[Note: This article was first posted on May 8 and had to be reposted due to website technical error.]
My last two posts (here and here) examined whether stocks are a good hedge against inflation. This is my third and last post in this series, at least for now. I reviewed many articles and research papers, and I found three common arguments against stocks as an inflation hedge:
- Stocks and inflation have a “strong” negative correlation, meaning that stocks do poorly when inflation rises.
- People confuse the long-term return from stocks with their ability to hedge inflation, which are two different things.
- Academic research established long ago that stocks are a poor inflation hedge.
In my last post, I tackled the first argument. After independently scrutinizing U.S. stock and inflation data, I found there’s a weak negative correlation between real (inflation adjusted) stock returns and concurrent inflation, and nominal stock returns are poorly correlated with inflation (nearly random). I also pointed out why the most appropriate comparison to inflation uses nominal stock returns not real returns.
That brings us to the second argument for why stocks are a poor inflation hedge: people confuse long-term stock returns with hedging. To avoid this problem, we need a clear definition of the term “hedge”.
What’s A Hedge Anyway?
We know that long-term stock returns have historically outperformed every other asset. But good long-term performance is no guarantee that stocks will consistently counteract shorter-term inflation changes. Unfortunately, media articles and even some academic studies use the term “hedge” broadly, which causes at least two points of confusion:
- What’s the relevant time frame for a hedge?
- What’s an effective hedge?
Here’s a basic initial definition of “hedge” from Investopedia:
- “A hedge is an investment to reduce the risk of adverse price movements in an asset…A perfect hedge is one that eliminates all risk in a position or portfolio. In other words, the hedge is 100% inversely correlated to the vulnerable asset. This is more an ideal than a reality on the ground, and even the hypothetical perfect hedge is not without cost.”
(The above definition focuses on “negative” or “inverse” correlations between investments. However, the perfect hedge against inflation would have a 100% positive correlation with inflation, where returns increase as inflation increases.)
Hedging Time Frame – Regarding the first point of confusion, the Investopedia definition does not address the likely time frame for a hedge. In my view, inflation hedging over the very long-term is illogical for most investors. For example, the annualized inflation rate since 1992 has been a low 2.3%. Hedges against inflation during this time would have generated consistently low relative returns, while the U.S. stock market performed well (the S&P 500 has an annualized return of 9.7% since 1992). Although often overlooked, the concept of hedging is most relevant over much shorter time frames than the decades-long horizons of most long-term investors.
The Effectiveness of A Hedge – Regarding the second point of confusion, Investopedia provides an unrealistic “perfect” example of a hedge, but no real-world example. Investopedia at least clarifies that we shouldn’t expect any real-world asset to act as a perfect hedge against anything. I liken this observation to one of Rick Ferri’s key tenants about asset diversification: the idea of consistently negatively correlated assets is a fantasy. Although for certain periods some assets may have a strong negative correlation, there are no sets of assets that increase in value and have consistent negative correlations with each other over long periods. For example, despite the fame of bonds as one of the best hedges against stock movements, as this graph from Ferri shows, the correlation between stocks and bonds is imperfect and has changed substantially over time.
Similarly, no one should expect stocks or any other asset to provide a perfect positive correlation with inflation. And to the extent that any correlation exists today, we should expect it to wax and wane, or possibly even reverse tomorrow.
An Adequate Real-World Hedge – Although we’re creeping closer, we still haven’t nailed down our second point of confusion about what’s an effective and realistic hedge. The clearest criterion for an adequate inflation hedge I found comes from a paper by Bekaert and Wang in 2010:
- “For existing securities to be good inflation hedges, their nominal returns must at the very least be positively correlated with inflation…Nevertheless, hedging may be difficult to accomplish in practice…”
In contrast, a Credit Suisse 2012 report on inflation hedges (based on the work of Elroy Dimson, Paul Marsh, and Mike Staunton) looks for a mere lack of correlation between inflation and real (inflation adjusted) stock returns. However, as I mentioned in my last post, the use of real returns in the Credit Suisse report biases their analysis toward a conclusion that all assets are poor inflation hedge. Consequently, I stick with the Bekaert and Wang definition for the rest of this post. I should also mention that Bekaert and Wang looked at inflation hedging over periods of 1 to 5 years. So, their approach fits with the shorter time frame that appears most relevant to our search for a real-world inflation hedge.
Academics Already Know That Stocks Are A Poor Inflation Hedge
This brings us to the third and last argument: academic research established long ago that stocks are a poor inflation hedge. First, a brief perusal of the internet reveals an ongoing academic debate on this subject. One example is from an article by well-known researcher Jeremy Siegel called “Stocks:The Best Inflation Hedge”. Siegel is often portrayed as perennially biased towards stocks. But here’s another example discussing the view of North Carolina State University Finance professor Richard Warr, where he says:
- “Most investors fail to appreciate stocks’ ability to hedge inflation because they confuse nominal and real earnings growth — a behavioral trait that economists refer to as ‘inflation illusion’.”
Second, if the inflation hedging ability of stocks is a dead and buried issue, it makes me wonder why researchers like Bekaert and Wang, Dimson et al., Warr, and others are still chipping away at this question as late as 2012. Most researchers like to discover new concepts rather than merely verify established theories. Nevertheless, we can explore this recent research to decide how settled the issue really is.
Using our definition from Bekaert and Wang, assets with returns that are positively correlated with inflation over shorter periods like 1 to 5 years represent an acceptable real-world inflation hedge. But even this fairly rigorous criterion fails to compare various investment assets to each other. A mindful outlook clarifies that many situations exist on a continuum between “good” and “bad”, or they can be “good” in one circumstance and “bad” in another. Therefore, we should also compare the relative inflation hedging ability of one asset to another.
I found many old and new studies using a range of methods to compare inflation hedging of various assets, sometimes with scant explanation of the methods employed. The Bekaert and Wang study is relatively recent (2010) and is one of the most well explained, comprehensive, and detailed studies I found. Also, because the conclusions of the Bekaert and Wang appear to be broadly representative of most of the other recent studies I found, I use their results below to compare the inflation hedging ability of several investment assets.
Bekaert and Wang calculate a so called “inflation beta”, which is one measure of the correlation exhibited in a time series regression. Beta is different from the R-squared values I’ve used in earlier posts. A study by Hewitt Ennis Knupp provides a good explanation of how to interpret inflation beta values:
- “Put simply, when inflation rises by 1 percentage point, an asset with an inflation beta of 1.0 will see its nominal return rise by 1 percentage point as well. Such an asset would be free of inflation risk, as its returns would rise and fall with inflation.”
Further, a negative inflation beta means that nominal returns decrease as inflation increases. Using the Bekaert and Wang hedge definition, assets with negative betas are poor inflation hedges.
Total Inflation Hedge – The following graph summarizes one-year inflation betas calculated in the Bekaert and Wang study for several commonly available investment options across 45 countries. (Their study also calculates betas for periods up to 5 years, and although the longer period results vary somewhat from the one-year results, the overall conclusions are largely the same.) The data used in their analysis span time frames between 1970 to 2010, although for some regions and countries, the available time period is considerably shorter. Each world region is represented by a dot. The error bars (based on standard errors) roughly estimate of how much each beta value could reasonably vary due to noise in the data.
The graph is based on so called “total inflation” for reasons that will become apparent when I talk about “unexpected inflation” below. The first thing that jumps out is the large spread of regional results and the associated errors around the estimates. Across all the regions examined, stock inflation betas could reasonably range from 1 (a great inflation hedge) to -2.5 (a poor inflation hedge). Nonetheless, the center of this spread for stocks is somewhat negative, as approximated by the North American inflation beta of -0.42. It’s also notable that even relatively large and stable developed stock markets can have positive inflation betas, as shown by the European Union (EU) inflation beta of 0.27. Further, Bekaert and Wang point out:
- “The inflation beta in the United States is indeed negative but it is not statistically significantly below 1. In fact, the coefficient became less negative by adding the recent crisis years, in which low stock returns and below average inflation went hand in hand.”
Recognizing all the uncertainty involved, stocks have roughly the worst range of inflation hedging betas among these assets.
The finding that bond betas are slightly better (more positive) than stock betas is somewhat surprising. In contrast, bonds came out as worse hedges than stocks in the Credit Suisse report as well as a report summarized by Early Retirement Now. Apparently, differences in data sets, time periods, and methods can yield substantially different answers. This is one clue that the relationships between asset returns and inflation are not as stable and pervasive as sometimes implied.
Unexpected Inflation Hedge – This next graph focuses on so called “unexpected inflation” betas from Bekaert and Wang.
Unexpected inflation is the part of inflation that’s not already “priced into” assets as part of normal market dynamics. Researchers have long posited that an inflation hedge could be more useful if it reacts particularly well to “surprise” changes in inflation. In earlier articles, I’ve referred to these inflation surprises as the “true risk” of inflation. Nonetheless, extreme attention to only the “unexpected inflation” results could be misleading. Bekaert and Wang point to a classic 1977 paper by Fama and Schwert, where they used both total and “unexpected inflation” betas to find the best inflation hedging assets.
In my view, the real value of my investments are determined by the total rate of inflation, not just the “unexpected” part of it. For me, an ideal hedge would address total inflation, which is composed of both routine inflation and “unexpected inflation” not just one or the other. When total inflation is high, I’d prefer to have access to higher returns, regardless of whether everyone originally expected inflation to rise substantially or not.
I’m even more suspicious of the “unexpected inflation” betas here, because the authors had insufficient data to calculate “unexpected inflation” for all these different regions and time periods. Instead, they assumed that markets generally expect existing rates of inflation to continue into the future. They defend this assumption at some length, but regardless, the “unexpected inflation” in this case is really just the change in inflation over the period analyzed. For example, if inflation was 2% at the start of the period and 2.5% at the end, they assumed that “unexpected inflation” was 0.5%.
Setting my suspicions aside, the “unexpected inflation” betas better differentiate the hedging ability of the assets, although there are still very large and overlapping uncertainties among the assets. Again, stocks fare the worst, while gold looks relatively compelling as a hedge against “unexpected inflation”, particularly in North America.
And the Winner Is…
The Bekaert and Wang study concludes that broad swaths of stocks are often a poor inflation hedge relative to other assets, which is generally consistent with most of the other academic papers I found. But as I noted above, the ranges of these outcomes is highly overlapping. For example,
Media articles too often fail to explain that these type of research findings aren’t certainties, but literal probabilities that are sometimes statistically insignificant. In most scientific fields, the insignificant negative correlation between U.S. stocks and inflation would not be actionable. In fact, if you point to any one colored data point in the total inflation beta graph, in almost no cases will that data point lie outside the range of beta uncertainties for the other assets. This is largely true for the “unexpected inflation” graph as well.
While you can say that gold is the “best” inflation hedge and stocks are the “worst”, such a conclusion completely ignores the uncertainties of the data including the variability over time and place. While media articles can elevate these conclusions to dogma, a close look at any of these research papers yields important qualifications like:
- Credit Suisse 2012 – “We show that such standard securities [stocks and bonds] are poor inflation hedges. When we expand the menu of assets to Treasury bills, foreign bonds,
real estate and gold, matters improve but mostly only marginally.” - Bekaert and Wang 2010 – “…[I]t is difficult to protect a portfolio against unexpected inflation in the long term as well using traditional asset classes.”
- Hewitt EnnisKnupp 2012 – “..[T]raditionally recognized “inflation-hedging” assets offer…limited benefits.
Further, the Bekaert and Wang study attempted to devise ideal inflation hedging portfolios by combining various sets of assets, but they couldn’t generate any portfolios that delivered a positive correlation with inflation. They go so far as to say:
- “In short, it is next to impossible to use an individual asset or a portfolio of assets to adequately hedge inflation risk.”
So, stocks are a “poor” inflation hedge by this measure, but so is pretty much everything else.
Balancing Inflation Risks and Returns
None of my discussion in this series of posts confuses assets as an inflation hedge with holding assets to achieve adequate long-term returns. But here’s the problem. No one invests in anything only because it’s a hedge. The bedrock goal of all investing is to grow the real value of your assets over the long term. If you don’t achieve that goal, then your portfolio is an unqualified failure. Although we don’t want to confuse hedging with long-term returns, ignoring those returns would result in some pretty stupid decisions.
For example, the Bekaert and Wang inflation betas suggest that an all-gold portfolio would be the best of a bad lot to hedge inflation. The Credit Suisse report notes that an all-gold portfolio has provided a 1% annualized real return since 1900. In contrast, stocks provided an annualized real return of 5.4% in this period. Using these annualized rates, if you invested $10,000 in each asset for a 40 year investing period, you’d have an inflation-adjusted $82,000 in stocks but only $15,000 in gold.
As many of these researchers point out, balancing inflation “risk” with returns is actually the very same problem as trying to balance volatility risk with returns. The more you obsess with reducing portfolio volatility, the more your long-term returns and investment goals will suffer. When an investor balances volatility and returns by selecting a 60%/40% stock/bond portfolio, no one accuses them of confusing volatility for returns. This observation reveals that the argument about confusing inflation hedging and long-term returns is a bit of a red herring. Accordingly, a mindful investor must prudently balance short-term inflation impacts with long-term returns.
You could take an all-stock portfolio and hedge it by converting 30% of the portfolio to gold. Using the annualized real returns for gold and stocks noted above, such a portfolio would have 4.1% in annualized real return as opposed to the 5.4% return for the all stock portfolio. Reducing your short-term inflation “risk” through hedging has a tangible cost in the form of substantially lower long-term returns. In this example that cost is 1.3% less in total annualized returns.
In earlier articles, the mindful conclusion was that short-term volatility is mostly just an issue of emotions, and the true investing risk is the potential for a long-term permanent loss. As a result, most investors benefit from a stock-heavy portfolio. Likewise, tailoring your portfolio to decrease the impact of short-term inflation gyrations may make you “feel” better, but it will cost you some of your long-term returns. A stock-heavy portfolio may not technically be the best hedge against inflation, but it’s still the best bet for most long-term investors.
TIPS to the Rescue
While it sounds like hedging for inflation is nearly impossible, many researchers including Bekaert and Wang have observed that Treasury Inflation Protected Securities (TIPS bonds) provide a good inflation hedge, which agrees with past Mindfully Investing posts. TIPS provide explicit inflation hedging by adjusting the principal and interest rates of a regular U.S. Treasury bond by the annual inflation rate, measured by the Consumer Price Index (CPI). TIPS will typically outperform similar duration Treasury bonds when inflation is positive, and under perform Treasuries during deflationary periods. Like other researchers, Bekaert and Wang found that TIPs, and similar bonds available in some countries, are probably the only predictable inflation hedge readily available. Increased return during increased inflation is literally built into these investment instruments. However, because TIPS also under perform stocks, adding TIPS to a stock portfolio is still going to cost you some long-term returns in most cases.
Conclusions
We need to dispense with the hunt for the unicorn called a “great” inflation hedge. The best we will likely ever do is a “decent” inflation hedge. If a “great” inflation hedge really existed, everyone would use it, which in turn, would probably cause such a hedge to stop working. This is the reason I’ve been using phrases like “decent hedge”, “sporadic hedge”, or “imperfect hedge”, when describing how stocks can hedge against inflation.
While detailed research shows that stocks have a weak negative correlation with inflation, this relationship is not pervasive over place and time, and most researchers recognize the large amount of noise in the data. This noise means that while stock returns may sometimes decline as inflation rises, at many other times stock returns hold steady or even rise with inflation. This noise is one reason that my simple correlation analyses between U.S. nominal stock returns and inflation in my last post showed that stocks move almost randomly relative to inflation changes. The mindful perspective is still that stocks are an “imperfect” hedge against inflation. Better inflation hedges (like gold) probably exist, but they come with the cost of sacrificing substantial long-term returns just to make you feel better about short-term inflation gyrations.
To get past the whole “hedging” argument, I propose we instead say that stocks provide a type of “inflation diversification”. No reasonable investor expects zero volatility from even the best portfolio diversification. By the same token, no reasonable investor should expect complete immunity to inflation either. Because nominal U.S. stock returns have often reacted almost randomly to inflation, stocks are an excellent way to diversify your portfolio for inflation. Just like with asset diversification, your stock returns are unlikely to consistently increase when inflation rises, but those returns won’t likely be entirely driven by inflation changes either.
Going further, the Credit Suisse report points out that if you add international stocks to your portfolio, you are also getting currency diversification. Because not all currencies inflate and deflate in lock step, international stock exposure can push the inflation beta for your stock portfolio in a positive direction. This is very consistent with the Mindfully Investing recommendation for stock diversification: most people should hold a moderate number of low-cost stock index funds including some foreign funds.
Here’s a simple example. Using the Bekaert and Wang total inflation betas, the weighted-average betas for two different stock portfolios are:
- Portfolio of 100% North America Stocks: -0.42
- Portfolio of 60% North America, 25% Foreign Developed, and 15% Emerging: -0.16
Let’s be clear that these weighted averages are not the exact betas for these portfolios, which would need a more elaborate calculation. Nonetheless, they give a rough estimate of how stock diversification can moderate negative inflation betas. The research discussed at Early Retirement Now suggests that you might be able to achieve even greater inflation diversification by emphasizing certain stock sectors in your portfolio. That study calculated certain stock sectors in the U.S. as having positive unexpected inflation betas in the 4 to 8 range, although their methods for arriving at these results are cited as “proprietary” and are not well explained.
One last note of thanks to Early Retirement Now and Actuary on FIRE, who are doing their own series together on inflation and investing right now. Their posts have been very valuable in my research and helped me work through this complex subject. Please check out their excellent posts on inflation!
