[The data on gold and real estate returns in this post were updated in January of 2022.]

Over the past few months, I’ve been posting a series on historical returns of various asset classes including:

• U.S. large-cap stocks
• U.S. government treasury bonds
• Cash
• U.S. corporate bonds
• Global stocks
• Small-cap and Value stocks

For those new to the historical returns series or Mindfully Investing in general, historical returns are an important piece of the puzzle in developing a mindful investment portfolio and investing plan that meets your long-term investing goals.

Today’s post on gold and U.S. real estate marks the last in this series.  You might be as tired of reading these as I am of writing them.  But given the great interest I’ve seen in the topic of historical returns, I expect that this series will eventually be among the most popular and useful posts that I’ve published.  The historical returns posts are turning out like an encyclopedia.  Any given article may be very interesting to a particular reader, but few people will want to actually read every article in the “H volume” all at once.

Most people put gold and real estate in the category of so-called “alternative” investments, which also include commodities (gold being one), annuities, syndications, debt, notes, note funds, ATM funds, currencies, and derivatives to name a few.  Gold and real estate are by far the most common alternative investments.  And if you consider owning a home as an investment, then real estate investing is so common that it’s hardly an “alternative” at all.  All the more reason to take a closer look at the historical returns of real estate along with gold, which is nearly as popular.  Let’s start with gold.

Historical Returns of Gold

When looking at return histories, it’s best to take the long view.  Because gold has been used as money for thousands of years, you can theoretically extrapolate amazingly long histories of gold returns.  For example, Claude Erb estimated that gold has produced a nominal annualized return of between 0.5% and 2% over the last 3000 years!  While it’s fun to consider millennia of returns, it’s difficult to argue that the value of gold in ancient Egypt as compared to medieval England, for example, has much relevance to future changes in the value of gold over the next few decades.

It seems much more appropriate to examine gold returns since the end of the “gold standard” around 1972, particularly when looking at gold returns in U.S. dollar terms.  Prior to 1972, the U.S. government valued the dollar based on gold, which made gold prices in U.S. dollars essentially arbitrary.

Using Portfolio Visualizer data, the long-term nominal (not inflation-adjusted) annualized return (compound annual growth rate; CAGR) from 1972 to 2021 was:

• 7.6% for gold

As a general comparison, U.S. large-cap stocks returned 11.2% over the same period.  While a few percent difference in annualized returns may not sound like a lot, it makes an absolutely huge difference to an account balance when compounded over many years.

However, the 7.6% annualized return for gold is a long-term average, which means that over shorter periods gold returns diverged substantially from this average.  Here are some additional descriptive statistics for the nominal annual returns from gold going back to 1972.

You may be interested in determining annualized gold returns between specific years.  Similar to my historical return calculators for stocks, bonds, cash, corporate bonds, global stocks, small-cap, and value stocks, this calculator provides annualized gold returns (both nominal and inflation-adjusted) between any two dates back to 1972 based on the Portfolio Visualizer data.

Historical Returns of U.S. Real Estate

There are many sources of U.S. home price data. And obviously, the growth of home prices has varied widely over time, regions, cities, and even neighborhoods within the U.S.  So, aggregating all these data into one set of returns statistics that accurately portrays the entire history of U.S. real estate is unrealistic.

Further, I’ve written before that rental properties are clearly the best type of real estate investment, because they return value both from price changes and rental payments over time, increasing cash-on-cash returns. But rental payments are even more variable across time and place than housing prices, and rentals only represent a small portion of U.S. real estate investing. Indirect real estate options such as mortgage note investing, as well as real estate debt and note funds, are also excluded from this analysis.

Nonetheless, I can at least summarize the data compiled and used by world-renowned experts who have struggled with normalizing all these data into a meaningful “average” of historical returns based on housing price data.  Specifically, Nobel Laureate Robert Shiller has compiled statistics on U.S. home prices that have been widely used and accepted as generally representative of U.S. real estate returns in aggregate.  Shiller has compiled data going all the way back to 1890.  But given that the further we go back, the less applicable the data may be to current modern-day housing markets, I’ve curtailed this dataset to 1928, following the lead of Aswath Damodaran of the Stern School of Business.

So, according to Robert Shiller data, the long-term nominal (not inflation-adjusted) annualized return (compound annual growth rate; CAGR) from 1928 to 2021 was:

• 4.2% for U.S. Real Estate (based on price changes only)

In comparison, U.S. large-cap stocks returned 10.2% over the same period.  To compare more closely with gold returns summarized above, the U.S. nominal real estate return since 1972 was 5.3%, as compared to 7.5% for gold.  So, gold substantially outpaced real estate as an investment since the end of the gold standard in 1972.

However, as with all these annualized averages, over shorter periods real estate returns diverged substantially from 4.2%.  Here are some additional descriptive statistics for the nominal annual returns from U.S. real estate going back to 1928.

And here’s a calculator that will provide the annualized return (nominal and inflation-adjusted) for U.S. real estate between any two years.

Historical Risks for Gold and Real Estate

Because higher returns are usually associated with higher risks of losing money, it’s prudent to evaluate the long-term balance of both returns and risks for every investment.  Volatility, as measured by the standard deviation of the routine ups and downs of returns over time, is the most common (but somewhat flawed) measure of investment risk.

In past posts, I’ve gathered volatility and return data for a wide range of asset classes covering the last couple of decades.  But the volatility data for some asset classes span a fairly brief period.  So, for this post, I focused on asset classes that have volatility data going back at least to 1972, to match the period of available gold and real estate data.  Here’s a graph plotting risk versus returns since 1972 for multiple asset classes.

The dotted line in the graph represents the best fit relationship for the risk/return data.  For real estate, daily price change data don’t exist, so I estimated comparable volatility for real estate using annual returns data since 1972.

The dotted line suggests that additional return is indeed accompanied by additional risks for most assets.  Interestingly, both gold and U.S. real estate diverged substantially from this relationship.  And unfortunately, they diverge mostly in terms of lower than expected returns given the risks (volatility) involved, at least in this period.

In fact, cash has produced nearly as good a return as U.S. real estate but with substantially lower volatility.   Meanwhile, gold exhibited the highest volatility of all these assets but produced returns not much better than holding a relatively safe 10-year U.S. Treasury bond.

Conclusions

These data all suggest that gold and U.S. real estate are relatively poor investments as compared to most other assets, both in terms of returns and risk-adjusted returns.  But as I’ve noted before, the problem with these sorts of conclusions is that they’re highly dependent on the timeframe of the assessment.

For example, in the 10-year period from 2001 to 2010, U.S. large-cap stocks suffered two bear markets while gold outperformed just about everything.  Here are the annualized returns of gold and real estate as compared to U.S. large caps in this period:

• Gold – 17.5%
• U.S. Real estate – 2.6%
• U.S. Large-cap stocks – 1.3%

It’s notable that during the 2008 Great Financial Crisis, real estate suffered pretty badly in tandem with stocks.  This came as a great shock to many homeowners and real estate investors because in the few decades prior to 2008, real estate had acted as a decent safe haven during periods of stock market turmoil.

As I’ve been noting throughout this series on historical returns, return histories almost always show a cyclical ebb and flow of relative performance when you compare just about any two asset classes or subclasses over the long term.  So, we can say that gold and real estate have “usually” underperformed stocks, but that’s no guarantee of what might happen in the next decade or two.

1 – The arithmetic average of annual returns differs from annualized returns (CAGR) as discussed more here, and from COC return.