Risk Mitigation with a Barbell Portfolio

It is extremely common for investors today, even those with professional financial advisers, to follow a version of buy-and-hold indexing portfolios with standard ETF or mutual fund products.

Invest 60 percent (or more) of your money in diversified equities and 40 percent (or less) in bonds to achieve a target return with a reasonable amount of risk. Re-balance annually. Hold for life. These types of portfolios have provided investors with 8-10 percent annual returns averaged over the past decades.

Many investors use period backtesting to justify further spins on this basic model. Go all U.S. stocks, allocate an outsized portion to small-cap value stocks or emerging market stocks. Use equal weight indices instead of cap-weighted indices. Or pursue other niche equity factors to try eek out small extra returns.

With most of the focus on the equity side, bonds are often overlooked. Generally investors will choose a broad bond fund, which includes Treasury bonds and investment grade corporate bonds. Others choose just Treasury bonds. Again, there are minor adjustments to this standard: adding international bonds, tilting to long-duration Treasury bonds, and so on.

Much of this standard, run-of-the-mill advice is based on lengthy backtests. While investors quote long-term returns of 8-10 percent annually, very few years actually would have seen those return figures.

The chart below shows what investors believe the returns look like in this type of investing model. An (almost) straight line up and to the right—nothing wrong with that.

Sources: TheRichMoose.com, S&P, FRED-Federal Reserve St. Louis

The danger of looking at long-term logarithmic charts is that the true experience of the investor is cloaked by the big picture view. Hindsight returns over long periods of time are great, but real investors live in the moment. They panic when their portfolios lose money and get overconfident when things are going well.

Instead of steadily higher returns as the long-term chart above suggests, investors experience many drawdowns in the 5-10 percent range and several that are much deeper still. Being an investor, including a buy-and-hold indexer means being in some level of drawdown much of the time.

Here is the same 60/40 portfolio plotting the drawdowns during this time period.

Sources: TheRichMoose.com, S&P, FRED-Federal Reserve St. Louis

The drawdowns of a 60/40 portfolio are severe enough to shake the confidence of most investors. The intensity only increases when investors slide up to 80 or 90 percent equities for better returns.

The Danger of Overlooking Bonds

Many investors have a profound misunderstanding of bonds in their portfolio. Quality bonds are easily dismissed as a drag on returns, a portfolio afterthought, or even old fashioned. (I admit I had some of the same reservations about bonds for a time.)

However, I've come to view bonds differently in the past few years. The way I see it, money invested in bonds is the money you have; money invested in stocks is the money you are willing to lose in order to obtain potentially large returns.

With this logic, bonds should be the core of a portfolio for most investors, not an afterthought. The more bonds an investor has in their portfolio the better they will sleep, the more stable their portfolio will be, and the more cash access they have available for daily spending without sacrificing the return drivers in their portfolio. Further, it is reckless to risk potentially catastrophic losses with high percentages of your money.

We quickly see that the standard 60/40 (or more aggressive) portfolio falls short of adequate stability. With a small portfolio allocation in bonds, it can be difficult to take money out of the bond side for daily expenses without adding too much risk into the portfolio. If we take money from the stock side, we are effectively cutting down the money tree.

For example, lets say an investor has a $1 million portfolio and needs $50,000 to buy a new car. If they only hold 10 percent bonds, they are effectively taking away half of their safe stuff. With 20 percent bonds, it would be a quarter. That could significantly reduce the stable portfolios of the portfolio and the ability to re-balance into stocks.

Bonds are likely to provide a moderate return over inflation over time. No one gets rich investing in quality bonds. Stocks can return 50 percent or more in a single year, but we know that even diversified stocks are prone to frequent drawdowns of 30 percent or higher. Major markets around the world, including the United States, have seen drawdowns exceeding 80 percent in equities.

To achieve a safe, stable portfolio that supports larger withdrawals, we need a lot of bonds. To achieve wealth building returns we need stocks.

The solution to this problem is a Leveraged Barbell Portfolio. With this approach, an investor can allocate the vast majority of their portfolio to quality bonds. A comparatively small amount of money can be allocated to highly leveraged stocks to achieve high overall portfolio returns.

Building a Barbell Portfolio Around Bonds

For the large bond holding, I believe the best option is to use short-term bonds. Shorter duration bond products are the least likely to lose value due to inflation or interest rate increases. Remember, the bond portion of the portfolio is for stability and moderate income.

There are a number of ways to properly access leveraged exposure to stocks for this style of portfolio. For most investors, the easy option is by purchasing highly liquid leveraged ETFs. There are a number of triple leveraged ETFs listed on U.S. exchanges. Some popular options are UPRO and SPXL (Large Caps), TNA and URTY (Small Caps), TQQQ (NASDAQ), and EDC (Emerging Markets).

Another option is buying deep in-the-money call options on the largest standard ETFs. To get the necessary exposure with call options, an investor could choose LEAPS options on SPY (Large Caps), IWM (Small Caps), EFA (International Developed), EEM (Emerging Markets), or QQQ (NASDAQ).

It is important to only use leveraged ETFs or call options to utilize leverage for this strategy. With both of these instruments it is impossible to lose more money than you allocate to equities. A call option might become worthless and a leveraged ETF might become worth nearly nothing, but they won't go into negative value. You can only lose what you put into the instrument.

This is unlike using margin to buy ETFs or buying indices with futures contracts. With margin and futures you can lose more than you allocate to the equities because your bond allocation will be used to secure the margin. A stop-loss would help, but will not have the same effect or ease of management.

Comparing Returns to a Bond-Based Portfolio

The chart above showed the returns of a standard 60/40 portfolio since 1950. (For data purposes I used T-bills instead of bonds in all charts.) An investor using triple leveraged ETFs would only need to have 12.5 percent of their portfolio in leveraged stocks to achieve similar annual returns to a 60/40 portfolio over the past seven decades.

Here's the comparison chart.

Sources: TheRichMoose.com, S&P, FRED-Federal Reserve St. Louis

On the surface, both charts look very similar. Two nice lines traveling steadily up and to the right. However, with a tiny 12.5 percent allocation to risky assets, the leveraged barbell portfolio has a much different characteristic in drawdowns.

Sources: TheRichMoose.com, S&P, FRED-Federal Reserve St. Louis

In nearly all of the drawdown periods, the Leveraged Barbell Portfolio experienced drawdowns that are approximately half the severity of the standard 60/40 portfolio. Many were less than half.

Only in the 1987 crash, which saw a massive one-day fall that most affects a daily leveraged product, did the Leveraged Barbell Portfolio experience a drawdown that is comparable to the standard 60/40 portfolio. Still, the portfolio never lost more than the amount of risk assets it held at the peak before the crash.

A Leveraged Barbell Portfolio can provide superior risk-adjusted returns. With more than 80 percent of the portfolio in quality bonds at any given time, the portfolio spins off interest income and provides easy access to large amounts of cash without distorting the risk profile or taking away from potential returns.

Comments & Questions

All comments are moderated before being posted for public viewing. Please don't send in multiple comments if yours doesn't appear right away. It can take up to 24 hours before comments are posted.

Comments containing links or "trolling" will not be posted. Comments with profane language or those which reveal personal information will be edited by moderator.

8 Replies to “Risk Mitigation with a Barbell Portfolio”

  1. Would any of the bond etf in your BEST CANADAIAN ETFS page qualify as quality bond to hold in the Leveraged Bartell Portfolio?

  2. Daren (Editor) says:

    I tend to prefer shorter duration bonds for this kind of strategy. Tilts towards higher credit grades are probably not a bad idea when using ETFs. I would also shy away from an ETF that holds a lot of BBB coupons as they are lower quality and barely a jolt away from junk status.
    I won’t give specific recommendations, but feel free to use my ETF page as a guide to some of the Canadian listed bond ETFs that I might choose for investing. Different ETFs have their own pros and cons depending on what you are looking for.

  3. I get your points. Thanks Daren.

  4. Interesting – thanks Daren. In this article you looked at matching a traditional 60/40 split, however, in other barbell articles you have looked at other allocations that with a 3x leveraged ETF would bring your equity position closer to 100%. Has your research and/or experience pointed to a generally optimal allocation to the leveraged equity portion?

    Also, I am currently using dual momentum with US listed ETF’s in my RRSP for the tax benefits. Was thinking of switching my TFSA from VGRO to a barbell portfolio. Using your example in this article, would you recommend short-term Canadian bonds and using Norbert’s gambit to switch 12.5% of my portfilio to USD to purchase UPRO or SPXL?

  5. Daren (Editor) says:

    There are several ways to try approximate 100 percent equities with 3x leveraged ETFs: none of them are perfect. For example, you can approximate based on long-term returns as I have done in this post. Before fees and taxes, that works out to roughly 12.5 percent. You can also measure it by historical portfolio drawdowns. In that method it is upwards of 25-30 percent. There are also impacts related to frequency of re-balancing and how re-balancing is managed if it is done more tactically.
    I think the issue of Canadian vs. U.S. ETFs depends on your personal situation and how you think about currencies. If you’re Canadian, spend in CAD almost always, and expect to live in Canada forever, a good argument could be made that it makes sense to have a decent portion of your portfolio in Canadian dollar denominated assets. Personally, I don’t think we can ignore that the world currently operates on the USD (and less on EUR). This to me means that a good way to hedge against currency risks is to go with USD, especially when your paycheque, pensions, and house are valued in CAD. I hold a lot of USD assets right now, but that’s also based on the current trend of a falling CAD/USD.

    For your situation, you will definitely have the 3x leveraged ETFs in USD. Now you need to decide on your remaining currency exposure. Going half/half on your bond side (example, XSB.TO + BSV) may be a decent compromise until you figure out your personal preferences and currency risk.

  6. Yanniel says:

    Hi Daren.

    I ran a couple of backtests with PortfolioVisualizer that seem to be at odds with your results:

    I ran the allocations below (reinvesting dividends) from 2010 to date (with annual rebalance).

    60% SPY + 40 % SHY has a CAGR of 6.47%.
    12.5% UPRO + 87.5% SHY has a CAGR of 4.33%.

    The leveraged portfolio came short by 2%.

    Any ideas as to what is causing these discrepancies? I am thinking your results cover a longer time frame; while my teats only go back 9 years.

  7. Daren (Editor) says:

    It’s purely a time frame issue. If you look at the comparison chart it shows the standard 60/40 (blue line) catching up to the LBP 12.5/87.5 (green line) since 2009. LBP set at 12.5/87.5 will underperform a standard 60/40 in an uptrending market since the approximated stock exposure is much lower than 60 percent (12.5 * 3 = 37.5). However, the LBP will outperform in downtrending periods. Put together, over full market cycles the LBP has better overall and risk-adjusted returns returns.
    You might feel a bit silly with the lower returns after a 9 year market half-cycle though. If it was easy everyone would do it.

  8. Yanniel says:

    Makes sense. Thanks.

Comments are closed