As a scientist, I view money management and portfolio construction from the same perspective that many "quants" do in that it's all about the numbers. Take the 60% stocks/40% bonds example from 1975 to today:
The data doesn't lie: growing $10,000 to nearly $600,000 (CAGR = 10.45%) is pretty good. Sure you sacrifice the returns of a 100% stock portfolio but you blow bonds out of the water with less "risk". However, if we were to consider risk as a permanent impairment of capital as opposed to the common misconception of it as volatility, the 60/40 portfolio suddenly becomes inadequate. For me, the most concerning aspect of this commonly touted "balanced" portfolio lies in the lack of diversification. Surely there must be some value for retail investors in including other assets in a portfolio.
At this point, my scientific training kicked in and led me to conduct research into different theories surrounding portfolio construction. Two primary sources influenced the way I will construct my portfolio for you today: Meb Faber's incredible portfolio analysis book, Global Asset Allocation, and Nicholas Nassim Taleb’s Incerto series and Antifragile, where he posits the idea of a “Barbell” organization when allocating assets. Based on these sources, I wanted to explore the idea of creating a barbell or pseudo-barbell portfolio embracing non-financial asset classes to become adequately diversified.
Key Portfolio Contents
In my view, portfolio construction seems to revolve around the two important concepts of wealth preservation and wealth growth. This does not preclude wealth preservation assets from growing in value but the idea is to distinguish between assets that have the potential for appreciation with limited downside risk and assets with appreciation potential and substantial downside risk. Here is a non-exhaustive list of several asset classes divided into those two key concepts:
|Wealth Preservation||Wealth Growth|
|Hard Assets||Precious metals, real estate||Commodities|
|Financial Assets||High-quality credit (AAA), Guaranteed Investment Certificates/Certificates of Deposit||High-yield credit, stocks, REITs|
Using the above asset classes in an attempt to construct and modify a barbell portfolio (90% T-Bill + 10% US Micro Cap), I came across an interesting combination that embodies the barbell idea of maintaining high returns and greatly minimizing drawdowns:
- 60% wealth preservation assets - 30% gold + 30% 10Y treasuries
- 20% US Micro Cap stocks
- 20% REITs
The asset mix may seem strange but I will briefly justify it below starting with its performance compared to two barbell portfolios and the 60/40:
Although not immediately clear from the graph, each portfolio mix has the following characteristics with a $10,000 starting balance:
|Portfolio||Final Balance||CAGR||Worst Year||Best Year|
|Beating the 60/40||$618,398.26||10.58%||-7.73%||56.06%|
|Barbell Portfolio w/ T-Bills||$107,151.80||5.96%||-2.52%||14.24%|
|Barbell Portfolio w/ 10Y||$317,009.31||8.80%||-6.56%||38.40%|
As you can see, the barbell portfolios do not do a very good job at providing returns but are best for preserving capital. Even taking a more liberal interpretation of the barbell and substituting T-Bills with 10Y treasuries does not allow for it to compete with the 60/40 in the long run. However, the seemingly odd asset mix I showed above seems to provide the best of both worlds with drawdowns of less than 10% while maintaining significant upside potential greater than the 60/40. With that being said, I will now explain the reasons I chose the asset mix of the 'Beating the 60/40' portfolio.
Asset #1 - 30% Gold
Gold is largely seen as a "barbarous relic" or "shiny doorstop" and is derided by Warren Buffett. However, I believe it is negligent to exclude it from a portfolio for the following reasons:
- Hedge against central bank mistakes: as Ben Hunt astutely pointed out on Bloomberg, gold acts as a hedge against central bank errors such as hiking rates too quickly or having no ammunition to fight the next economic downturn.
- Hedge against financial mistakes/portfolio stabilizer: a major function of gold is to preserve wealth over time and act as insurance against financial calamities such as massive inflation or deflation. Given the startling devaluation of fiat currency and unsustainable government debt loads, gold deserves consideration as a wealth preservation device.
Courtesy of Tyler Durden & Bloomberg
- No counterparty risk: as JP Morgan once said, 'Gold is money, everything else is credit'. No matter the amount of fiat currency in circulation, the size/weight/purity of gold remains constant. Moreover, gold is a tangible element and its inherent value is not reliant on any other person/institution, making it one of the very few assets with no counterparty risk.
A good example of using gold in a portfolio to avoid counterparty risk while maintaining liquidity is in the portfolio of Edelweiss Holdings as of January 2018:
Despite having 35% of the portfolio in the preferred currency of miners and pirates, Edelweiss Holdings has performed exceptionally well compared to MSCI World Index:
That performance equates to a CAGR of approximately 18.19% compared to Berkshire Hathaway's CAGR of 8.92% since 2001. Maybe gold is not such a bad idea after all!
Asset #2 - 30% 10Y US Treasuries
At first glance, it would appear that a barbell portfolio would shun 10Y treasuries as they expose the investor to interest rate risk and sit in the middle of the yield curve. However, in a 2014 study by Thomas Flavin and colleagues in the Journal of International Financial Markets, Institutions, and Money, 10Y treasuries performed better than 1Y treasuries during periods of financial shocks. With that in mind, take a look at the 'Beating the 60/40' portfolio with T-bills:
As you can see, substituting 10Y treasuries with T-bills greatly impacts performance. If desired, shorter-term treasuries/T-bills could be used as they have less drawdowns than 10Y treasuries over time but you would need to accept lower portfolio returns. It is interesting to note the approximate growth in wealth is nearly identical around 2010-2011 for the portfolio with T-Bills and the traditional 60/40.
Asset #3 - 20% US Micro Cap Stocks
With the safer assets allocated to prevent major losses in the portfolio, adding risk assets is required to enhance portfolio returns. Picking and choosing between various large and mid-cap names is redundant and will only act as a drag on future returns. Take a look at US micro cap stocks (yellow line) in comparison to the total US stock market (blue line) and the global stock market ex-US (red line) since 1995 when global market ex-US data is available:
Allocating 20% of the portfolio to this asset class definitely opens up the possibility of losing substantial capital; 56.61% in one go to be precise. However, David Desjardins recently penned a column in Reuters I highly recommend to understand why micro caps are worth the risk:
In conclusion, success in micro-cap stocks is often about sacrificing first order and linear effects for the power of higher order or non-linear movements. The goal should be to profit from extreme outcomes while focusing on minimizing roll losses. A new generation of value investors will focus on finding value in higher order movements where the markets are less efficient.
In layman's terms: micro cap stocks are where you can find an edge and greatly increase potential returns. With a dearth of analyst coverage and media attention, these stocks may be more difficult to thoroughly research compared to large-cap names with big investor relations departments but it can be done. This is a real advantage as it is unlikely you will out-analyze somebody at a large financial firm whose job is to examine large-cap names on a daily basis.
Asset #4 - 20% Real Estate Investment Trusts (REITs)
In Meb Faber's portfolio construction book, Global Asset Allocation, he profiles many of the greatest investors' portfolios and comes to the conclusion that diversification is essential. In particular, adding real assets to a portfolio greatly stabilizes returns over time. This is evidenced by the approximate portfolio of Marc Faber (author of Gloom, Boom, Doom report) containing 50% real assets including 25% in real estate. His portfolio had positive real returns every decade since the 1970s, giving a good indication that real estate (and gold) should be key parts of a solid portfolio.
In the context of portfolio construction, REITs may act as the closest proxy for real estate. I recognize that REITs are financial constructions based on real estate and can have substantial volatility during financial shocks but, as they are based on real assets, they will suffice. Infrastructure assets can also be included in this category especially given the urgent need for rebuilding many of our crumbling bridges, water infrastructure, etc.
Conclusion - From Barbell to Diversified Portfolio
The 'Beating the 60/40' portfolio, to provide for a balance of steady growth and protection from ruin, consists of 60% wealth preservation assets (30% 10Y treasuries + 30% gold). The remaining assets sit in US micro cap stocks and REITs as a representation of risk assets. At first glance, this portfolio seems anathema to anyone looking to grow wealth while investing as there is 60% allocated to "safe" assets. As the data above demonstrates, it is possible to balance protection of capital with growth while diversifying away from only stocks and bonds.
From a macroscopic asset class allocation perspective, it looks like a true barbell portfolio of 90% safe assets/10% risk assets is not realistic if any return is expected. While the barbell performs well in drawdowns, it underperforms once risk assets recover. This raises the possibility of using a pseudo-barbell strategy when the market is extremely overvalued and then shifting more assets back to risk-on after a major drawdown to preserve and greater compound existing capital. Furthermore, the true barbell portfolio Taleb refers to in his books involves complex derivatives providing payoffs exceeding those of the US micro-cap index shown above, meaning it is difficult, if not impossible, for a retail investor to fully implement.
Taking a critical look at the 'Beating the 60/40', I can observe some potential issues not evident during backtesting. For one, the large allocation to 10Y treasuries sit in the middle of the yield curve and are susceptible to rising interest rates more than short-term treasuries or T-Bills. While I will not go into the probability of interest rate changes, it is something worth considering in the future if you believe the 30-year bull market in bonds is truly over and 10Y treasuries could suffer. In theory, you could experiment with different AAA bond combinations at various durations and yields to 'optimize' the portfolio but would, in my view, unnecessarily add complexity.
I am also acutely aware of the irony of trying to move away from the 60/40 stock/bond split only to invert the risky/safe idea and move back to a 60/40 asset split. On a conceptual level, this suggests that an individual can position themselves in a conservative manner while enabling strong growth if they want to put in the effort to research micro-cap stocks and real estate/infrastructure investments.
Disclosure: I am/we are long GLD. I wrote this article myself, and it expresses my own opinions. I am not receiving compensation for it (other than from Seeking Alpha). I have no business relationship with any company whose stock is mentioned in this article.