Thursday, September 5, 2013

September 2013: What Will Drive Equities Higher In The Next 12-Months?

This rally, which started in 2009, has now run 1640 days, which is well over double the average 'cyclical bull market' of the last 80 years (article). It is, in fact, the third longest cyclical bull market of the post-depression period. The percentage gain (140%) is also nearly twice the average and the third highest.

It is, in other words, the right time to ask what, if anything, will drive the indices higher before they instead head lower.

(Note: 'Cyclical' bull markets are distinct from 'secular' bull markets. A secular bull market (1982-2000) is generational and comprised of several cyclical bull and bear markets, each lasting two years on average (as above). Read further here or here. This post is only focused on the cyclical trend).

There have been two distinct stages to this bull market.

Stage One ran from 2009 to late 2011 and was driven primarily by fundamentals: strong sales growth and strong earnings growth (EPS doubled; sold blue line below). In Stage One, fundamentals outpaced the appreciation in the indices.

The second stage has run from late 2011 to the present. Fundamentals have slightly improved but Stage Two was overwhelming about the indices becoming revalued upwards. Since the end of 2011, less than 20% of the gains in the SPX can be accounted for by EPS growth; the remaining 80% has been multiple expansion. The chart below shows the rise in the trailing 12-month (TTM) PE; it has increased by one-third in Stage Two.

If fundamentals drove Stage One and valuation drove Stage Two, what will drive the next leg higher?

To simplify, think of the market as being driven by three primary variables: sales, margins and valuation multiples. For the market to go higher, some combination of higher revenue growth, higher profit margins or higher multiples will be needed.

We will look at each in turn below, but here are the main conclusions:

  1. Sales growth, like economic growth, has been slow for the past two years. The good news is top-line growth might increase in the next year. The bad news is that sales growth has the least leverage on future returns.
  2. Margins, on the other hand, are already at the top of their long term range and are unlikely to expand further; in fact, they may well contract as interest expense and labor costs rise. Moreover, margins have a very large impact on future returns; even a small drop in margins can negate a large increase in sales growth.
  3. Valuation multiples are also already at the top of their range based on a number of measures. A substantial expansion in multiples would be an outlier event. Multiples, like margins, also have a highly leveraged effect on future returns.
Net, expecting the double digit returns of the past several years to continue over the next 12-months is not odds-on. The best-case realistic assumption is that equities appreciate no faster than the growth rate of their revenues (2-4%); there is also a reasonable likelihood equities decline.

Sales growth is the most likely bright spot

Sales growth is typically fast early in a bull market, and then decelerates. What is remarkable with the current rally is how quickly sales growth decelerated (chart below shows year-over-year SPX sales growth). 

During the last cyclical bull market (2003-07), sales grew an average of 7% per annum. Likewise, sales growth in 2009-10 was more than 10%. But in 2011 it slowed to 2% and it declined further to 1.7% in 2Q13. 

Forward guidance for FY13 sales growth was 3.2% in March. This dropped to 2.4% in June. Now, its even lower, at 2.1%. That will make two back-to-back years of 2% sales growth.

If you are looking for a macro reason for sluggish sales growth it is that the financial crisis wiped out about 40% of American household net worth (here). Regaining consumers' financial footing, and therefore their ability to spend, has been slow. This was expected by those who have studied prior financial crises.

A dramatic reacceleration of sales growth in FY14 is not likely. Analysts currently expect 4.3% growth but history suggests that these estimates will decline as the next year draws closer (chart).

The macro data are very consistent with decelerating SPX sales growth of 2%. GDP (net of inventories) is at the lowest level of growth outside of the start of a recession in the post-war period (article):

You can similarly look at durable goods orders (chart), consumer expenditures (chart) and rail volumes (chart), all of which tell the same tale of decelerating, sluggish growth.

Despite this, there is some reason for optimism. Roughly half the SPX's sales and profits come from outside the US. A recession in the Eurozone has been a drag over the past 2 years (first chart) and the latest figures are encouraging of better growth ahead (second chart). Not a dramatic reacceleration, but enough to make 2-4% growth feasible.

So, the good news is that revenue growth, which has been 2% the last two years, can reasonably be expected to be at that level or higher (4%) in the next year, as analysts currently expect.

What does that imply for SPX over the next 12-months?

The chart below looks at margin assumptions (horizontal row) and sales growth (vertical). We will use this same model throughout this post. Here's how it works:

  • The first table (top left, titled "1. EPS") calculates the resulting EPS for SPX in 12-months time based on sales growth and margin assumptions.  
  • The second table (top right, "2. EPS Chg yoy") calculates the year-over-year growth in EPS from today based on table 1. 
  • The third table (bottom left, "3. SPX (FTM)" calculates the level of SPX in the next 12-months (FTM: forward 12-months) based on the current PE of 15.6. 
  • The fourth table (bottom right, "4. SPX Chg yoy") calculates the resulting year-over-year appreciation in SPX from today based on table 3.
If margins (9.5%) and multiples (15.6x) stay the same as today, then the SPX index would be 1700-1730 in 12-months time (blue box, table 3). This is equivalent to 3-5% appreciation from today (blue box, table 4). This is equal to the growth in EPS (table 2).

Two conclusions are worth emphasizing:
  1. First, 3-5% in appreciation in SPX is markedly lower than what investors have become accustomed to over the past four years (i.e., 25% compounded annually). 
  2. Second, sales has the least impact on SPX of the three variables. Even doubling the sales growth rate (from 2% to 4%) raises appreciation to just 5%. As you will see, margin and multiple have a much greater impact on appreciation. 

Margins are not likely to expand, and may contract

Margins rose rapidly in 2009-2011 and then flattened at 9% - 9.5%. Moreover, this is the same level at which margins peaked during the 2003-2007 bull market.

Analysts are nonetheless bullish that margins will rise another percentage point over the next 12-months, to 10.3%. Is this reasonable? There are some reasons to be very skeptical.

Margins are not just at 10-year highs, they are at 70-year highs. An assumption that they can go ever farther from the historical average is high-risk, for several reasons.

Why are margins so high?

First, because interest rates have been historically low, thus reducing interest expense by at least half their levels from the last bull market. For example, in 2012, interest expense was 1.8% of sales for the SPX; in comparison, the 15-year average is 3.9%. As rates rise (a process that has already started and which would continue if revenue growth accelerates), margins could fall 1-2 percentage points.

The second reason margins are high is wage expenses are low. Wage expenses are currently at over 50-year lows. Again, if economic activity increases, unemployment will fall and wage inflation will increase, lowering margins.

So, the bad news is that margins are unlikely to increase and, in fact, will likely be under pressure going forward.

What does that imply for SPX over the next 12-months?

Turning to the same model as above, if sales growth rises (4%) and multiples stay the same as today (15.6x), but margins fall by even 50bp, the SPX index would be at same 1640 level it is today in 12-months time (blue box, table 3). If margins fall a full percentage point, SPX would be 1550 next year (6% lower; blue box in table 4).

Two conclusions are worth emphasizing:
  1. First, margins have high leverage on SPX's forward returns. Even dropping a half percentage point in margins, to 9%, would negate the positive impact from sales growth doubling from 2% to 4%.
  2. Second, flat appreciation, or even a fall in SPX, would be unexpected by most investors accustomed to double digit growth each year. 

Much higher valuation multiples than today's have been outlier events

Because forward guidance is subject to significant revision (chart), we prefer to look at trailing 12-month (TTM) valuation multiples.

During the height of the 2003-2007 bull market, the brief peak PE was 16.5x. In the chart below, multiples can be seen to have reverted downwards every time after exceeding 16x (red shading) during that period as well as at the July 2013 market peak.

Note: 2009 multiples are high as 2008 profits collapsed from the prior year, then quickly recovered in 2009 and onwards

How does 16x compare to historical norms?

Since 1870, the median PE has been 14.4x (average of 15.5x). But this masks significant changes caused by the technology bubble in 2000. Before 1990, the median PE was 14x; since 1990, it has been 21x.

What is clear is that since 2000, multiples have been declining towards their historical norms, and that means that 16x is slightly (about 1x) above average.

It's also clear that periods of above average PEs (like the past 20 years) alternate with periods of below average PEs.

Can we expect multiples to go much higher above the mean? The short answer is probably not, for several reasons.

First, PEs are correlated inversely to treasury yields; the recent period of higher PEs corresponds to a secular decline in yields since 1980. The reverse is true prior to 1980. If yields are beginning an uptrend, PEs will be pressured to decline (or at least not rise).

The financial rationale for this relationship is this: market values are the sum of future discounted cash flows. As yields rise, the discount rate rises and the value of future cash flows decline.

We can also calibrate current market valuations using other methods besides a 12-month PE. One is a 10-year PE, also called a Graham-Shiller PE, that smooths out cycles by using EPS over the prior decade (article). Based on this approach, the current valuation is 24x, about 40% above the mean (16x) and in the top 89th percentile over the past 140 years. Higher valuations than today's have been outlier events.

Another valuation metric is Total Market Cap/GDP (Warren Buffett's favorite metric). At 108% of GDP, the current market cap is about 1.5% below the 2007 peak and well above the ratio's levels before the mid-1990s. Higher valuations have only occurred once before (i.e., during an outlier event). This implies that at current multiples, the market will likely grow at the same rate as GDP (i.e., 2%) going forward.

Similarly, price to sales is now at the post-bubble highs and about 7% above the average, meaning, again, that current multiples imply equity appreciation at the rate of sales growth (2-4%).

Finally, using Tobin's Q, which measures market capitalization relative to business replacement cost (explained), equities are priced at their high end versus the past 100 years. Higher valuations than today have been outliers. Like the other measures, it means that multiple expansion is unlikely to be the main source of future growth but, instead, it will have to come from growth of the business itself (in this case, it's balance sheet).

So, in the absence of an outlier event like the technology bubble, valuation multiples are not likely to substantially expand from current levels. Optimistically, a range low of 15.5x (like today) to a range high of 16.5x (the post-bubble peak) is realistic.

Turning to the same model as above and assuming a high PE (16.5x), the maximum upside in 12-months is 1730 (+5%). If margins fall to 8.5%, even with sales growth of 4% and a higher PE, the index would be flat.

Likewise, assuming a 16x valuation, future values range from 1590 (-4%) to 1680 (+2%).

At current multiples, the index would range from 1640 (flat) to 1550 (6% lower).

Two conclusions are worth emphasizing:
  1. First, valuation multiples, like margins, have high leverage on SPX's forward returns. 
  2. Second, even if multiples expand to the 2007 peak of 16.5x, the upside to SPX is just 5% (1730). If multiples are capped at 16x (which has been a notable barrier since 2005), the index is likely to be flat to lower next year. 

* * *

2013 will in all likelihood end with above average gains. This follows the above average gains last year. Since 1960, annual sequential gains above average (green arrows, below) have tended to be followed by below average gains in the third year (i.e., mean reversion; red arrows). The fundamentals outlined in this post suggest that this a likely occurrence for next year. 

The only times in the past 50 years that the US indices have had more than two sequential above average years was at the end of the 16-year 1949-65 secular bull market and the end of the 18-year 1982-2000 secular bull market (yellow arrows). 

Valuation is the last factor on our order-of-priority Weekly Market Summary table each week. It is not that fundamentals are unimportant; in the long run, they are the most important factor driving equities. But using it as a timing mechanism is difficult, and exuberance can render all calculations moot. In short, valuation is like brushing your teeth; you know it is important, but the risk is in the indeterminate future.

Still, the message here is clear: fundamentals have been a nice tailwind for equities the past four years; it is more likely to be a headwind over the next 12-months.