This is the fifth installment in our series on how individual investors can assess their financial health.
At every stage of investing, you should periodically ask yourself how much risk you can realistically tolerate. The primary way to measure the risk level of your portfolio is to look at its allocation of stocks vs. bonds. Although some stock and bond ETFs are riskier than others, your first decision has to be how much of your investments to put in stocks and how much in bonds.
One standard rule of thumb that’s a good place to start is the “age in bonds” axiom. According to this guideline, you invest a percentage of assets equal to your age in a broad bond index, and the balance of your portfolio in a diversified stock portfolio. The idea here is that your portfolio should become more conservative as you get older. This makes sense for two reasons:
- You tend to get wealthier as you age, so any given percentage loss from your portfolio represents an increasingly larger dollar value.
- You are gradually converting your human capital (your ability to work and earn money) into financial capital (investments) as you age. And as you get older, your financial assets represent a larger and larger fraction of your lifetime wealth potential.
For these reasons, it makes sense to manage this pool of assets more conservatively as time goes by.
Beyond “Age in Bonds” – Choosing Your Allocation of Stocks and Bonds
The past decade provides a powerful example of the tradeoffs between risk and return. The table below shows the year-by-year returns for portfolios comprising different mixes of an S&P 500 ETF (IVV) and a broad bond ETF (AGG). The returns include the expense ratios of the ETFs, but no adjustment is made for brokerage fees.
Source: Author’s calculations and Morningstar
Over the 10-year period from 2004 through 2013, a portfolio that is entirely allocated to the S&P 500 ETF has an average annual return of 9.2%. In its worst year over this period, 2008, this portfolio lost almost 37% of its value. As the percentage of the portfolio allocated to stocks declines, the average return goes down. But the worst 12-month loss also becomes markedly less severe.
We cannot say, with any certainty, that these statistics for the past ten years are representative of what we can expect in the future, but they do provide a reasonable basis for thinking about how much risk might be appropriate.
Ask yourself: If these figures are what you could expect, what allocation of stocks vs. bonds would you choose? Would you be willing to lose 37% in a really bad year to make an average of 9.2% per year? Or would you prefer to sacrifice 1.5% per year to reduce the potential worst-case loss by one third? If so, the 70% stock / 30% bond portfolio provides this tradeoff.
Planning around Improbable Events
One might object that 2008 was an extreme case, and that such a bad year is unlikely to recur with any meaningful probability. One way to correct for this potential bias towards extreme events is to assume that returns from stocks and bonds follow a bell curve distribution, a common way to estimate investment risk. Using the data over the last ten years to estimate the properties of the bell curve (also known as the “normal” or Gaussian distribution), I have estimated the probabilities of various levels of loss over a 12-month period.
Estimated 12-month loss percentiles for a ‘normal’ distribution (Source: author’s calculations)
When you look at the figures for the 5th percentile loss, you can see what might be expected in the worst 5% of 12-month periods for each of the five portfolio types. For example, the 100% stock portfolio has a 1-in-20 chance of returning -21% or worse over the next twelve months. Note that a loss of 35% for stocks, similar to 2008, is estimated to have a probability of 1-in-100.
It’s important to point out that the ability to calculate the probability of very rare events is very poor. Perhaps 2008 really was a 1-in-100 probability event, but we don’t know that with any certainty. The most catastrophic events (what Nassim Taleb has famously dubbed “Black Swans”) are so severe and outside our normal range of experience that they tend to catch us totally off guard.
Moshe Milevsky, a well-known retirement planning expert, suggests that rather than thinking in terms of probabilities, it’s sensible to set your portfolio’s risk to a level that ensures that the worst case outcomes are survivable. Based on that, it’s prudent to choose a portfolio risk level that won’t ruin you if there’s another year like 2008. If you can survive a 12-month loss of 23% (the average of the worst loss for this allocation over the past ten years and the estimated worst-case 1st percentile return), for example, you can afford to hold a 70% equity portfolio.
If your investments in stocks don’t approximate the S&P 500, the stock portion of your portfolio may be considerably riskier than the table above implies. Allocations to emerging markets, small companies, and technology stocks can be very volatile. The examples shown here provide a starting point in determining risk. Combining a wider range of asset classes can provide important diversification benefits beyond their individual risk levels, but this topic is beyond my scope here.
The past ten years have provided examples of very high returns and very low returns from stocks. This period gives us a useful basis for testing our tolerance for volatility. Many readers, I imagine, will find that their risk tolerance—self-diagnosed from looking at the tables above—corresponds reasonably well to the “age in bonds” rule. If your choice of risk levels is too far from these levels, a closer look is needed—and perhaps a talk with an investment advisor.