Book Reviews, Economics

Book Review of Probable Outcomes by Ed Easterling

The full title of this book is “Probable Outcomes: Secular Stock Market Insights”. This is the 2nd book written by Ed Easterling. His first book Unexpected Returns is reviewed here, which explains in detail his Financial Physics model used in this 2nd book. It is recommended that readers start with the first book before reading this 2nd book.

Probable Outcomes uses Ed Easterling’s Financial Physics model to perform a scenario analysis as to the outcome at the end of this decade (2019). In his previous book, he identified that inflation is the one key driver of stock returns, as inflation impacts both the P/E and earnings, while other factors have remained stable and hence predictable. In this 2nd book, he highlighted that real economic growth rate has slowed down over the 2000s decade, and highlighted economic growth rate as the 2nd driver in his Financial Physics model. He puts forth 3 scenarios each for inflation rate and economic growth rate, which results in 9 outcome scenarios.

I like how the book explains how inflation and economic growth rate impacts the P/E and EPS of the stock market, and how that is modelled. While the book shows the 9 possible outcomes, the reader is left to figure out on his/her own what the eventual outcome would be.

One good thing that Ed carried over from his first book, is the use of full coloured pages printed on solid quality paper. That always gives a very good feel to the book.

Finally, one interesting bit that I learned from this book is that P/Es do not “revert to the mean”, they are driven by inflation. The fact that you see P/Es “reverting to the mean” when you plot a graph of historical P/Es and draw a horizontal line for the average, is simply because the average is duh, the average, which is why P/Es cycle around it. To forecast P/Es however, you don’t forecast that it reverts to the mean, rather that you forecast a level based on your inflation expectations.

Stock Market Remains Early in Secular Bear Cycle

  • As of 2010, the stock market remains early in the secular bear cycle (i.e. valuations are still too high)
    • S&P 500 P/E ratio at 19.9 (year-end 2009) is still above the average of 15.5x over 1900 to 2009.
    • S&P 500 Dividend Yield at 2.5% (2009 normalized dividend yield) is still below the average of 4.3% over 1900 to 2009.
    • Inflation (measured by CPI) is low at near zero, below the average of 3.3% over 1900 to 2009.
    • Interest rates (measured by 20-year Treasury yield) is near 4%, below the average of 4.9% from 1900 to 2009.

Real GDP Growth and Inflation are Two Major Uncertainties

  • There are two major uncertainties in forecasting how this current decade (2010 to 2020) will play out:
    • Major Uncertainty #1: Real GDP growth rate
    • Major Uncertainty #2: Inflation

Major Uncertainty #1: Real GDP Growth Rate

  • Almost all of an economy is the conversion of labour into goods or services. A small portion of the final goods is basic materials. As such, the two essential elements of real GDP growth are population growth (i.e. growth in quantity of labour) and labour productivity (GDP per capita).
  • While production drives consumption, which in turns drives production, the starting point is always the first conversion of labour into products.
  • Real GDP CAGR has been 3.3% from 1950 to 2009 as well as 1900 to 2009.
    • From 1900 to 1949, the 3.3% comprises of 1.4% population growth CAGR and 1.9% GDP per capita growth CAGR. Inflation in the GDP was 2.2%.
    • From 1950 to 2009, the 3.3% comprises of 1.2% population growth CAGR and 2.1% GDP per capita growth CAGR. Inflation in the GDP was 3.5%.
  • In the recent decades, 1970s, 1980s, and 1990s Real GDP CAGR were 3.2%, 3.0%, 3.2% respectively, however the 2000s Real GDP CAGR was 1.9%. The question is whether this 1.9% is the ‘new normal’.

Major Uncertainty #2: Inflation

  • Inflation is the overall level of increase in the aggregate price of goods and services across an economy.
  • Measures of inflation
    • Wholesale level (Producer Price Index and Commodity Price Index)
    • Retail level (Consumer Price Index and Personal Consumption Expenditures)
      • Consumer Price Index, all urban consumers (CPI-U) produced by the U.S. Bureau of Labor Statistics (BLS)
    • Economy level (GDP deflator)
      • Real GDP (GDP-R) measures the increase in the quantity and value of goods and services.
      • Nominal GDP (GDP-N) measures the actual price of all goods and services.
      • Inflation is measured by the GDP deflator (GDP-D) = GDP-N / GDP-R * 100. GDP-R is this year’s production expressed at the previous year’s prices.
    • GDP-D and CPI reflect fairly similar levels of inflation, but CPI does not reflect consumer substitutions that occur between periods while GDP-D adjusts to the actual economy of goods and services.
  • Deficits and Inflation
    • Compare debt growth with economic growth
      • Fiscal deficits add to national debt.
      • If the economy grows at a rate higher than the rate of debt growth (debt growth rate = deficit / existing debt), over time it will lead to a low debt ratio (debt / GDP).
      • If debt grows at a higher rate than the economy growth rate, debt ratio will ultimately be unsustainable.
    • Fiscal deficits can lead to increase in money supply
      • If deficits are permanently funded with government debt which are then monetized, money supply increases.
      • If the money growth exceeds economic growth, inflation arises.
      • Hence fiscal policy can lead to growth in money supply which can lead to inflation.
    • Economic growth leads to increase in money demand
      • Economic growth creates a demand for monetary growth.
      • If the demand for money growth (i.e. economic growth) > money supply growth due to fiscal deficits, inflation decreases. The Fed can then control the money supply to control inflation.
  • Velocity of Money
    • Effective money supply = Units of money * Velocity of money (i.e. no. of times money circulates around the economy).
    • U.S. Nominal GDP is ~2x the money supply, i.e. velocity is 2.
    • Velocity of money also impacts the level of inflation.
  • Deflation is extremely destructive
    • Deflation causes asset prices and nominal income to decline.
    • Debt interest and principal are usually serviced by income. Hence deflation would lead to loan defaults, which leads to liquidation and foreclosure of collateral, which leads to further asset price declines, and the spiral continues.
    • Most labor contracts do not adjust downwards with deflation. Hence obligations remain fixed. When funding sources decline under deflation, companies’ profits and government budgets get hit.
    • People anticipating deflation will sell their assets to pay down debt because they foresee asset values will decline. That causes asset values to decline further.
    • People will also delay purchases because they expect goods to be less expensive later. This slows the velocity of money and reinforces more deflation.


  • There is a long-term economic growth cycle (e.g. expansion and recession), and within that, a shorter-term business profit cycle. The business profit cycle (aka EPS cycle) is affected by industry competition, over/under supply/demand, etc.
  • Since corporate revenues is a proxy for GDP-N, and profit margins tend to fluctuate around an average level, corporate profits (i.e. EPS) have a strong fundamental relationship to GDP. However, the profit cycle is more variable and fluctuates around the more stable economic cycle.
  • The underlying long-term EPS grows at a rate which consists of inflation and real growth. Actual EPS will swing up and down around the underlying long-term EPS trend. The average over 1950-2009 of the 3-year CAGR of the S&P 500 EPS was 5.6%.

Measuring P/E

  • Normalized Earnings
    • To obtain the underlying long-term EPS trend, EPS needs to be normalized to ‘flatten’ the EPS cycle. Two methods:
    • Method #1: Robert Shiller (Yale): Trailing 10-year average for EPS in current dollars – Use reported EPS for the past 10 years, adjust each EPS forward for the effects of inflation, then average the adjusted EPS value across all 10 years.
    • Shiller’s approach produces a value that lags by ~5 years in real terms, i.e. it did not adjust for the underlying long-term real growth in the economy and earnings. Even after adjusting for inflation, the 10 EPS values are not yet “comparable” for averaging because those EPS values which are further back in the past needs to be adjusted upwards more to account for real growth.
    • If real growth is not accounted for, Shiller’s method will result in a ‘normalized’ EPS that is lower than the underlying long-term EPS. This is fine so long as it is compared with other measures obtained in a consistent manner. Generally, Shiller’s normalized EPS value should be increased by 15% to bring the inflation-adjusted EPS to the present in real terms.
    • Method #2: Crestmont: Regress EPS vs. GDP-N. The predicted value of EPS given a level of GDP-N, is a reliable normalized value.
  • Measuring the Price
    • If the objective is to assess the current level of valuation, P/E should include the current market index and a version of normalized EPS.
    • If the objective is to assess historical relationships or to develop a measure of the historical average P/E, the long-term series for P/E should be based upon the average index across each year (e.g. average of the closing price for each trading day) and a version of normalized EPS.
    • Shiller produces monthly data for the index by averaging across the days in each month.
  • Dividend yield as a cross-check
    • P/E calculated using operating earnings tend to overstate earnings because one-time losses are taken out but one-time gains are usually not adjusted.
    • In recent decades, the normalized payout ratio of dividends is about 45% of earnings. Note that payout ratio over reported earnings seems volatile, but it is relatively stable as a percentage of normalized earnings.
    • To validate the P/E, invert the reported dividend yield and multiply it with the average payout ratio, i.e. implied P/E = P/D * 45%, since D = 45% * E. The result should be similar to the P/E using normalized EPS.

Impact of Inflation on Valuation

  • P/E is driven by inflation, not by interest rates
    • Inflation increases earnings growth rate, but also increases the discount rate. Since inflation does not always transfer completely into earnings growth, and the discount rate can increase more than the inflation rate due to the added economic uncertainty that comes with high inflation, P/E multiples drop as experienced in secular bear markets.
    • Deflation lowers the discount rate but also lowers the future nominal EPS. The price of stocks drop further due to lower EPS than what is offsetted by the lower discount rate, hence P/E multiples drop. Hence for the deflation scenario, it is not true that lower interest rates mean a higher P/E multiple.
    • At low inflation, a fairly valued P/E will be high.
    • Hence the main driver of P/E multiples is actually inflation, rather than interest rates.
    • Even with deflation, the discount rate stops near zero and does not go negative because people have the ability to simply hold cash. Hence bonds do well during deflation.
  • P/E do not “revert to the mean”, the mean is not some “natural” level that P/E should go to. Because the mean is simply a statistical measure of the long-term average, people who think that P/Es revert to the mean will simply be right about half the time.
  • Secular cycles are driven by long-term trends in the inflation rate
    • Cyclical bull and bear markets are short-term surges or falls in valuation due to euphoria or crisis.
    • Secular bull and bear markets are long-term periods of revaluation due to shifts in the overall inflation rate.
    • When the level of P/E is inconsistent with the expected trend in the inflation rate, it is more likely positioned for a short-term cyclical move.

Impact of Real Growth Rate on Valuation

  • Assuming that (i) lower economic growth leads of lower earnings growth, (ii) long-term profit margins remain unchanged, (iii) inflation rate remains unchanged, (iv) expected return for stocks and bonds remain unchanged, then a decrease in EPS growth needs to be compensated by an increase in dividend yield, since the expected return remains the same. For dividend yield to increase, P/E needs to decline.
  • To estimate the required decline in P/E:
    1. Assume a normalized dividend yield of 45%.
    2. Convert the current P/E prior to the growth rate decline into an implied dividend yield (i.e. dividend yield = E/P * 45%)
    3. New dividend yield = current dividend yield + estimated reduction in EPS growth rate (e.g. new D/P = current D/P + 1%)
    4. Convert the new dividend yield into an implied new P/E (i.e. P/E = P/D * 45%).
  • If P/E starts relatively high, then a higher decline is required to provide the required dividend yield increase. If EPS growth drops by 1%, to increase the dividend yield by 1% requires P/E to drop from 22 to 15, 15.5 to 11.5, or 10 to 8.
  • [My note: Total return = capital gain + dividend yield = [ (new P/E * new EPS) / (old P/E * old EPS) – 1] + dividend yield. If we assume that P/E remains the same, the question will be to solve for the change in dividend yield required to keep the total return the same, in the event the growth rate of EPS decreases (not necessarily mean it goes negative).]

Possible Scenarios

  • 3 inflation scenarios
    • Deflation: inflation rate going from 0% to -5%.
    • Price stability: inflation rate staying constant at 1%.
    • Inflation: inflation rate going from 1% to 5%.
  • 3 economic growth scenarios
    • Trend: Growth rate of 2% for 2010s and beyond
    • Abberation: Growth rate of 3% for 2010s and beyond
    • Reversion: Growth rate of 4% from 2010 to 2019 (to compensate for the 1.9% growth over the 2000s), 3% afterwards
  • P/E impact
    • Deflation and Inflation scenarios will drive P/E down to ~10.
    • Price stability scenario will drive P/E to ~22.
    • Abberation and Reversion scenarios do not affect the P/E because the long-term growth rate is still 3%.
    • Trend scenario means lower economic growth and lower earnings growth. This lowers P/E in a manner explained above.
  • 9 scenarios (2010 to 2019)
    • Abberation x Deflation
      • P/E = 10 in 2019
    • Abberation x Inflation
      • P/E = 10 in 2019
    • Abberation x Price Stability
      • P/E = 22 in 2019
      • EPS ~= $85 in 2019
    • Reversion x Deflation
      • P/E = 10 in 2019
    • Reversion x Inflation
      • P/E = 10 in 2019
      • EPS ~= $113 in 2019
    • Reversion x Price Stability
      • P/E = 22 in 2019
    • Trend x Deflation
      • P/E = 8 in 2019
      • EPS ~= $55 in 2019
    • Trend x Inflation
      • P/E = 8 in 2019
    • Trend x Price Stability
      • P/E = 15 in 2019
  • Forecast real and nominal returns can be found in Crestmont Research’s article here.



No comments yet.

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s

Copyright © 2005-2014 All Rights Reserved.

Enter your email address to follow this blog and receive notifications of new posts by email.

Blog Stats

  • 399,703 hits
%d bloggers like this: