There has been a lot of hand-wringing going on about the Federal Reserve's latest, open ended quantitative easing program - or commitment to buy $40 billion of mortgage backed securities per month until high unemployment stages a "substantial improvement."
Some have argued that the Fed is stoking the fires of hyperinflation and advocates storing gold coins in Canadian vaults and stocking backyard bunkers with plenty of guns and gasoline. Others scoff at these naysayers, pointing out that the Fed has engaged in two previous rounds of QE with inflation not even batting an eye. So, who is right, and what are the implications for the S&P 500 and inflation hedging in our portfolios?
QE and Money Supply Growth: The first two rounds of QE meaningfully increased the money supply, with 1-year growth in M1 - or the Monetary Base plus Travelers Checks, Demand Deposits and other Checkable Deposits - peaking at record levels in both June 2009 and July 2011. However, while growth in M2, which augments M1 with retail money-market funds and savings deposits, has risen to multi-decade highs as well, it has lagged the growth of M1 significantly, one sign that money supply increases are not being transmitted effectively to the real economy. This is in stark contrast to the 1960s-70s, the mid 1990s, and the mid-2000s, when M2 growth was above M1 growth, stoking inflation (see chart below).
Indeed, another measure of "slack" in the economy - Capacity Utilization - remains well below its peak for the last cycle and may actually be starting another cyclical decline (see chart below).
It is this difficulty in transmitting money supply growth to the real economy, illustrated partially by the M1-M2 divergence, that caused the Fed to engage in increasingly aggressive steps to try to stimulate, as John Maynard Keynes would call them, "animal spirits."
Money Supply Growth and Inflation: But, will inflation stay dormant forever? Just because we have engaged in record or near-record levels of M1 and M2 growth since 2009 and seen no inflation (due to the popping of a massive asset bubble), will this continue to be the case in the future?
Milton Friedman would argue not. In 1972, Friedman published a paper demonstrating an almost two-year lag in the transmission of money supply growth to inflation. A 2002 paper by the British central bank (click here for reference) affirmed Friedman's general findings that it takes well over a year to transmit money supply growth to inflation.
The following chart graphs the current annual change in the M2 money supply with the annual change in the Consumer Price Index two years in the future (e.g. graphing CPI change against M2 lagged two years). While transmission has been less clear in the 1990s-2000s than in the 1960s-70s, it is still evident that changes in M2 growth lead to increases in inflation. Indeed, the correlation coefficient over the entire period is 0.47.
If we look at a longer-term chart of the relationship between money supply growth and CPI, we can see even more clearly how money supply growth acts as an anchor for CPI growth. The following chart graphs 5-year cumulative, rolling changes in M2 money supply and CPI. CPI growth is more volatile than money supply growth, reflecting the influence of expectations and other factors in the economy, though we can see how growth in CPI has returned over and over to closely track the growth in money supply after diverging. Indeed, over the last five years, CPI has grown a cumulative 11% (or 2.1% annualized), while M2 money supply has grown a cumulative 18% (or 3.4% annualized).
The implication is that, once animal spirits return, CPI would need to grow a good deal faster than ~2% to make up for recent monetary growth.
Inflation and Stocks: Contrary to popular belief, equities as an asset class are not a good inflation hedge. Andrew Ang, Marie Brière, and Ombretta Signori, in an excellent recent article in the Financial Analysts Journal, found that the average "inflation beta" of the S&P 500 is -0.52. Many individual stocks had high positive inflation betas - a desirable quality if you are hedging against rising inflation - but the betas were very unstable over time and the authors found it nearly impossible to build a reliable inflation hedge portfolio from individual stocks. Similarly, Antti Ilmanen devotes a section in his new e-book from CFA Research Institute - "Expected Returns on Major Asset Classes" - to the negative correlation between inflation and equity returns. Ed Easterling of Crestmont Research has also written extensively regarding the link between price instability (e.g. high inflation or acute deflation) and low cyclically adjusted P/E rates.
The following chart comes from Crestmont's presentation, "Financial Physics," on the links between real growth, inflation and P/E rates. Easterling's research demonstrates that cyclically adjusted price-to-earnings ratios for the S&P 500 are much lower during periods where inflation diverges meaningfully from price stability (e.g. ~1-3% inflation/year).
Conclusion: The bottom line is that secular bull markets generally begin when the price on the S&P 500 has sunk to around 5x-10x normalized earnings, far lower than the current 22.89x normalized earnings (see chart below), and that the key driver for P/E ratios to compress to those levels is either a period of acute deflation (where future nominal earnings growth is expected to decline) or acute inflation (where money supply has grown faster than the economy and causes an increase in prices).
The Fed's stimulus measures have been designed to ward off the specter of deflation at the possible expense of higher inflation. If high ongoing growth in M1 and M2 money supply begins to transmit strongly into inflation over the next 1-2 years - even in the absence of another deflationary shock - this could cause a decline in cyclically adjusted P/E ratios and, by extension, the S&P 500.
What to do? George Martin at Wood Creek Capital Management wrote an excellent article in the Journal of Alternative Investments (reproduced by his firm at this link), that found Treasury Inflation Protected Securities (GM:TIPS), commodities, timber and farmland to have the best inflation hedging capabilities. Within commodities, he found positive hedging properties for broad indices like the S&P GSCI or the Deutsche Bank Liquid Commodity Index - which can be tracked by index products such as the Powershares DB Commodity Index Tracking Fund (DBC), the iShares GSCI Commodity-Indexed Trust Fund (GSG) or the Harbor Commodity Real Return Strategy Fund (HACMX).
Martin also found that the commodities with the highest positive correlation to inflation between 1991 and 2008 were Aluminum, Copper, Crude Oil, Heating Oil, Soybean Oil and Unleaded Gas. Investors can target these commodities specifically through ETNs and ETFs such as JJU (Aluminum), JJC (Copper), USO (Crude Oil), UHN (Heating Oil) and UGA (Gasoline). The hedging properties of these products, though, can be impacted by contango in the futures curves, which can lead to underperformance of the commodity spot prices. Indeed, the WTI crude oil futures curve is currently in contango (longer-dated futures prices higher than shorter dated prices), while the curves for Brent crude, Heating Oil, Copper, and Gasoline are in backwardation. Interestingly, gold was found to be a better return diversifier and crisis hedge than it was an inflation hedge.
With respect to timber, there are four major REITs that own timber land - Weyerhaeuser (WY), Potlatch (PCH), Plum Creek (PCL), and Rayonier (RYN) - as well as the Guggenheim Timber ETF (CUT) that diversifies across several timber REITs. There is currently no REIT for farmland.
In addition to commodities, Antti Ilmanen has found that value equities, volatility selling and momentum strategies (e.g., buying recent winners and selling recent losers) have a positive correlation to the inflation risk factor.