I try more and more to understand ideas from the ground up, but with a recent increase in partisanship and decrease in nuance, I decided I would read one study or academic journal a month for 2018. This would not only help increase my knowledge on various topics, but do so a way that came directly from the source and not an interpretation of the source. A less stringent but still worthwhile part of this initiative is to read across topics, though I’m willing to consider ‘economic rates of return’ and ‘minimum wage’ as separate topics (since I read a lot about economics). What follows are the studies and notes from each reading (updated monthly).
The Rate of Return on Everything
Jorda, Knoll, Juvshinov, et al // January // Topic: Economic Rates of Return
I stumbled upon this from multiple sources, with the first being Tyler Cowen, and was immediately intrigued because I’m often drawn to high-level or historical points-of-view — I find it helps me place today inside of the larger human story. As with any study, it’s never the final word, but this one gives a good overview of the rates of return on various assets, which is important because:
- There aren’t many studies or resources that collect and combine centuries-worth of economic data (the other famous one is Piketty’s). Regardless of their interpretations, this is a valuable dataset for others to use in their research.
- Conclusions drawn from these historical rates or return informs a number of existing topics that are currently being widely discussed, including economic inequality, secular stagnation in the aftermath of the global financial crises, and the addition of housing data in understanding the rate of return on different asset classes.
- Global vs USA: since my world is generally U.S.-centric, it’s important to note the U.S. sometimes differs as compared to other countries and to the aggregate global numbers. Specifically in regards to economic inequality (r - g), this study does not include data from places like China or India, which has had significant affects on global income inequality (specifically, reducing it).
- Time Horizons: economic data often gets grouped into pre-world wars, interwar, and post-wars time periods, as these three different periods can offer a more nuanced view than simply aggregating and averaging numbers over 150 years. Studies, including this one, will at times also segment time periods around the 1980 mark (1950–1980 and 1980-now), as that is often when the U.S. and global changes started affecting economic indicators (for instance, productivity slow downs happened in the U.S. starting in the late 1970s).
- “Real rate of return” is the return on investment minus all of the costs (e.g. transaction costs, inflation-adjusted, etc.) AND this study takes into consideration volatility, which means that while equities may have a slightly higher rate of return as compared to housing, housing outperforms equity over a longer period as there’s less of a chance you have a negative return in any given year with housing (“although aggregate returns on equities exceed aggregate returns on housing for certain countries and time periods, equities do not outperform housing in simple risk-adjusted terms.” p. 23).
- Generally, “housing” does not refer to a single homeowner, but housing as an investment vehicle.
- One should note that these findings cannot (and maybe should not) be easily translated into personal financial decisions. This is more a method to better understand the world.
Summary of Findings
The study breaks their findings into four groups: risky returns, safe returns, risk premiums, and return vs growth.
Risky assets, specifically equity and housing investments, have a similar rate of annual return, averaging 7%, though housing is significantly less volatile. Housing outperformed equity pre-WW2, and equity outperformed housing post-WW2, but at a significantly higher volatility rate as it has “experienced many pronounced global boom-bust cycles” (as compared to housing) — ranging from 16% to -4%. Given a longer time horizon, compound returns in housing allowed it to outperform equities. Unsurprisingly, housing does not appear to be correlated across countries like equities, partially due to housing being less globally tradable as well as regional shocks.
On the other hand, safe assets, which consists of treasury bills and government bonds, have had significantly lower returns than risky assets, as expected, on average about 1–3%. However, the volatility depends a lot on the aforementioned periods, having a low rate of return during peacetime period, a negative rate or return during the two world wars, and a high rate of return during the interwar period and external shocks (e.g. 80s inflation crises). Basically, there’s low risk in safe assets, except during periods you may want to rely on that low riskiness (e.g. wars, inflationary periods, etc.), and so safe asset returns are more volatile than you’d think.
The risk premium is the real rate of return on risky assets minus safe assets (in essence, the premium you pay — or the return you don’t receive — for choosing a safe asset). Long-term the risk premium averages 5%, but has also been volatile, largely driven by the volatility of the safe rate of return (as opposed to changes in the risky rate of return, which is relatively stable, and what I would’ve thought to be the more volatile asset). An important note from the authors stated, “risk premiums stayed curiously and persistently high from the 1950s and 1970s, persisting long after the conclusion of WW2. However there is no visible long-run trend, and mean reversion appears strong.”
Over the long run, return vs growth (r vs g) seems to show that the return on wealth (r) is greater than the growth of GDP (so, r > g). Though, breaking it down into periods shows that the world wars swapped the dynamic so that r < g. There also seems to be no evidence in the data that there’s a correlation between r - g and g. Despite this not surfacing in this study, it is notable that while the U.S. has seen in increase in income inequality since the 1970s, global income inequality is decreasing (note: wealth inequality is a different story) — largely due to China.
Conclusions - aka how does this relate to society
Housing is an important ingredient in understanding rates of return, and more specifically, housing investments have had roughly the same rate or return as equity, but with significantly less volatility (standard deviation of returns for housing is 10%, vs 22% for equities), which means that housing has been a better investment if you consider compound returns (6.6% vs 4.6%).
Economic inequality has been a topic of much discussion over the last 10–15 years, often for good reason. The data in this study seems to support Thomas Piketty’s theory that r > g, which can lead to increases in economic inequality (again, wealth vs income is an important nuance). On average, the rate of return on wealth (r) has been roughly 6% while GDP growth (g) has been 3%, and r has exceeded g in 13 of the last 15 decades (the two being both World Wars). However, this study also sheds more light on the theory by differentiating rates of return by asset classes. For instance, wealthy households seem to earn a higher return on their portfolios, potentially due to differences of assets in their portfolio, which would add nuance to Piketty’s theory. A final point that is of particular interest to me (having read a lot about economic inequality and technology/innovation): the stock of wealth has increased over the last few decades while keeping it’s high rate of return (e.g. no diminishing returns to scale), and a potential conclusion to be drawn from this is that “the elasticity of substitution between capital and labour may be high” (p48). In layman’s terms, society could continue to substitute labor for capital (like technology, among other things), without decreasing its rate of return on investment. A lot of potential policy implications from this.
Secular stagnation is associated with low rates of return, driven by an excess of savings or a general unwillingness to borrow and invest (a major theory resurfaced over the last 10 years by Larry Summers). While this may still be a potential theory, this study’s data puts it into question once we split “rates of return” into our two categories: risky and safe — and as we know, while safe returns have fallen recently, risky returns remain persistently high.
Investment opportunities, while less interesting to me, are still relevant conclusions to be drawn from this study. For instance, while difficult to achieve, diversifying a housing portfolio based on region appears to be fertile ground. Continuing with housing, “the low covariance of equity and housing returns over the long run reveals attractive gains from diversification across these two assets classes that economists, up to now, have been unable to measure or analyze.” Historically, safe and risky returns have been correlated, but more recently, this has weakened, and “turned negative from the 1990s onwards…[suggesting] that safe assets have acted as a better hedge for risk during both the Great Moderation, and the recent Global Financial Crises.” p44.
Banking crises “appear to be relatively more frequent when risk premiums are low. This find speaks to the recent literature on the mispricing of risk around financial crises.” p42.
Government Financing. The flip side of low or negative rates of return on safe investments have allowed the government cheap or free financing, since safe investments include treasury bills and government bonds (and the money to buy these goes to the government). Similarly and more specifically, if the rate of return on safe assets is less than GDP growth, then a reduction in debt-to-GDP is possible even with the government running modest deficits (the flip side requires the government to have budget surpluses). Historically, this has been true, as GDP has been 3% while the safe asset return rate has been 2%. All of this has been particularly important in helping to finance both World Wars.