7.3 Diversification across asset classes (stocks, bonds, and cash)

20160731_155901

I established in Articles 7.1 and 7.2 that buying funds can provide easy diversification at little cost, particularly for stocks.  At the same time, some moderate diversification of stock types has the potential to incrementally reduce volatility or boost return performance of your investment portfolio.  And if stock types become substantially less correlated in the future, the benefits of stock diversification could become more than just “incremental”.  The main question left is whether and how to diversify across the major asset classes of stocks, bonds, and cash.

Stock/bond mixes

You’ve probably noticed that many investing guides recommend some combination of stocks and bonds, purportedly so you get the benefits of both.  How does that investment approach compare to owning only stocks or only bonds?  Here are the return and volatility information from a recent Vanguard study for a few possible combinations of stocks and bonds:

 

 

 

 

 

As you would expect, the risk/return profile for combinations of stocks and bonds is somewhere between the profile for 100% of either investment by itself.  Just like I discussed in Article 7.1 regarding stocks, we can place combinations of stock/bond portfolios on our volatility vs. return cross plot.  This graph shows the general relationship between stocks and bonds in the period 1971 to 2015 in an investment portfolio using the calculator available at “Portfolio Visualizer”.

stock-bond-cross-plot

Similar to the stock diversification example in Article 7.1, there is a “tangency portfolio” of about 20% stocks and 80% bonds that historically has moderately boosted portfolio returns while adding little additional volatility.  As compared to our diversified stock portfolio, the volatility of this 20/80 portfolio is quite low (about 6%), which appeals to many people.  However, the returns are also several percentage points lower than has historically been achieved with stocks, which is a large performance difference in this game of compounding returns over long periods.

Selecting a stock/bond mix – While the 20/80 portfolio answer appears elegant, if we use a different period of past stock/bond performance, we would get a different answer.  I examined this “back testing” issue in the stock diversification discussion in Article 7.1.  Here is an example of the same type of risk/return plot, over a very similar total period, but separating out the data by consecutive decades.

efficient-frontiers-by-decade

So, we might arrive at a totally different conclusion regarding the right balance of stocks and bonds depending on what slice of history we choose to consider.  In the case of the various curves in the above graph, the amount of stocks to hold in your portfolio could range between about 10% and 50%.  Put another way, if you chose a 50/50 stock/bond portfolio based on an analysis of one of these periods and then held that portfolio from 1960 to 2004, during most of those decades you could reasonably conclude you had the wrong mix of stocks and bonds resulting in additional risk without much additional return.

It’s worth noting again that these risk/return plots are useful, but they also illustrate the limitations of using historical data to predict future portfolio performance, because history will never be exactly repeated in the future. Such evaluations draw from only 100 to 150 years of stock and bond history.  Because many people will be investing for periods of 50 years or more (for example starting at 35 years old and dying at 85), all available history is only providing just 2 or 3 unique periods of similar length to back-test against.  In statistical terms, our unique sample size is very small (n=3).  It seems highly unlikely that the next 50 years will be exactly the same, or even vaguely similar to any one of the hand full of relevant historical examples that are available for comparison.  (Note that some people try to overcome this limitation by conducting an analysis that uses multiple overlapping segments of historical risks and returns data and then identifying the optimal portfolio among all possible outcomes.  However, this “Monte Carlo” approach is evaluating many similar segments of data multiple times over, and it still only draws from the same 150 year total period, which is still likely to differ significantly from the next 50 years.)

This type of risk/return analysis is the source of a common recommendation to hold a retirement portfolio of 60% stocks and 40% bondsWith a relatively even balance of the two assets, bonds moderate volatility and the overall portfolio is still expected to achieve relatively decent annual returns.  The stabilization provided by bonds is often referred to as the “ballast” function in a mixed portfolio.  The mixed portfolio “ship” moves forward at a decent speed and on a more consistently even keel.  Importantly, stocks provided about 71% of the return in a 60/40 portfolio from 1928 to 2014.   So, most of the “speed” is provided by stocks and most of the “even keel” is provided by the ballast of bonds.  But what are we giving up in possible additional returns to stabilize this perceived risk represented by volatility?  Here is one graph from Meb Faber showing the historical returns of the 60/40 portfolio as compared to all stocks and all bonds.

returns-60-40-bonds

With a 40% mix of bonds the long-term return on $100 is about $13,000, as compared to about $51,000 return when investing in stocks alone.  We need to be careful when projecting returns over very long periods, but we still see significant divergence between the two portfolios in periods as small as 10 years.  Here is another example over a perhaps more relevant duration of about 40 years using Portfolio Visualizer.  The blue line (Portfolio 1) is 100% stocks and the red line (Portfolio 2) is 60% stocks and 40% intermediate duration bonds.

60-40-bonds-port-visual

In this example, with an initial $10,000 investment in 1972, the 100% stock portfolio ends with $170,000 more in value than the 60/40 portfolio.  More importantly, this historical back test covers the unprecedented 30 year bond bull market of falling yields and rising bond prices that won’t be repeated any time soon.  As discussed in Article 6.2, the future returns for bonds are expected to be very low because of today’s historically unprecedented low interest rates.  So, moving forward, the 60/40 portfolio is likely to perform even worse than implied by the above graph.

Balancing risks and returns with a stock/bond mix – In the stock diversification discussion (Article 7.1) we came to the somewhat surprising conclusion that adding small amounts of highly volatile stocks could potentially boost returns while hardly impacting volatility.  Let’s see if the opposite is true for stock/bond mixes.  If we add a small amount of bonds to a mostly stock portfolio, can we decrease volatility without damaging our stock returns too much?  Using Portfolio Visualizer again, here is a comparison of a 100% U.S. stock portfolio (Portfolio 1) to a portfolio with 20% in a total bond fund instead (Portfolio 2; an “80/20” portfolio).

80-20-port-visual

It’s worth pulling in some detailed statistics on these two portfolios as provided by Portfolio Visualizer.

stock-bond-stats-1-port-vis

As expected, the 100% stock portfolio (Portfolio 1) did better than the 80/20 Portfolio (Portfolio 2) in terms of final balance and annual return (noted as “CAGR” in this table), with Portfolio 1 coming out about $77,000 better on an initial $10,000 investment.  But the volatility (“std. dev.”) of the 80/20 Portfolio was more than 3 percentage points lower, and the maximum drawdown and worst year were both about 9% better.  (But it still appears unlikely that most individual investors would be particularly calmed by a one-time 32% drawdown instead of a 41% drawdown.)  As with other analyses of this type, using a different time period will yield different results, but in almost all cases, the two portfolios perform in a similar neck-and-neck fashion, with the 100% stock Portfolio usually providing slightly better returns.  The period from 2000 to 2015 is particularly interesting once again, because provides a slightly different answer.

stock-bond-stats-2-port-vis

In this case, the 80/20 Portfolio (Portfolio 2) actually provided a slightly better return and with lower volatility.  Given uncertainties about the future, this lends some credence to balancing the 100% stock portfolio with at least some ballast from bonds.  But once again this result must be viewed in the context of the constant and unprecedented bond bull market through this period.  That favorable bond market can’t be repeated in the next 15 years because bond yields are now so low.

Whether you prefer a good chance of slightly more money (with 100% stocks) or a good chance of a slightly smaller maximum draw down (with the 80/20 portfolio) is really a personal decision.  However, from a mindful perspective, it feels like nothing much is lost by taking a reasonable chance of suffering a 41% draw down instead of a 32% drawdown, and something is clearly gained by having a good chance of an added percent or so of return over the long term.

Mindful conclusions about stock/bond mixes – Despite the likely poor returns of bonds moving forward, having some portion of your portfolio serving the ballast function of less volatility and less downside potential may be important for some investment plans.  In all likelihood, using a relatively even balance of stocks and bonds today will result in substantially diminished returns for the foreseeable future.  And as we’ve seen, the conventional reason given for accepting these diminished returns is so that you can avoid the anxiety produced by stock volatility along the way.  Or perhaps more fairly, the conventional concern is that individual investors are too emotional to stick to a “roller coaster” plan involving mostly stocks and will panic sell during market declines, resulting in lower actual returns than if they had followed a more “balanced” plan.  That’s probably why William Bernstein, a widely respected author on investing, has said:

  • “For most people, over the long term it probably is optimal to be 100% in stocks, but almost nobody can execute that optimal strategy.  So, a suboptimal strategy that you can execute is better than an optimal one you can’t.  For most people, the best path is one that involves owning a large amount of bonds so they can sleep at night during the bad times.” [emphasis added]

This is the crux of the problem, and the Mindfully Investing website is dedicated to the idea that mindfulness can solve that problem.  If you are mindful or working toward mindfulness, you are the one and only person able to determine whether you can in fact endure a roller coaster plan involving mostly stocks.  I recommend you should literally meditate on this question and consider your likely attitude when your investment account plummets by something like 50%.  The less mindful investor is probably better off with a relatively even balance of stocks and bonds, and/or or using an investment management professional who can help them avoid panic decisions.  As we have discussed in Article 4.2, the more mindful you are, the more likely it is that you can maintain your plan, even a very aggressive one.

Our mindful conclusions about stock bond mixes are mostly consistent with prominent authors addressing asset allocation.  Again using William Bernstein as an example, he makes some key conclusions (paraphrased by me) in theIntelligent Asset Allocator”:

  • Even the most risk averse investor should own some stocks, because of their potential to increase bond returns.  Small amounts of stocks in a bond portfolio can actually decrease volatility.
  • The addition of a small amount of bonds to a stock portfolio reduces volatility, while reducing returns only slightly.  (And I will add that a mindful investor may not find reductions in routine volatility particularly helpful.)
  • Don’t put too much emphasis on recent outcomes (“chasing performance”) or risk/returns over spans of less than two or three decades.  Some assets that have had poor showings in the past are likely to recover and vice versa.

We have not yet talked about cash in the asset allocation mix.  As you may have guessed, cash will also have some impact on the volatility of stock portfolios as well, which gets to our next question.

What about cash?

Recall the mindful definition of cash from the introduction to Article 6.  Specifically, cash is highly liquid (meaning you can convert it into money in hand without much delay or hassle) and broadly includes relatively short term bank certificates of deposit, bank accounts, and money market accounts that can currently return up to 1 to 2% annually (as of November 2017).

It’s reasonable to say that there is no conventional wisdom when it comes to cash and investing.  This Wall Street Journal article summarizes answers from 13 “experts” on the question: “Should you keep cash in your portfolio?”   The answers are all over the map, and in some cases are diametrically opposed.  Fortunately, a mindful investing approach does not rely on expert opinions.  Using this article as an example, the primary knocks against holding cash in a portfolio are:

  • Cash provides no return – This appears to be a rather narrow view, because a reasonable definition of cash can equate to a 1 to 2% return (without inflation adjustment), which is not too different than the current expected return on intermediate term bonds (as of November 2017).
  • Cash is better used to pay down debts – This is a reasonable point, but because I am talking mostly about investing for the future, I am operating under the assumption that you don’t have an unreasonable debt burden and large debts like mortgages will be paid off by the time you retire or otherwise need your money.
  • Cash has no place in a qualified retirement account or similar because short term money can’t be withdrawn without penalties – This is an important nuance about pre-retirement vs. retirement status that is better addressed in Article 8 on investing over time.

The primary cited advantages of cash are:

  • Cash provides ballast and peace of mind for risk adverse investors – This is not very mindful, but true for most people.
  • Cash is necessary to fund short-term non-investing expenses – The basic idea is to have cash available to fund short term expenditures, which in these “expert” answers ranges from 3 to 7 years.  Dramatic market declines can occur very quickly, so using cash reserves in the short term allows other investments to recover and helps avoid the real risk of a permanent loss.
  • Cash is useful to respond quickly to changing investment opportunities and market/economic events – Although a few of these experts warn against “market timing”, this may be a concept worth considering.

On balance, we can mostly dismiss the primary knocks against cash as being over simplistic or as addressing relatively specific scenarios.  That suggests that there may be some advantages to having some cash.

Let’s start with the question of cash as ballast.  In this view, cash is just another asset that we can freely consider in asset allocation just like stocks and bonds.  I’ve already discussed that in mixed portfolios bonds provide ballast because of their lower volatility and lower returns.  Do cash or bonds (or both) work better in the ballast role?  First off, the correlation relationships between cash and other assets are interesting and rather rare.  This graph from Schwabe gives some sense of cash correlations over the last 10 years.

cash-correlations

As the title of the graph indicates, cash is pretty well known to be the least correlated of assets.  That suggests cash should have a decent potential to improve the balance of risk/return in a mixed portfolio.  The risk/return cross plot for a cash and stock mixed portfolio looks nearly identical to the bond/stock cross plots I’ve already presented, except the returns and volatility are even lower.  So instead, let’s look at investment growth in the volatile period of 2000 to 2015 for a 50/50 mix of stocks and cash (Portfolio 2) as compared to 100% stocks (Portfolio 1) using Portfolio Visualizer again.

cash-graph-1

In this period, cash served the ballast function very well, moderating declines in 2002 and 2008.  Further, the cash ballast portfolio actually kept ahead of the 100% stock portfolio for nearly 13 years, with the cash ballast portfolio eventually falling behind in the stock bull period from 2012 to 2015.  Given its low correlation with stocks, could cash be even better than bonds for the ballast function?  At times it can be.  Here is a graph of a period of rising and then falling interest rates with 100% stocks (Portfolio 1), 50/50 stocks and cash (Portfolio 2), and 50/50 stocks and 10-year government bonds (Portfolio 3).

cash-graph-2

Portfolio 2 with cash ballast actually performs a little better than Portfolio 3 with bond ballast for much of this period, and they end up in about the same place.  To some extent I am emphasizing periods of economic conditions that make the cash option look good.  To be fair, let’s look at these same three portfolios over the entire period available from Portfolio Visualizer.

cash-graph-3

As you would expect, over the long term the bond ballast (Portfolio 3) does better than the cash ballast (Portfolio 2).  The bonds are providing substantially better return over the long term, while cash is actually losing inflation adjusted value over this period.  However, it’s pretty clear by now that cash can perform the ballast function in a mixed portfolio very similar to bonds.  Going a step further, you can approximate the effect of bond ballast using smaller amounts of cash as ballast instead.  Here is one more graph over the entire available period with the cash ballast Portfolio 2 reduced to 30% cash instead.

cash-graph-4

The end result using less cash ballast is nearly the same as using larger amounts of intermediate term bond ballast.  Ballast should function to decrease the volatility of the portfolio over time.  With cash as ballast, the above graph shows the ride along the way is much less bumpy than using 100% stocks and not too different than using larger amounts of bond ballast.  For example, the maximum draw downs in this period were:

  • 41% for the all stock portfolio (Portfolio 1)
  • 26% for the 70/30 cash ballast (Portfolio 2)
  • 19% for the 50/50 bond ballast (Portfolio 3).

We’ve gone through a lot graphs to show a pretty simple point.  Cash can serve as portfolio stabilizer just like bonds, and because cash is so stable, you can use less cash as ballast and achieve the same outcome as more bond ballast in a mixed portfolio.

Given bonds and cash can perform very similar portfolio functions, how are we to ultimately decide on using bonds or cash (or both) for the ballast function?  The second two advantages of cash summarized from the “experts” article above help break the tie.  The two advantages of cash are that 1) it can be used to fund short term expenditures without forcing you to sell long term investments at a loss and 2) it can be used to buy depressed assets when opportunities arise.  Bonds don’t have those advantages.  Although bond volatility is less than stocks, we have seen several examples already where bonds can lose money in the short term and may take several years to recover from those deficits.  Selling bonds in the short term to either fund expenditures or pursue investment opportunities may result in permanent investment losses, which is the primary risk we are trying to avoid.  The obvious disadvantage to cash is that it provides very little return, usually below the rate of inflation.  However, as I have discussed previously, bonds are currently in the exact same situation and are expected to fail to provide a positive real return in the near future.  Until interest rates rise substantially, bonds aren’t providing their historical key advantage over cash of a positive real return.

Mindful Conclusions about diversification of stocks, bonds, and cash

Our primary mindful conclusions across the entire Article 7 series include:

  • As we saw in Article 7.1, adequate stock diversification is likely obtained by buying a moderate number of low cost index funds that contain stocks varying to some degree across the main stock types.  Such a simple diversified stock portfolio may turn out to provide some risk/return advantages, but it is not guaranteed to outperform a less diversified stock portfolio in any given future time span.
  • Per Article 7.2, if you choose to have a substantial bond allocation in your portfolio despite that bond returns will likely mimic cash returns for the foreseeable future, a mindful perspective suggests there is certainly no harm in holding a more diverse array of bonds.
  • In Article 7.3, we found that the normal advantage of bonds over cash as ballast in a mixed portfolio with stocks is currently absent, because bonds are not expected to provide a real return above inflation anytime in the foreseeable future.
  • At present, our mindful conclusion is that cash has more advantages than bonds as ballast in a portfolio of mixed investments, because cash can be used for short term expenditures and investment opportunities with no risk of permanent loss (other than slow and persistent inflation erosion).
  • Therefore, the most mindful long term mixed portfolio for right now is likely heavy on stocks (in the 80% or more range) and contains a relatively small amount of cash ballast.

Again, these conclusions are only valid while the status quo continues.  When intermediate bonds yields become measurably better than cash, say about 4%, it is probably time to reconsider using bonds as ballast instead of cash.  We have seen that the total expected return of bonds is highly correlated with the bond’s current yield.  So, another way to decide when to switch to bonds as ballast might be when the current bond yield is a percentage point or two above the current and expected rate of inflation.  That would mean that bonds are likely to provide a real positive return, while cash ballast would slowly lose value due to inflation.  Regardless, holding too much cash for too long is clearly the one method that guarantees decreasing the purchasing power of your money over time.  Cash should almost never constitute a major portion of your portfolio for long periods.

The approach of using cash for ballast now and bonds for ballast when interest rates rise, does not completely address the advantages of cash as a reserve for short term expenditures and investment opportunities.  So, even when bond yields improve substantially, it may make sense to hold some cash in a stock/bond/cash mixed portfolio. The question then becomes how much to hold of each.  The advantages of cash are related to short term factors, and the advantages of bonds and stocks in portfolios are related to long term factors.  To finally and fully determine the “best” stock/bond/cash mix we therefore need to look at the issue of investing over time, which is addressed in Article 8.

 

2 comments

  1. Roadrunner says:

    Very detailed and informative article and I can tell the same about the whole site. It has a great structure, you should think about publishing it as an e-book once you’re finished with all chapters.
    I really enjoyed all articles I read so far here.

    • Karl Steiner says:

      Thanks for the vote of confidence. I am a fan of your site, and I am still in the process of working my way through it. Lot’s of good information!

Leave a Reply

Your email address will not be published.