A little-understood alternative asset class
Investors spend a lot of time, energy and money trying to find the right alternative asset classes or diversifiers to add to their portfolios. Advisers can suggest to clients that patience may be the best alternative asset class in their portfolio. In fact, patience is a portfolio diversifier that clients can actually control.
Let’s take a look at the performance of portfolios composed of large-cap U.S. stocks, non-U.S. stocks and commodities based on several different levels of patience over the period from 1970 to 2015.
The benchmark level of performance was a simple buy-and-hold approach, the quintessential manifestation of patience. The 46-year average annualized buy-and-hold return for large-cap U.S. stocks (the S&P 500) was 10.27%. The return for developed non-U.S. stocks (MSCI EAFE Index) was 8.79%, and the return was 6.92% for broad-basket commodities (S&P Goldman Sachs Commodity Index).
Investor patience was simulated here by the use of trading triggers — that is, threshold levels of performance (percentage moves up or down) that would cause a transfer of money out of the investment account and into a 100% cash account. The initial starting point was assumed to be the client’s investment account. Thus, the first trigger event would result in moving all their money out of the investment account and into the cash account.
LEVELS OF PATIENCE
For example, let’s consider a highly impatient client who has -3%/+3% trigger levels. This client will bail out of the investment account if it experiences a monthly return of
-3% or worse and move all the money to a cash account at the start of the next month.
The client will then earn interest each month (based on what T-bills earned that month) for as long as the funds are in the cash account. This same investor will not transfer their money back into the investment account until asset class has experienced a monthly return of at least +3%.
A somewhat less trigger-happy investor will have higher trading triggers, such as -5% and +5%, -7% and +7% and so on. Here, we’ll review the results of selected investor patience levels that ranged from -3%/+3% to -15%/+15%.
As the chart “Got Patience?” shows, if using simple mirror-image trading triggers, such as -4%/+4%, there were only three combinations that produced a higher return than the buy-and-hold approach if the investment account was composed of 100% large-cap U.S. stocks.
Specifically, a trading trigger tandem of -7%/+7% produced a 46-year average annualized return of 10.32% (just 5 bps above the buy-and-hold return) and required 38 trades over the 46-year period (using a cost figure of $50 per trade). Even if the trading cost was assumed to be only $10 per trade, the 46-year return increased only to 10.425.
Trading triggers of -9%/+9% produced a slightly higher level of return (10.48% with a total of 24 trades), as did trading triggers of -10%/+10% (10.38% with a total of 20 trades). Using -7%/+7% triggers, the investor spent 82.4% of the 46-year period in the investment account and the remainder of the time in the cash account.
Using impatient trading triggers of -3%/+3%, the investor essentially churned himself with 92 trades, and ended up with an 8.53% 46-year average annualized return. A lack of patience produced inferior results compared with simply investing for the long haul.
Even if a trading cost of zero was assumed, the investor who traded 92 times using -3%/+3% triggers ended up with an 8.89% 46-year average annualized return — still well below the 10.27% buy-and-hold return. With this impatient approach, the investor was in the investment account 73.4% of the time and the cash account 26.6% of the time.
If non-U.S. stocks were used as the investment account asset, there was only one mirror-image trading trigger tandem that generated a higher return than the 8.79% produced by a simple buy-and-hold approach.
Using -7%/+7% triggers, the investor executed 47 trades (at a cost of $50 per trade) over the 46-year period and ended up with an annualized return of 8.93%, or just 14 bps higher than the buy-and-hold return. Clearly, this was not worth the effort. Using these trigger levels, the investor spent 75.2% of the 46-year period in the investment account. Assuming a lower trading cost of $10 per trade, the 46-year average annualized return improved slightly to 9.05% using trading triggers of -7%/+7%.
A DIFFERENT STORY
Using commodities as the investment account asset creates a very different story. Here is a case where every mirror-image set of trading triggers produced an outcome better than buy and hold.
The best mirror-image trading trigger tandem was -5%/+5%, which resulted in 65 trades and produced a 46-year average annualized return of 11.38%, far surpassing the 6.92% buy-and-hold return. A trading cost of $50 was assumed. If the trading cost was assumed to be $10, the 46-year return increased to 11.42%.
Here’s how the sequence unfolded: Investors started out with all their money in their investment account on Jan. 1, 1970. Their investment account suffered its first monthly loss of 5% or more in October of 1973 (a return of -10.4%). At the start of November 1973, all the money was transferred to their cash account and earned the monthly return of cash for as long as it was there.
In December 1973, the monthly return of the GSCI was 15.8% — higher than the needed trigger of +5% to summon the investors back into their investment account. Thus, on Jan. 1, 1974, the investors moved all their money back into the investment account. Using triggers of -5%/+5%, there were 65 trades, half of them out of the investment account and the other half back into the investment account. The investor spent 66.1% of the time in the investment account and the balance of the time in cash over the 46-year period.
Over this period, the optimal trading triggers for an investment account composed of 100% large-cap U.S. stock were -5% and +6.5%. These triggers resulted in 50 trades (at $50 per trade) and a 46-year average annualized return of 10.62%, or only 35 basis points higher than the buy-and-hold return of 10.27%. The 46-year return was 10.74% if assuming a $10 trading cost.
The investor spent 73.7% of the 46-year period in the investment account. Clearly, if the investment account and cash account are exposed to taxation, this is not worth the effort. Even if the accounts are tax-protected, the value of using trading triggers for large-cap stocks is questionable. Long-term patience would seem to be the more prudent approach.
The optimal triggers for international stocks were monthly returns of -5% and +7.25%. These triggers produced a 9.83% average annualized return and 53 trades, and had the investor in the investment account 65.9% of the time. Trades were assumed to cost $50. At a $10 trading cost, the 46-year return was 9.94%.
The buy-and-hold return was 8.79%, so this trading approach produced a benefit of 104 bps (assuming a $50 trading cost). Taxation would probably completely swallow that advantage, but if the money is tax-protected, using trading triggers did demonstrate some value.
Whether or not the trading triggers of -5% and +7.25% will be optimal going forward is unknowable, which makes the case for a less trigger-happy and more patient approach.
The optimal triggers for commodities were monthly thresholds of -5% and +4.5%. These triggers produced an 11.73% average annualized return and 65 trades, and had the investor in the investment account 67% of the time. Trades were assumed to cost $50. At a cost of $10 per trade, the performance improved to only 11.77%.
The buy-and-hold return for commodities was 6.92%. Thus, using trading triggers of -5% and +4.5% performance was increased by 481 bps (assuming a $50 trading cost). Historically, commodities are asset class that benefited from being managed a bit more impatiently.
BUY AND REBALANCE
There is another alternative to buy-and-hold versus trading triggers: buy and rebalance. Rather than using trading triggers that require monthly monitoring, simply build a multi-asset portfolio and rebalance each asset class back to its prescribed allocation at the end of each year.
Four different portfolios are illustrated in “The Other Alternative.” The first portfolio consisted of 50% large-cap U.S. stocks and 50% non-U.S. stocks. It was rebalanced at the end of each year to those same allocations, assuming a rebalancing cost of $50. Its 46-year average annualized return was 9.68%, with a 17.94% standard deviation, a worst one-year loss of -40.19% and a worst three-year drawdown of 40.37%.
Adding a third asset class (commodities), the return increased slightly, standard deviation dropped by almost 18%, the worst one-year loss increased slightly and the worst three-year drawdown improved substantially. Given the perceived volatility of commodities, this may seem odd, but it reveals the value of adding a low-correlation ingredient into a portfolio.
Commodities had a correlation of -0.05 to large-cap U.S. stocks and a 0.08 correlation to non-U.S. stocks over this 46-year period. A correlation of 0.00 between asset classes is the goal, so these correlations are terrific. By comparison, the correlation between large-cap U.S. stocks and non-U.S. stocks was 0.66 — a bit higher than we would like it.
When cash was added as a fourth asset class, the return naturally declined, but the risk measures all improved substantially. Finally, a diverse seven-asset portfolio was assembled that consisted of large-cap U.S. stocks, small-cap U.S. stocks, non-U.S. stocks, real estate, commodities, bonds and cash — all equally weighted.
Performance bumped back up to 9.67% (the same level of the two-asset, all-equity portfolio), but standard deviation declined to 10.31%, the worst one-year loss was lower than the other three models and the worst three-year drawdown was the lowest, at -13.40%. If rebalancing had zero cost, the 46-year return for the seven-asset portfolio was 9.78%.
NO BAILING OUT
This buy-and-rebalance approach also requires patience. A patient investor doesn’t bail out on asset classes that are experiencing turbulence during the year. In fact, rebalancing means investors will give the underperforming asset class some extra money at the end of the year to bring its allocation up to the needed level.
And here is the emotionally hard part — the needed money to do the rebalancing will be acquired by selling a bit from the best performers each year. Sounds crazy, doesn’t it? Rebalancing turns out to be counterintuitive — which is why it works.
If the seven-asset portfolio was not rebalanced each year, the 46-year return was 9.54% (versus 9.67% if the portfolio was rebalanced annually at a $50 rebalancing cost). Not a huge difference, it may seem — but it resulted in an ending account value that was $35,600 lower than the annually rebalanced seven-asset portfolio (assuming a starting investment of $10,000). If the rebalancing cost was assumed to be zero, the rebalancing advantage was over $68,000.
“All things come to those who wait” is a phrase that has advocated patience since the 19th century. In all likelihood, patience has always been a rare attribute. Indeed, patience acts as a diversifier or “alternative” against our tendency to act rashly. For investors, the important companions of patience are broad diversification and periodic rebalancing.