About a month ago on CNBC there was a segment regarding ETFs vs. “hand picking individual stocks” where two guest luminaries debated their respective opposing positions. They also had recent $300K Jeopardy! winner Arthur Chu on the segment, and Chu said both of those approaches were far too risky for him. Chu said he invests almost exclusively in (relatively) safe bonds.
One of the guests then said “Arthur! You're so young! You're barely 30 years old! You should be in 100% equities!”
That doesn’t seem right, I thought. 100% equities? Really?
Then I re-read the 15-year-old article “Why Not 100% Equities?” published in the Journal of Portfolio Management (winter 1996) by now-current hedge-fund legend Cliff Asness (a Fama protégé by the way). Asness was still at Goldman Sachs when he published the article.
The long-term annual ROI of equities (stocks) is much higher than the analogous long-term ROI of bonds. The traditional 60/40 portfolio splits the difference. See the following chart, which is 2nd nature to most investors, and is typically used to argue for the superiority of stocks over bonds. Indeed, in only a handful of 10-year timeframes does a 60/40 portfolio outperform a 100/0 portfolio:
In the above chart, the S&P 500 is used for equities, and the Ibbotson total return series for long-term corporate debt.
So, the obvious question is why bother to have any bonds in your portfolio if you are a long-term investor and can tolerate the risk?
Asness goes on to say perhaps the most important lesson of modern finance is that under reasonable assumptions the choices of (1) which risky assets to hold, and (2) how much risk to bear are independent choices. Under some simple assumptions, an investor chooses a portfolio of risky assets to maximize the portfolio's Sharpe ratio. Then, given the maximal Sharpe ratio of the portfolio P, the investor then chooses the proper mixture of P and riskless cash. This mix will vary from investor to investor because of differing tolerances for risk, but the relative weights among risky assets will stay constant. Feasible portfolios that maximize expected return for a given amount of risk are said to be "efficient."
The following table gives the data from 1926 through 1993 (remember, the paper I'm summarizing was published in 1996), restating the conclusion of the 1st chart posted:
Note that the comparison isn't really fair, as the 100% equities portfolio has substantially more variability (risk) in it than the 60/40 portfolio or the 0/100 portfolio.
Constructing a new portfolio makes the comparison more fair. Imagine an investor has already determined that (a) the 60/40 portfolio is the optimal portfolio of risky assets, and (b) the desired amount of risk is the same as a 100% stock portfolio (risk means standard deviation, of course). For a $1 investment, a NEW portfolio can be constructed by purchasing 20.0/12.9= $1.55 of the 60/40 portfolio, financing the extra 55 cents by borrowing.
The following charts restate the former, including this new "levered 60/40" portfolio:
Asness shows that for the exact same amount of risk as a 100% equity portfolio, one can instead have a levered 60/40 equity/bond portfolio that provides a higher compound annual return. Yes, a
higher ROI for the
same amount of risk. He uses his results to show that even very long term investors (e.g., 100-year investors such as university endowments) probably should not have 100% equities even in light of the historical superiority of equity returns relative to bond returns.
(He's financing the 55 cents of borrowing for each $1 invested by borrowing at whatever the then-current 1 month T-Bill rate is).
From 1926 through 1993, with the same initial investment of $1, the 100% equity portfolio grows to $800 while the levered 60/40 portfolio grows to $1291. Even though a 100% bond portfolio grows to only $40, using bonds in conjunction with stocks and leverage leads to an investment that grows to $1291. The investor who owns 100% stocks must bear the same risk and receive only $800.
Asness goes on to analyze scenarios with differing ways to look at risk (e.g., worst-case scenarios, etc) plus lots of other cases as well.
*****
I don't know what the data look like since he published his article in 1996 – there have been a ton of bad things the past 15 years such as the collapse of LTCM, the Russian debt crisis, the Asian (financial) Flu, the dot-com collapse, the bursting of the housing bubble, the Great Recession…
But in light of a talking-head on CNBC advising the Jeopardy winner to be in 100% equities, I thought I'd pass this along.
Please register to participate in our discussions with 2 million other members - it's free and quick! Some forums can only be seen by registered members. After you create your account, you'll be able to customize options and access all our 15,000 new posts/day with fewer ads.
