Rule of Thumb: Valuation Edition

A little over a year ago, we wrote about a simple “rule of thumb” calculation that Vanguard founder Jack Bogle has discussed in order to quickly predict 10-year future annualized market returns.  With markets having rallied significantly over the past 12 months, we thought it might be interesting to quickly run back through the "back of the envelope" calculation and see what it’s telling us about return prospects over the next decade in US and global markets.  Then, we’ll look briefly again at a few other US-focused valuation indicators to see how the "back of the envelope" calculation lines up with those that have a long statistical history.  

The Bogle Rule, as we’ll label it, states that one can approximate annualized future 10-year market returns by adding together predicted annual nominal GDP growth and the dividend yield at the start of the period, then adding or subtracting an annual percentage based upon the distance of current P/E valuation from the historical average, thus accounting for multiple expansion or contraction. The expansion/contraction factor is calculated "back of the envelope" by dividing the percentage distance from the median by 10.

Here’s a quick and dirty example.  If nominal GDP growth (GDP growth not adjusted for inflation) is expected to be 5% per annum, the current S&P 500 dividend yield is 2.5% and the market happens to trade 20% below median, expected annual total returns over the subsequent decade would be 5% + 2.5% +2%, or 9.5%.  

How do things currently stack up right now in the US and among the major international equity indices?

For the US, we’ll go with 2.5% real GDP growth over the next decade per year, and 2.5%, giving us approximately 5% nominal GDP growth.  The current S&P 500 dividend yield is 2%.  Using the Shiller long-term 10-year P/E, US markets are currently overvalued.  The current P/E is 24.3x vs. the long term median of 16.5x.  The P/E would have to contract by 32% total to get back to median levels.  We’ll lop off 3% per year from the "back of the envelope" calculation for multiple contraction.  5% + 2% - 3% gives us an annualized total return calculation of 4%.

Across the broader EAFE, which encompasses developed economies in Europe and Asia, we’ll assume slightly lower economic growth and inflation prospects than the US due to structural economic issues and assign a value of 4% for nominal GDP growth.  The current dividend yield for the MSCI EAFE Index is 3%.  The MSCI EAFE is currently trading at 19.1 approximately 13% above the long-term median.  We’ll remove 1% per annum for multiple contraction.  The back of the envelope calculation comes in at 6% per annum.  

Turning to the MSCI Emerging Markets Index, we’ll assign slightly higher values for real GDP and inflation than the US and give the emerging markets roster a 6% per annum nominal GDP forecast.  The MSCI Emerging Markets Index’ current dividend yield is 2.6%.  Furthermore, the index is currently trading at 15.6x, approximately 6% below typical long-term median equity market valuation.  As such we’ll add 0.5% per annum for multiple expansion.  Overall, the back of the envelope calculation comes in at 9.1%.  

Surely, these assumptions could be wildly off the mark.  Therefore, as a point of comparison, let’s compare the US numbers to predicted US values derived from two valuation metrics with a long standing statistical backdrop, the Shiller P/E and the Q-Ratio.  

As mentioned above, the current 10-year P/E for the S&P 500 is 24.3x.  At the current levels, the predicted value of nominal total 10-year future annualized S&P 500 returns is 5.3%, below the long-term median 10-year annualized return of 9.1%.  The Q-ratio, which is basically an approximation of the traditional Price to Book ratio using market book values, stands currently at approximately 1.00.  At the current level, predicted 10-year annualized nominal total returns come in at approximately 4%, in line with our simple prediction above.  

Keep in mind that the statistical relationships between the Q-ratio and Shiller P/E and future 10-year market performance are reasonably strong.  Correlation for the Q-ratio and returns is -0.72 (the negative correlation tells us that higher valuation produces lower returns and vice versa).  Correlation for the Shiller P/E is -0.68.  

Overall, it doesn’t appear that our back of the envelope prediction for the US is outlandishly off the mark.  Take all three indicators, two of which have a strong statistical history, and we think it’s safe to say there is a high probability future S&P annualized returns will come in decently below the long-term average of 9.1%.  On the other hand, international markets, especially emerging markets, seem to be positioned for better 10-year annualized returns from this point in time.  

Valuation and the potential for multiple expansion or contraction play a big part in these forecast differentials.  Multiple expansion and contraction have always figured prominently in long-term return outcomes.  It may take a while sometimes, but eventually investors must reckon with reversion to the mean.

Again, and we can’t emphasize it enough, 10 years is a long time and markets don’t move in neat straight lines.  Nor does over or undervaluation mean that markets will begin moving in the short-term.  Take October 2003.  At that time, the Shiller P/E stood at a more overvalued 25.7x, with a predicted annualized total return per annum of 4.7% over the ensuing 10-years.  Total annualized returns came in at a better than predicted 6.8% per annum (still approximately 2.5% per annum below the long-term median).  As we all know very well, the market movements that transpired over that 10-year period were very messy (to put it kindly).  From that overvalued position in October 2003, markets continued to rally for another four years before dropping a harrowing 60% in 2008 to 2009, only to rally sharply from the March 2009 lows to the present.  

It’s better to think about these long-term prognostications with a generalist perspective.  In the US, if you’re thinking about planning for retirement or your child’s college education in 10-years, it’s probably not smart to assume that we’re on the verge of a 1980s/1990s super-bull performance repeat.  Alternately, for the doom and gloomers of the world, don’t necessarily count on massive market carnage; there’s a good chance that the market produces decent, if unexceptional, returns. 

It’s the economic data, Stupid! Or, wait, is it??

Turn on any financial, news, or political program in America, and inevitably hosts and guests will spend a good portion of the time discussing economic growth and its relationship to market performance.  Of course, all of us got an extra helping of this type of analysis while the budget shutdown was in force, with the general theme being that governmental infighting will hold back the economy, and in turn, the ability of the S&P 500 to reach new highs this year.  

In practice, a historical statistical relationship between market performance and GDP growth is non-existent in many cases.  Comparing annual S&P 500 returns with annual US nominal and real GDP figures from 1950 through 2012 produces some interesting results using simple regression analysis.  Here’s a quick summary of the results:

• From 1950 to 2012 using annual data, there is basically no statistical relationship between yearly real and/or nominal GDP and S&P 500 performance (correlation coefficients of -0.05 and 0.11, real and nominal)
• Likewise, there is very little if any statistical relationship between trailing 10-year compounded GDP (nominal or real) growth and 10-year trailing S&P 500 returns (correlation coefficients of 0.21 and -0.23, real and nominal).  
• There isn’t a statistical relationship between GDP in a given year and S&P performance in the following year, nominal or real.
• There is, however, a decent statistical relationship between S&P 500 performance in a given year and real GDP growth in the subsequent year.  This relationship did not hold up when running the analysis with nominal GDP numbers (correlation coefficient of 0.63 using real GDP, but only 0.19 using nominal).  

Here are the basic numbers:

What are some of our quick takeaways from this simple statistical run?

• Making investment decisions based upon current or prior economic data is probably a waste of time (and opportunity).  Certainly many prior studies have demonstrated this, but it never hurts to repeat.
• Equity markets seem to be a decent discounting engine when it comes to future economic growth.  Markets seem to get ahead of the data on economic growth.  This brings us back to point one.  Bottom line: if you’re waiting to see the whites of the economic data’s eyes before making a decision, you’re probably going to be late to the party.  
• With an R-squared value of 0.395 (which basically says that S&P 500 performance “explains” approximately 40% of next year’s GDP), that leaves a lot of room for other variables like valuation or interest rates.  Our gut feeling here is that even if one has particularly accurate predictive capacity on GDP growth in future years, the investor with that information could still end up with a poor investing outcome, all things being equal.  Short-term market decision-making based upon predicted future economic outcomes alone is probably a losing proposition for most investors.
• We do know that statistically, 10-year CAPE P/E values have a decent statistical relationship to 10-year future annualized market returns.  Soaking it all in, an investor is much better off paying attention to general valuation levels and their potential impact on future returns than wringing hands over economic data releases, especially when valuations reach extreme levels (perhaps greater than 25x on the overvaluation side or less than 10x or 12x for undervaluation).  Of course, the longer term view and patience is essential when it comes to using valuation.  Overall, enjoy your days and don’t let the latest BLS release on employment or the latest government GDP release ruin your day or month.  It does make for good water-cooler talk and keeps the TV talking heads in business, though.
• If the market is a decent discounting mechanism and moves ahead of the data, not with it, and one is concerned about managing downside risks, it might behoove those interested in risk management techniques to explore employing simple longer-term technical trading rules, such as simple moving average triggers.  As discussed in the past, some intermediate to longer-term moving average rules have produced solid historical outcomes from a risk-adjusted return basis.  The market as a whole knows more than we do individually.  If markets are showing signs of breaking down, it’s often not wise to fight the tape.  The same can be said for rising markets, as demonstrated in recent months and years.  
• Ultimately, the market doesn’t care about our theses based on current or backward looking information and makes mince meat out of those stuck in the analytical gobbledygook.  Many investors were blindsided in ’08 looking backward and accepting the “all is well” economic news at the time.  Likewise, many investors have missed the current rally hung up on economic and political news flow.  Odds are the next market calamity, like the prior ones, will only become truly apparent in hindsight.  And, like the past, most investors will end up selling at precisely the wrong moment by using real-time economic data (and by paying attention to the talking heads in real time).

Valuation Update:

Once again, it’s been a few months since we updated our valuation tables.  As in the past, we’ve included stock market indices for 13 countries across the developed and emerging market spectrum.  We also throw in a few developed and emerging market indices to show how the broader international areas/regions stack up against one another.  We take three indicators broadly used by many value-oriented practitioners—10-Year CAPE P/E, Price to Book, and Enterprise Value to EBITDA—rank the countries in each category from highest valuation (relative overvaluation) to lowest valuation, then average across the categories to provide an overall score to determine position.  The lower the average score, the higher the relative overvaluation and vice-versa.  All ratios are based fundamental data provided by Bloomberg.

Here are updated results as of 10/4/13:

Valuation Conclusions:

• While the overall MSCI Emerging Market index has taken a relative beating this year, reflected in the fact that the MSCI Emerging Market Index sits in the lower portion of the table, India remains the most overvalued across all three indicators on a relative basis.  Among the BRIC country indices, the India SENSEX Index has held up reasonably well over time.  Brazil, Russia, and China indices are all decently below 2008 highs, but the SENSEX has held its ground, even though overall economic and market prospects have become cloudier.  Perhaps the valuation tables above indicate that Indian markets will play catch up and pay a price over the next several years in terms of poor relative performance.

• The US comes in at a close second to India in terms of relative overvaluation.  We’ve discussed in past valuation posts and in our chartbook that historical analysis shows that future 10-year total annualized returns for the US are predicted to fall three to four percentage points below the long-term historical average per annum.  Australia, UK, and Japan round out the top five portion of the table when it comes to individual countries.  Japan has witnessed a phenomenal run over the past year on the back of the Abenomics announcements.  Fundamentals are going to have to start catching up, though, for momentum to continue.  Japanese markets have basically moved sideways over the past five to six months and momentum is waning.  Australia benefitted mightily from China’s growth over the past decade.  There could be a moment of truth for Australian investors as markets adjust to new realities about China’s future economic path.  The UK has rallied in recent years, but equity market prices have remained ahead of broader economic realities.

• Within the developed market category, France, Spain, Italy, and Greece continue to occupy the bottom spots, ex-Russia.  Valuations have recovered off the rock-bottom levels observed during the main portion of the European crisis.  Still, markets continue to price in significant pessimism.  These markets have a strong chance of providing outsized relative returns over the next decade, if history is any guide.  Economic momentum is beginning to turn in the periphery countries.  Like an individual value stock, all it takes is a few better than expected economic surprises to attract investor capital into the vacuum.  

• Looking at the broader indices, the MSCI EAFE, representing developed markets ex-US, scores better than the United States, while the Emerging Markets Index remains the most attractively valued on a relative basis.  Breaking down regional international indices, Europe scores better than Asia. 

• Russia continues to occupy the bottom of the table.  Like the broader Emerging Markets Index, Russia’s equity markets have basically treaded water since 2006/2007 performance wise.  Nonetheless, Russia has rarely been able to command robust valuation multiples over time.  Russia’s reputation among international equity investors remains very spotty due to numerous issues surrounding trust, corporate governance, misappropriation, corruption, and other issues.  

• Based on valuation alone, we’d still argue that Emerging Market equities should provide better overall total returns over the next decade compared to their developed counterparts, though investors would probably be wise to also look beyond the BRIC countries for opportunities.  Within developed, we continue to believe that European equities will outperform their Asian or US counterparts over the next several years, most likely driven by rebounds in the periphery.  Again, as the southern European economies begin to show signs of stabilization, there could be significant room for multiple expansion.

• Our standard “disclaimer” applies as usual.  Many fundamental indicators prove their efficacy over multi-year time frames and, thus, shouldn’t be considered great short-term timing indicators.  Don’t take this as indication that we expect Indian markets to collapse 50% over the next 12 months, or that we expect Italian markets to skyrocket.  From this point forward, over time, history shows the odds work against the countries/indices at the top of the table and for the ones at the bottom.