Q2 2012: The Allocation of Capital

In the long run, valuation drives returns. And over the long term, asset prices have a strong relationship to economic factors like earnings and interest rates. However, in the short run, prices tend to disconnect from economic fundamentals and are affected by a myriad of inputs. This can be readily observed by comparing the movement of a country’s stock index relative to the change in its economic growth. In a normal year, the US economy grows by 2-3%, with most years ranging between -1% and +4%. US stocks will return 7-12% in normal years, with outliers ranging from -30% to +20%.

Asset class returns tend to revert to long term trends over time. When an asset class (such as real estate) averages mid-single digit growth for hundreds of years and then suddenly returns 20% per year for 5 years, there is a high probability of low or negative returns in the years to follow.

Booms and busts, terms given to periods of abnormal price action, are influenced by policy decisions, demographics, and innovation. When an asset class returns outsized profits for an extended period of time, investors will often try to explain it as a “new era” or a shift in growth. But outsized returns are temporary. So-called “bubbles” are only truly evident AFTER a crisis causes prices to collapse. Factors like interest rates, taxes, demographics and inflation are central to asset class returns. During the stock market boom of the 1980’s and ’90’s in the US, inflation was waning, demographic trends were strong, tax rates were falling and interest rates were dropping. In addition, policy shifts created easy credit and lending conditions. It was a recipe for outsized returns. During that 20-year period, the S&P 500 index rose by 16% per year versus a 9.9% long-term average. In the 10 years following that historic run, US large stocks returned just 2.5% annually, bringing the average total return back in line with history.

A Fundamental Asset Valuation Model

A comparison of an asset class’ price movement to its normalized economic factor (earnings in equity assets and interest rates in debt assets), provides statistical evidence of a long-term relationship that can be used to model a going-forward return assumption. The model can be applied to numerous asset classes including: US large company stocks, US small company stocks, developed international stocks, emerging market stocks, investment grade bonds, foreign government bonds, high yield bonds, real estate and commodities.

Using history (over 100 years in some cases), we can observe the range of extreme valuations in each asset class. In the US large-cap company asset class, we use the S&P 500 index and the Dow Jones Industrials as proxies. Over the past 100 years, large US stocks have ranged from extremely cheap (reference 1940 and 1980 as examples) to extremely expensive (reference 1929 and 1999 as examples). When US large companies are cheap, they can produce annual returns of 14-16% or more for the following ten years. When they are expensive, they can result in very low returns of 2-4% annually.

When making capital allocation decisions, we compare each asset class relative to its own historic range and also compare the forward implied returns relative to other asset classes. The model’s value lies in a basic investing tenet: buy assets when they are cheap. This provides a powerful fundamental framework for asset class allocation decisions and answers two key questions. What can I expect my investment to return in the future, and how does that compare to other investment choices?

The power of the fundamental asset valuation model is in identifying long-term opportunities and allocating into above-average return expectations.

US large-cap stocks, (last 100 years) have an average annual return of 9.9%. Using a normalized earnings estimate whereby the cyclical ups and downs of the economy are adjusted, we can look at the long-term relationship between earnings and price, calculate the average of the relationship and compare it to long-term returns. By plotting the current asset price relative to a normalized, long-term earnings trend, we can solve for the current implied going-forward return and compare it to prior periods when the earnings and price relationship was similar.

When price and earnings are “normal”, we should expect to receive the long-term return of the asset class. We call this the “fair market value” (FMV). In the chart below, 50 years of US Large Company history is plotted along with the model’s FMV for the asset class.

CHART 1: Fair Market Value

As you can see, there are periods where US large-cap stocks have been historically “cheap” and other times when they have been expensive. Going back to the early 1960’s, the model’s expected going- forward 10-yr return was only 3%. Prices were high relative to their long-term relationship with corporate earnings. Over the next 10 years, US large-cap stocks averaged approximately 4% per year.

By 1980, US stock valuations had become cheap relative to their long term historic price to earnings relationship. After years of stagnant price movement and moderate underlying economic growth, valuations were low. The fundamental asset valuation model expected a going-forward 10-yr return of 14%. US stocks went on to yield an annual return of 16%.

Buying into stocks below the FMV line has resulted in greater than 10% annual returns while adding to stocks above FMV offers a below average result. In 2009, US large-cap stocks declined below FMV for the first time in 20 years and offered a rare value opportunity.

In the next chart, we have plotted the model’s predictive returns over the past 50 years. As you can see, the model suggested a long-term return of nearly 14% in 1982, which turned out to be 15 years of the best annual returns in the history of stock investing. Likewise, in 2000, the model suggested that stocks had reached extreme overvaluation levels and were offering less than a 2% annual going forward return.

CHART 2: High and Low Predictions for US Large Cap

Using similar comparative analysis for major growth and income asset classes, we can look at the relative value opportunities in a broader perspective and make decisions about how to best allocate capital. In Chart 2, the high and low extremes for US large-cap stocks were 14% and 2% and the current projection is for 6% returns. Using these data points for reference, refer to Chart 3 for a relative comparison of each major asset class and their respective ranges. The top and bottom of each box represent the high and low of historic models. The yellow diamond represents the current projected return. As you can see, each asset class has its own unique range and its own current projection.

CHART 3 High and Low Expectations for Asset Classes

This fundamental model provides a visual comparison of current investment opportunities. In the far right column, the cash yield is nearly zero and is at an extreme low. The Federal Reserve’s stated “zero interest rate policy” (ZIRP) designed to bolster economic growth through low cost borrowing is shown in this chart. In fact, ZIRP has impacted all US-based bond investments. After nearly 30 years of reducing interest rates to stimulate growth, rates have reached historic low levels.

In fixed income, only foreign bonds offer attractive yields relative to their historic range of returns. Our model suggests that US investment-grade bonds will return roughly 3.5% annual in the coming years.

Current expectations of the fundamental asset valuation model

The fundamental asset valuation model is signaling that a deep value opportunity is unfolding in parts of the developed international equity markets (western European countries). The going-forward return assumption suggests that over the next ten years, investing in the stocks of major European companies can outpace similar returns in US companies. Investors will need to have patience. Most undervalued opportunities are due to economic crisis or similar problems that need to be resolved. The current European debt crisis is no different.

The fundamental asset valuation model suggests that expected going-forward returns in fixed income allocations (investment-grade bonds) can be improved by investing in the 7-10 year maturity range and in foreign bonds. Expected going-forward returns in equity allocations can be improved by adding to international developed and emerging market economies.

Expected going-forward returns are lower than historic averages. This is mainly due to the impact of low income yields in many asset classes. In a balanced allocation, yield from US Treasuries and from US stock dividends are 2% versus the long-term historic income yield of 4.5%.

In order to achieve returns equal to historic average, investors will need to allocate to asset classes with higher yields (like high-quality international value stocks and foreign government bonds) and also seek out strategies like covered calls to generate higher yield without increasing risk.