Thursday, May 19, 2011

Grantham's Bear Case - Ari Paul


Shortly after I posted Siegel's bullish narrative last week, a friend (thanks Alejandro) sent me a recent essay by Jeremy Grantham that serves as the perfect counterpoint to Siegel's bullish narrative. Grantham runs GMO and manages about $110 billion in assets. He is known for his long-term approach to predicting asset returns (latest 7 year forecast) and his bearishness over much of the last decade. In the essay (part 1 here, part 2 here), Grantham argues that the entirety of Jeremy Siegel's bull case is really just a 200 year hydrocarbon bubble. Grantham believes that global GDP growth has been above trend for two centuries because of the discovery of incredibly cheap energy...and that the age of cheap energy is over.

Grantham's argument:
The recent boom in commodity prices has reversed the downward trend of the last century. Even as the population increased more than fourfold over the past one hundred years, commodities produced negative returns because we were discovering and innovating even faster than we were consuming. The recent supply fundamentals suggest we're entering a new era. Productivity growth in agriculture has fallen from 3.5% a year to 1.5% a year. Growth in the oil supply has nearly stagnated and the average cost per barrel has more than doubled in real terms in the last 10 years, after staying roughly steady for 60 years. Every extra barrel of oil, ounce of copper, and bushel of wheat is becoming exponentially more expensive.

Economists generally believe that every $10 increase in the price of oil reduces US GDP growth by about 0.25%. The demand for oil by emerging economies (especially China), continues to rise at a dramatic rate, while supply growth stagnates. What will keep oil from climbing over $140 and eliminating US GDP growth? Optimists hope for sudden innovation in alternatives like solar or wind, but any transition will require at least 10 years to have a meaningful impact, and probably more like 20 years.

Grantham also focuses on the impossibility of sustained compound growth. As a colorful example, he asks us to imagine what would have happened if the Ancient Egyptians started with a cubic meter of physical wealth (say gold) and compounded it at a rate of 4.5% a year. How much would they have today? Their gold would completely fill more than one thousand solar systems. His point - if our demand for resources rises at all with our growth in GDP, then indefinite compound growth is impossible, and will catch up to us sooner than most realize.
Speaking practically, Grantham believes that bets on resource production and resource efficiency will pay off over the long-term. However, over the next 18 months he thinks there is a significant risk of commodity prices falling sharply because of a "blip" in Chinese growth, better weather, and the end of QE2.

To paraphrase Warren Buffett, it's never paid to bet against American ingenuity. Malthus predicted that we'd all die from famine because the world couldn't support a population of more than a billion people. Today we have nearly 7 billion with less famine than in Malthus' time. Who knows what innovations and discoveries tomorrow will bring?

Sunday, May 15, 2011

Economic Commentary: Siegel's Bullish Argument - by Ari Paul


I returned to my alma mater, Upenn, for my 5th reunion this weekend. I had the pleasure of attending a lecture by Jeremy Siegel. Siegel is a Wharton professor, the author of "Stocks for the Long Run", and the leading researcher of the returns of asset classes. He made a convincing argument that stocks are currently fair to undervalued and explained why the market never overshot as far to the downside as many bears expected in 2009.
The cornerstone of Siegel's research is the above graph. Invested over 200 years, a $1 investment in gold would be worth just $4.02 today (adjusted for inflation). A $1 investment in treasury bonds would be worth $1,530. A $1 investment in equities would be worth $700,000. While the equity returns are overstated because Siegel does not adjust for taxes or trading costs, the basic conclusion is valid. Perhaps more impressive is the consistency of the returns. The average annual real equity return has been 6.7% (after inflation) over the past 200 years. It has also been between 6 and 7.5% over any 50 year period you look at.

He recently extended the study to look at 16 countries with 100 years of data. While equity returns had a bit more range (from 3% to 9%), the difference between equities and bonds was remarkably consistent. Countries with an average real equity return of 3% had treasury returns near 0; countries with a 9% equity return had treasury returns of about 5%.

So, are stocks cheap? According to Siegel's 200 year logarithmic stock chart, stocks are currently slightly below trend. He also discussed the traditional P/E metric and said that the average is 15. In a low interest rate environment like we have now, 18-20 is more common. The current PE of the market is about 13, suggesting the market is perhaps 35% undervalued (more on this later). Siegel suggested that many stocks in China are remarkably cheap; the top 10 dividend paying Chinese stocks have a yield of over 5%, PE ratios of <13, and expected growth of 12% a year.

The role of interest rates also explain why the market failed to overshoot to the downside as much as prominent bears like Jeremy Grantham and Cliff Asness expected. In the 1970s and great depression, P/E ratios fell below 8. In March of 2009, they bottomed around 10. Siegel argues that in the 1970s and 30s, long-term interest rates on government debt provided a much more attractive alternative. When interest rates are 6%+, investors are much more willing to dump stocks. With the incredibly low interest rates in 2009, pension funds were unwilling to reduce their stock holdings; if they locked in a return of 3% in bonds, they would effectively be admitting to insolvency since they require a higher a return to meet liabilities.

As for future returns, Siegel is optimistic for structural reasons as well. His research suggests that long-term stock returns depend on long-term GDP growth, and long-term GDP growth depends on productivity growth. While short-term GDP growth is calculated as consumption + investment + government spending, long-term GDP growth depends almost entirely on productivity growth. Productivity growth depends on the number of people in a position to innovate and their ability to communicate. For example, before WW2, women were generally not part of the research and entrepreneurial community in the US. When we double the number of potential innovators, we more than double the rate of innovation because of network effects.

Today, technology allows a professor at Wharton to collaborate easily with a professor at South China University. The number of global innovators and their ability to collaborate is growing at fantastic speeds, which should lead to rapid increases in productivity, which will support strong global GDP growth and thus strong global stock returns.

Now for the criticisms of these arguments -
Siegel naively trusts earnings. While the earnings of non-financial US companies are of high quality today, the earnings of Chinese companies are certainly not. A significant percentage of Chinese stocks that are listed on US exchanges directly or through ADRs are reporting fraudulent financial data. If the PE ratios of Chinese stocks were really 12, they would be great buys, but the real PE is higher. They may still be great buys, but we need to dig deeper.

Second, Siegel makes a compelling argument that the low interest rate environment of 2009 is what generated the high valuation levels (relative to what bears expected). He also acknowledged that interest rates are likely to rise significantly (he guesses 3%) over the next 2 years. That means that while stocks may have been a great buy in 2009, they are not any longer as the market will (and possibly is already) anticipating the future higher interest rates. Given his own methodology and interest rate assumptions, fair in the S&P 500 is 1320 to 1530 (versus a current value of 1337.