same problem – why do we want to compare the S&P 500 of today with its forebear of the late 19th century when it was 12 railroad companies? Changing accounting standards and index composition tend to force relevant comparison towards the modern era, where the above mentioned biases start to take over. Meanwhile, Tobin’s Q is theoretically as well as empirically flawed, as explored in previous publications3.
This is where a form of relative valuation can come in handy. If, as theory would indicate, a company’s share price (and therefore that of broader equity indices) should encompass all future cash flows that this company/index will generate in the future discounted back to a present value, then on our conservative assumptions detailed in Box 1, US equities in particular, remain inexpensive. Such valuation metrics also have their critics of course, but by separating out some of the building blocks involved in valuing equity, some light is shed on some of the assumptions that the market may be making to justify current prices. As explored in a bit more detail below, this leaves us suspecting that there is room for bond yields to rise considerably and equities still to be unremarkably valued on well grounded growth assumptions.
“We still expect the world’s major economies to continue muddling along, and this should continue to support sales and profitability growth for the world’s corporate sector, long after the base effects of dollar and oil fade”
Box 1 - The dividend discount model
The dividend discount model (DDM) is a method of valuing a company’s stock price based on the theory that its stock is worth the sum of all of its future dividend payments, discounted back to their present value. In other words, it is used to value stocks based on the net present value of the future dividends. Within this model, there are three important moving parts: the expected growth this company or index will generate over its life, the level of profitability growth the company/index can sustain in the long-term, and how to discount those future cash flows to factor in the risk inherent in having to wait for them.
These DDMs can achieve tremendous sophistication, by modelling various periods with different characteristics – for example, periods in which companies undergo above-trend earnings growth. For us, we would prefer using the 1-stage Gordon Growth Model, which posits that the long-run fair value of an equity index’s share is determined by the formula below:
FV = D(1+G)/(ERP+BY-G)
FV = Fair value per share
D = Current dividend per share incl. net buybacks
G = Long-term nominal growth in earnings
BY = Post-crisis average risk-free rate
ERP = Post-crisis average equity risk premium
The reasons for preferring the 1-stage DDM (Gordon Growth Model) over more sophisticated multi-stage DDMs are two-fold. Firstly, the proportion of the fair value accounted for by the extra stages in multi-stage DDMs tends to be relatively small. Since the fair value per share is relatively insensitive to the extra stages incorporated, the potential benefits of increasing the model’s sophistication is muted compared to the costs of invoking the extra assumptions required. Second, we recognise that all DDMs are ultimately subject to many onerous assumptions, and therefore can only provide us with a long-run approximation of an equity index’s fair value. Based on this understanding, we think the simplicity of the Gordon Growth Model would provide protection against the danger of spurious precision, while being parsimonious at the same time.
Admittedly, plenty of subjective discernment is required when it comes to selecting the appropriate figures for the long-run variables. By far the most contentious of them all is the ERP, which is given by the formula below:
ERP = DY+G-BY
DY = Dividend yield incl. net buybacks
G = Long-term nominal growth in earnings
BY = Current risk-free rate
For the long-run nominal growth rate (G), we opted for a historical growth rate of 4% based on long-run S&P 500 earnings growth since 1871. Alternatively, one might opt to use a higher growth rate of 6% based on a shorter timeframe of 50 years (1966-2016), depending on how conservative one desires to be4. Meanwhile, our preferred risk-free rate is approximated by the current 10-year US Treasury yield. Historically, both the dividend and net buyback yield has hovered around the 2% level over the past fifteen years5. Putting them all together, this yields a sensible ERP range between 0% - 7% for the S&P 500 between 2000-2015 (Figure 11). Interestingly, the US ERP has been trending above its long-run average of 4% ever since the post-crisis economic recovery. To take account of this, we would prefer to settle with a post-crisis ERP average of 5.5%.
Putting these numbers together would imply that as of late-March, the S&P 500 index is undervalued by approximately ~5%. Furthermore, by tweaking each variable, we can obtain the varying degrees of undervaluation/overvaluation for the S&P 500 corresponding to each alternative scenario. From our sensitivity analysis, we see that the S&P 500 is currently inexpensive according to most plausible alternative scenarios (Figure 12).
Another US and global recession is of course inevitable, but in our view not yet imminent. If we are right about the latter point, then there may be a disconnect for investors still to exploit. As explored in more detail above, we do not have to work hard in terms of building block assumptions to find upside in our Dividend Discount Model. While other valuation measures are perhaps less conclusive, we remain comfortable concluding that equities are inexpensively valued. The clear signs of mistrust in corporate fundamentals, highlighted by the outperformance of ‘safe’ dividend paying companies (S&P dividend aristocrats), is misplaced in our view (Figure 13). The still steady relationship between free cash flow (aka ‘cash earnings’) and earnings (Figure 14), with previous deteriorations in this metric proving a good indicator of deteriorating corporate fundamentals, would support this view. Our strong overweight recommendation on Developed Markets Equities remains well-founded. Our preferred regions remain the US and Europe ex UK.
2 Factset Earnings Guidance – Q4 2015
3 Compass, December 2013 and January 2014
5 How dilution and share buybacks impact equity returns – J.P Morgan, 2014