Pre-1958, Stocks Yielding More than Bond Yields was the Norm
Before 1958, the accepted wisdom was that the S&P 500 dividend yield (and earnings yield) should be higher than US treasury yields, because stocks were riskier than bonds and therefore should have the higher yield. Robert Shiller's data measured an average yield 1.8% above Treasuries between 1871 and 1958.
|Hat Tip: Safe Haven.com|
Post 1958, a “New Normal” of Stock Capital Gains, Not Dividend Yields
By the late 1950s, the memories of Great Depression of the 1930s—i.e., of destroying equity values while offering deflationary gains to sovereign bond holds—had finally become a dim enough distant memory that the idea of growth-stock investing producing capital gains began to take hold, i.e., capital growth in the value of stocks became more attractive than dividend income. The mantra became, “who needs yield when you got capital gains?
|Source: Data source:Shiller/econ.yaleedu.|
1982~2000 Unique Period in that Capital Gains Exceeded Dividend Gains
This secular shift had no apparent impact on the historical fact that long-term stock returns are driven more by dividends than capital gains until around 1982. The 1982 to 2000, however was marked by a dramatic fall in nominal bond and dividend yields. As the runaway inflation caused by two oil shocks subsided, capital gains provided more than twice the return of dividend yields as the market capitalization rate (P/E multiple) underwent a secular expansion. But as the above table indicates, this period was dramatically different from historical experience in that capital gains on stocks were greater than the return provided by dividends, whereas dividends have historically provided the bulk of total real stock returns.
New Normal Linked to Market Capitalization Rates, Not Dividend Yields
Thus the “new normal” post 1958 was codified into the difference between the earnings yield of a particular stock or asset and the long-term bond yield, or what is often called the “risk premium”. The valuation methodology used to gauge the relative attractiveness of stocks versus bonds became known as the Fed Model, i.e., when the forward earnings yield on the S&P 500 is less than the 10-year bond yield, stocks are too expensive, i.e., it does not pay to hold stocks because you can earn more in a Treasury bond with less risk. Conversely, when the forward earnings yield on the S&P 500 is above the 10-year bond yield, stocks have gotten relatively cheap to bonds.
This model however makes no consideration for earnings paid out to shareholders, or the dividend yield.
Japan Was Back to the “Old Normal” Well Before the 2008 Financial Crisis
PIMCO’s “new normal”, i.e., years of substandard growth, low stock returns and depressed bond yields, is what we would term the “old normal” in terms of the relationship between stocks and bonds that existed prior to 1958. For essentially all investors currently active in the market, the “old normal” is an unknown phenomenon they have never experienced—except in Japan. The Japanese stock market peaked in December 1989 at 38,900, and has been in the “mother of all bear markets” since. As Japan’s Heisei Malaise wore on, the Fed Model (i.e., the relationship between earnings yield and bond yield) began to break down, i.e., earnings yields above bond yields no longer signaled a rally was imminent. Further, the dividend yield on Japanese stocks began to trade higher than the yield on Japanese government bonds.
Prior to late 2007, dividend yield moving above 10-year JGB yield was the trigger for a strong rally in Japanese stocks, in 1998, 2003 and 2005. Sadly, the “magic” of stocks yielding more than risk-free bonds being a major buy signal worked in Japan right up until it didn’t, i.e., November 2007, when the free-falling Nikkei 225 just kept falling, pushing the dividend yield to multi-decade highs, even as JGB yields continued to drop. Aggregate yields have been trading only slightly above or significantly below JGB yields since, and JGB yields continue to trade at historically low yields.
"Old Normal” versus “New Normal”: The Primary Difference is Inflation
According to analysis by the late Peter Bernstein, the long-running positive correlation between bond yields and dividend yields defined the perceived fundamental relationship between these two variables, as the coefficient of correlation from 1970 to 1999 was a powerful +76.7%. However, the correlation between bond yields and dividend yields was negative from 1954 to 1969, at -58.8%.
As every investor is aware, dividends are variable over time and bond coupons are fixed from issuance to maturity. Thus inflation is one of the biggest risks to fixed-income securities, while dividends represent a long-term hedge against inflation. The correlation in returns therefore should therefore be just the opposite of the relationship prevailing post 1958 - negative during inflationary episodes and positive when the price level is relatively stable. According to Peter Bernstein’s work, the correlation between the two series has been high when inflation is high and volatile, while the correlation has been low when inflation is low and steady.
Determinants of Bond Yields
Most text books will describe the major determinants of bond yields as the following.
1. Interest rate risk premium: Investors in bonds will demand a premium (higher coupon) for longer-dated securities as compensation for holding securities for the long-term when their prices are very sensitive to interest rate changes.
2. Inflation premium: Because prices could rise in the future and erode the purchasing power of the fixed income stream from bonds, investors demand extra return for future inflation risk.
3. Default risk premium: Investors also demand extra compensation for the possibility of the issuer of the bond being unable to meet required interest or principal payments.
4. Liquidity premium: Is the amount of extra compensation investors demand for the difficulty of converting the bonds to cash at their true value.
The Risk-Free Nature of Sovereign Bonds
The U.S. treasury market is the deepest and largest financial market in the world, meaning large amounts of funds can enter and leave easily, with no liquidity premium. The U.S. government is also considered one of the best credits in the world, meaning there is no default premium. Further, the extraordinary monetary policy of the Fed and its aggressive use of its balance sheet has all but guaranteed against interest rate risk, meaning the interest rate risk premium in U.S. treasuries has also shrunk to an imperceptible level. Finally, the default scenario for economic growth coming off the worst financial crisis in over 80 years is weak economic growth, balance sheet restructuring and excessive levels of government debt as far as the eye can see, implying deflation, not inflation. The inflation premium for U.S. treasuries, historically seen as around 4%, has also been all but completely rung out of treasury yields, leaving only a safe-haven discount.
On the other hand, the sovereign bond yields of Spain and other Southern Euro nations reflect high, a) default risk premiums and b) liquidity premiums. Indeed, much of the balance sheet problems being experienced by Euroland banks is that holdings of Euro sovereign bonds are now considered “risky” assets with high default risk premiums rather than the “riskless” sovereign debts they were considered when first bought.
Ironically, while Japan is the most indebted country in the OECD, JGB yields were acting like US treasury yields long before US treasury yields hit 60-year lows, much for the same reasons, i.e., a) inflation risk was minus because of deflation, b) there was no liquidity premium as Japan's JGB marekt is also deep and large, c) equity risk-adverse domestic institutions continue to park the bulk of Japan's savings in JGBs, d) growth expectations for Japan's economy (and therefore potential price pressures) are essentially zero.
Determinants of Stock Prices
The textbook explanation of stock prices are that, a) the current price of a stock represents the present value of the expected flow of future dividends (earnings) plus realizable net asset value, with the value of future cash flows/dividends being discounted to present value based on the assumed “riskless” discount rate, which is usually the government bond yield. From another perspective, the stock price is equal to forward earnings x market capitalization rate (P/E multiple) + realizable net asset value.
In most models used to determine “fair” value, the major determinants of a stock's price include, a) dividend growth, b) risk premium, c) intermediate growth rate, and d) long-term growth rate. The risk premium in turn is determined by the “risk-less rate”, which is government bond yields, while the long-term growth rate is usually assumed to be the growth of the economy as a whole. Thus the biggest variable in stock prices is the expected growth rate, both for the individual company and the economy as a whole.
So What are Stock Yields Trading Above Government Bond Yields Telling Us?
In the case of the Nikkei 225, the earnings yield began to move above JGB yields as early as 2001, and has stayed above JGB yields since 2004. By this measure, Japanese stocks have been a screaming buy since 2004. Yet the benchmark Nikkei 225 has since fallen to a new historical low.
The dividend yield has also been above the JGB yield since late 2007. Why have these stocks relative to bond indicators broken down?
Basically, we believe it is because essentially all growth and inflation expectations have been squeezed out of both stock and bond prices over the last 20 years in Japan. Further, even though Japan’s public finances have deteriorated markedly during this period, Japan continues to be able to finance its massive public debt through domestic savings, meaning there still is no default or inflation premium reflected in historically low JGB yields.
With stocks in a semi-permanent (20-year) bear market, domestic institutional investors are giving up on generating enough returns from domestic equities to pay for a growing funding shortage in domestic public and private pension funds, and have been structurally unwinding the very strategic/cross holdings that skewed market valuations so horribly during the bubble years. In other words, the equity cult in Japan died a long time ago. Long live equities.
|Source: Nikkei Astra|
|Source: Nikkei Astra|
Thus what the Japanification of U.S. equities and treasury yields means is a structural breakdown in the bonds /equities allocation decision, semi-permanently in favor of bonds.
If the last 20 years in Japan’s financial markets are any guide, U.S. treasury yields (and market capitalization ratios) will continue to fall--even though 10-year treasury yields have already been falling for 30 years--and dividend yields will continue to rise—not so much because of a sharp increase in dividend payouts, but because of further declines in stock market indicaes as growth expectations are wrung out of individual as well as aggregate stock prices.