Editorial note: this post is part of the Cornerstone Foundations series that explains our firm’s philosophical thinking on matters of critical importance to retirement savers and retirees alike.
Market and economic data are slowing this time of year, so we thought it was a good opportunity wax philosophical. Kidding aside, Cornerstone Financial Services (CFS) employs a financial planning philosophy that is dynamic and more tailored to individual needs than traditional strategies of bygone eras, and we want to expand on some of those ideas here.
This Cornerstone Foundations post will focus on two concepts – portfolio diversification and sequence of returns (SOR) risk – and how CFS approaches both of them to benefit clients.
Stock Portfolio Diversification
Optimal diversification within individual equities is a popular topic, especially since sector rotations have been so prevalent. At CFS we have two proprietary individual stock models – U.S. Equity Dividend Growth and U.S. Equity Outliers (we also offer seven exchanged-traded fund models). Each is an equally weighted baskets of stocks. We’re often asked questions around the decision that determines the number of stocks making up each model, so we’d like to take some time today to discuss our philosophy on individual equity portfolio diversification.
As portfolio model managers, we believe the biggest trade off in constructing a portfolio is between specialization and risk reduction. In other words, the fewer stocks we have to research to include in the portfolio, the better. With fewer selections, our understanding of the underlying companies improves by forcing us to concentrate on our best ideas which consequently gives us better chances at generating excess returns.
Thus, our positions are high on conviction and allow us to wholeheartedly believe in the thesis of these picks.
But on the flip side, the fewer stocks there are in a portfolio, the greater the likelihood of volatility. And if the chances are greater for more volatility, they’re also greater risk for the chance of outsized losses. In general, fewer stocks mean more risk.
So, what’s the right balance? At what point is peak diversification achieved, therefore guiding our firm principles regarding the number of stock positions in our two model portfolios? We’ll show you by comparing diversification benefits across different portfolio characteristics.
It’s important to note that much of this information, including the following three charts, came from quantitative research from the CFA Institute, which is an organization dedicated to the advancement of investment industry ethics, integrity, and professional standards. The CFA Institute has its thumb firmly on the pulse of the investing industry, and we think its data is useful. This data has been quantitatively proven across research paper after research paper but we have chosen to illustrate the CFA data for simplicity and authoritative purposes.
To lay the groundwork, all the data you will see was constructed from a random portfolio with a given number of equally-weighted stocks in each style. Volatility was calculated using monthly returns from 2005-20. This procedure was conducted 100 times, averaging the volatility across all iterations. It’s important to understand these are randomized trials done several times over, not cherrypicked data.
For each style cohort you'll see, there is an average standard deviation (which measures relative riskiness) for each portfolio based on the number of stocks held. We’ll look at different scenarios, like large-cap versus small-cap stocks, to analyze how the number of stocks held affects portfolio volatility.
Market Capitalization
Our first example compares large-cap and small-cap portfolios. The average volatility of a portfolio holding 10 large-cap stocks is 20%. A more diverse large-cap portfolio of 40 stocks has an average volatility of 17%. So, adding 30 additional stocks to the mix only lowered risk by 3%, which doesn’t seem like a worthwhile volatility mitigator.
At 10 stocks, the small-cap portfolio had an average volatility of slightly more than 32%. Adding 30 more equities lowered it to 25%. The additional effort of adding 30 stocks decreased relative risk by about 7%.
In both scenarios, the standard deviation flatlines around 25-30 stocks (green box in chart):
*This material is strictly for illustrative, educational, or informational purposes only
Dividend Payers
Since one of our stock models is dividend based and the other more non-dividend based, we felt it important portfolios of dividend paying stocks and non-dividend paying stocks.
In the below chart, the non-dividend portfolio’s volatility only fell by 5% when going from 10 to 40 stocks. In the dividend-paying portfolio, going from 10 to 40 stocks only produced 3% less volatility.
But once again, we can see how optimal volatility mitigation occurs around the 25-30 stock level:
*This material is strictly for illustrative, educational, or informational purposes only
Growth and Value
We found the comparison between growth stocks and value stocks to be quite interesting because most investors think a growth-oriented portfolio will be significantly more volatile than a value portfolio. While our equity models don’t distinguish between value and growth (instead mostly distinguishing by dividend payers and non-dividend payers), we thought it was important to examine this scenario because many investors and funds do make such a distinction. So, we wanted to see if there are volatility mitigation differences.
Spoiler alert: there really isn’t much of a difference in relative risk between value and growth stocks.
The chart below shows the minimal variation in volatility as the number of stocks increased in both growth and value portfolios. Risk reduction was consistent across both investing styles. Again, the sweet spot of relative risk mitigation is 25-30 stocks:
*This material is strictly for illustrative, educational, or informational purposes only
The lesson here is growth versus value is an irrelevant classification when it comes to portfolio risk construction. Of course, that runs counter to many investors’ thinking, but the data doesn’t lie.
How CFS Does It
Circling back to how we do things here at CFS within our individual equity portfolios, we found the above studies to be helpful and illustrative. In the charts above, we can see the optimal effective diversification strategy in terms of number of stock holdings when using an equal-weight strategy is 25-30 stocks across different strategies (e.g., large-cap vs. small-cap, etc.).
We’ve embraced this methodology at CFS. We construct our individual equity portfolios so they hold exactly 25 stocks at a 4% weighting each, rebalancing quarterly so the ratios remain aligned. It’s yet another example of how we use data to drive our decisions.
The information above has guided us over the years to settle on equity portfolio models being constrained to 25 individual holdings. Research proves that holding more positions does little to mitigate portfolio risk and volatility.
In fact, we think holding excessive positions in a portfolio is essentially a false flag when it comes to diversification. Thus, we intentionally restrict our equities models to our 25 best, fact-based, research-backed ideas.
Sequence of Returns
Another big topic for our clients (and all retirement savers in general) is sequence of returns (SOR) risk.
Essentially, SOR risk is a situation where sequencing or ordering of yearly rates of return affect the overall returns, depending on when money is deposited or withdrawn during that span. Here are four common SOR risk examples:
- Borrowing against a residence to invest elsewhere
- Margin accounts
- Pre-retirement savings (i.e., contributions)
- Post-retirement spending (i.e., withdrawals)
The crux of SOR risk is that we can predict results much better over long periods of time than short periods of time, and SOR relates specifically to short periods of time. While we're still considering long timeframes, the interim payments and withdrawals are made over short time periods where we cannot predict the ups and downs of the underlying asset values.
In other words, volatility makes market timing impossible. Of course, all investors know the futility of market timing (or they should).
To illustrate the SOR concept on a macrolevel, we wanted to look at hypothetical yearly returns of the S&P 500 over a three-year period where return rates were 0%, 10%, and 20%. When we include yearly contributions, you can see the effects of sequencing.
In the below two examples, we'll start with $0 and add a yearly deposit of $1,000 to see how sequencing affects the results for contributions:
*This material is strictly for illustrative, educational, or informational purposes only
The ending balance for Sequence #1 is more than 10% greater than Sequence #2. Keep in mind this is only for a short period of time. For longer periods, the difference can be significantly more dramatic.
So, unless you can predict the short-term swings in the market, you can't accurately predict the effects of SOR. Remember, market timing is impossible.
But how can we reduce this risk? More on that later.
First, let’s look at SOR risk related to withdrawals. This is important because while debunking the “4% Rule” we found that over a 30-year span there is little consistency on what an ending balance will be in a retirement account. Some come close to running out of money, while others gain HUGE in retirement, even as withdrawals occur.
Consider that if you were 25 years old in either 1973 or 1983 AND magically could predict the future, you could easily do a time value of money calculation to see the savings needed to reach a retirement goal of $2 million in 30 years. But as you can see, even a one-year difference can change the results by millions of dollars, and it’s all due to SOR risk:
*This material is strictly for illustrative, educational, or informational purposes only
In the table above, the average return in a 30-year span was about 10.2%. But retiring in 1973 versus 1983 produced a $1 million difference in ending principal. The reason is that when you consider SOR for savings, you want higher returns to occur later in the span than when you’ve accumulated. In 1973-2002, you benefit from the big tech runup in the late 1990s. In 1983-2012, you get that runup, but then suffer from market crashes caused by the tech bubble bursting and the subprime mortgage crisis.
Now let's look at retirement spending. This is where retirees can really be hit by SOR risk.
Let’s say you retired in 1973 with a $1 million portfolio, equally split between stocks and bonds, and employed a rule to spend 4% of your assets annually to live in retirement.
At the end of 30 years, your account balance would be down to about $500,000. If you needed it for only 30 years, that’s great. But what if you needed it for 35 years? That could be a problem. In contrast, using the exact same scenario but retiring one year later (1974 instead of 1973), the account would be worth $3 million:
*This material is strictly for illustrative, educational, or informational purposes only
To be sure, this example is extreme. But the numbers don’t lie – SOR risk is real. There are two main reasons for this huge one-year difference.
First, inflation was rampant in these early spans, causing withdrawals to be larger in 1973. Second, the market took a significant hit in 1973 that didn’t affect retirees in 1974.
One of the biggest problems with the “4% Rule” is it can suffer a double whammy of market downturns and inflation. In severe cases, withdrawals early in retirement in adverse market conditions can severely diminish a nest egg’s ability to provide income throughout the whole retirement. The worst cases occur when retirees rely heavily on investment accounts for income in severe bear markets with high inflation. In these scenarios, retirees are forced to sell diminishing assets to fund their increasingly costly lives.
The traditional thinking on how to avoid SOR risk when it comes to retirement planning or distribution planning is to plan for the worst case, which is the primary concept behind the Monte Carlo model. But we find that this strategy can lead to a suboptimal retirement due to continual spending cuts and income levels that can’t handle inflation.
So how does CFS eliminate or mitigate SOR risk?
Our philosophy is to use a goals-based planning approach that is dynamic (i.e., adjustable on a yearly basis) and based on income, income, income. We aim to elevate the amount of guaranteed income in retirement, so you have a stable base from which to live. This often comes from pensions, annuities, Social Security, interest, and dividend yields.
The larger this base is, the more it does to minimize the need for income from selling assets/withdrawing from retirement accounts. And the smaller this “withdrawal” number is, the more it reduces (or eliminates) SOR risk during your distribution years.
This is goals-based planning in a nutshell. It is flexible to serve individual needs and runs counter to the “4% Rule,” which is more rigid. Our goals-based process requires active involvement and attention paid to what’s actually happening. Conversely, the “4% Rule” is a lazy attempt to “set it and forget it,” which as we’ve seen, can be dangerous for retirees.
SOR risk is real. Which approach would you rather take in combatting it – one that’s active and dynamic, or one that sets a course, regardless of conditions?
Links to third-party websites are being provided for informational purposes only. CoreCap is not affiliated with and does not endorse, authorize, or sponsor any of the listed websites or their respective sponsors. CoreCap is not responsible for the content of any third-party website or the collection or use of information regarding any websites users and/or members.