Why Is My Retirement Portfolio So Bond Heavy as I Age?

Original article posted on "Separating Value From Bias" Substack here: #13 Why Is My Retirement Portfolio So Bond Heavy as I Age?

If you're in your 50s and allocating to a target date fund or general retirement portfolio, you'll notice your portfolio start to have a heavier bond allocation as you near retirement

So if any of you still remember my last post, I spoke about how life insurance and annuity products help maintain the negative stock-bond correlation.

This is important in retirement because a positive correlation would mean that if you have a loss in your stock portfolio, you would also have a loss in your bond portfolio—all while you are making withdrawals from your portfolio.

This creates what’s known as a “sequence of returns” problem where the portfolio is being depleted by both the losses in the portfolio and the retirement withdrawals to such an extent that it can’t recover from them in the future.

In this case, the client is actually withdrawing funds at the portfolio’s lowest point which inhibits future growth.

Life insurance and annuity products help mitigate this sequence of return risk problem in retirement by protecting the client’s bond portfolio from absorbing a loss.

To be fair, they also mitigate this problem by having higher pre-tax yields, lower tax-liabilities, arbitrage benefits from risk-pooling, and allowing more growth in the equity side of the portfolio. But I’ll cover those in future articles.

It’s this sequence of return problem that I want to address in more detail in the context of retirement planning for this post.

But before we get there, I need to do a better job of explaining why bonds are used heavily in retirement in the first place.

Why are Bonds Used Heavily in Retirement?

Most of you are invested in some sort of generic retirement fund that essentially mimic what’s known as a target date fund.

The “target date” here is the approximate year that you plan on retiring. So a target retirement 2040 fund is for someone planning on retiring in 2040 while a target retirement 2060 fund is for someone planning on retiring in 2060.

A target date fund changes the allocation of your portfolio between stocks and bonds dependent on how close you are to retiring.

The farther away you are from retiring, the heavier the stock part of your allocation is.

The closer you are to retiring, the heavier the bond part of your portfolio allocation is.

Let’s take a look at a couple of target date funds from Vanguard and their stock/bond allocation for people with different target retirement dates.

Let’s assume that the people in these funds all plan on retiring at age 65. This helps us to give an approximate age to each person in each target date fund

Table 1: Stock/Bond allocations of Target Date Funds Based on Year of Retirement

So we can see in the table above that the closer the investor gets to retirement age of 65, the more the fund moves away from investing in stocks and more towards investing in bonds.

This is known as a “glidepath”. The portfolio is slowly “gliding” towards a heavier bond allocation as the client nears retirement.

Chart 1: Sample Glidepath of Retirement Portfolio that Increase Bond Allocation of Portfolio as Client Ages

Table 2: Stock and Bond Returns and Risk 1928-2023

In other words, why stop investing in the higher earning asset (stocks) and start to slowly replace it with the lower earning asset (bonds)?

The answer is a result of the fact that stocks tend to be riskier over the short-term than bonds are.

In Table 2 above this is indicated by the higher standard deviation of stocks (19.23%) compared to that of bonds (7.67%).

This means that there is a higher chance of stocks deviating away from its compound annual return than that of its bond counterpart and the magnitude of this deviation is higher.

For example, assume that the return of both stocks and bonds deviate by 2 standard deviations in the negative direction in a given year from their compound annual return.

That would mean that stock return for the year would be -29.13% (=9.33%-2*19.23%) but the bond return would only be -8.87% (=6.47%-2*7.67%).

So while both stock and bond returns are negative in this scenario, the stock return is more than 3 times more negative than that of the bond return.

This becomes more apparent when we look at the historical data and ask what are the chances of stocks and bonds having losses over various time horizons (eg 1 year, 5 years, 10 years, 15 years)

The below table helps us answer that.

Table 3: Chance of Loss in Stock and Bond Investments over Various Time Horizons During the Years of 1928-2023

The above table shows us that there is a larger chance of having a loss in the stock portfolio than there is in the bond portfolio over a given period of time.

Furthermore, what’s not shown in the table above is the magnitude of the loss. Due to the higher standard deviation of the stock portfolio, if a loss is observed in the stock portfolio, almost half the time that loss is greater than 10%. However, there’s only a 12% chance of a loss in the bond portfolio being greater than 10%.

So not only is the frequency of a loss higher with a stock portfolio, but the severity of that loss is also higher.

So by replacing stocks with bonds in a client’s portfolio, you are reducing the portfolio’s chance of having a loss as the client nears retirement as well as the magnitude of that loss.

You might look at the table above and notice that historically if you invested long enough in stocks (i.e. 15 years or more), you have a 0% chance of having a loss in your portfolio.

In other words, even if markets absorb a large loss (like the one we had in 2008) if you give it enough time, historical data shows that you will recover your losses (however, note that applying historical data to future projections has its limitations).

The problem is that when you are nearing retirement you don’t always have enough time to let markets recover before you need to start drawing down on that portfolio to pay for your lifestyle expenses in retirement.

And the worst thing that can happen here is that you have a series of losses just before or as you are retiring and then you start withdrawing funds from your portfolio after you retire to support your lifestyle.

As we mentioned earlier, the reason why this is so bad is because in this scenario you are withdrawing funds while your portfolio is at its lowest which doesn’t leave the remaining portfolio enough funds to fully recover.

Let’s take a more in-depth look at this with a math example that looks at the effects on the portfolio of having negative early year returns first and the impact that taking withdrawals has on the portfolio (versus not taking withdrawals).

For example, let’s say you have $1M and have the choice of either of the following returns over the next 10 years.

Option A has all the positive return years during the first 5 years and all the negative returns during the last 5 years.

Option B’s returns are the opposite of Option A. With Option B all the negative returns are in the first 5 years and all the positive returns come later.

However, the compound annual returns of Option A and Option B are the same after 10 years:

Table 4: Changing the Sequence of Returns from Positive Returns First to Negative Returns First

Since it’s multiplication, multiplying the returns forward or backward doesn’t matter.

You get the same result either way. I’ve illustrated Option A’s annual return calculation below:

Option A: (1.10)*(1.09)*(1.08)*(1.07)*(1.06)*(0.94)*(0.93)*(0.92)*(0.91)*(0.90)=0.9764

Option A: 0.9764^(1/10)-1= -0.33%

We can take these returns along with our starting $1M portfolio value and multiply them out to see what our portfolio value would be at the end of each year as well.

Table 5: Portfolio Values with Different Sequence of Returns without Withdrawals (in Millions)

When we look at both of these paths, we can see that as expected Option A starts off significantly higher than Option B for the first 5 years.

This makes sense since in the first 5 years Option A has all the positive returns and Option B has all the negative returns.

In case you wanted to see an example of the math, I’ve included the first five years for you for both Option A and Option B:

Option A: 1M*(1.10)*(1.09)*(1.08)*(1.07)*(1.06)=1.47M

Option B: 1M*(0.90)*(0.91)*(0.92)*(0.93)*(0.94)=0.66M

At the end of 5 years, the portfolio value of Option A is more than twice that of B ($1.47M vs $0.66M).

However, after that point the portfolio value of Option A starts to decrease while that of Option B starts to increase such that at the end of 10 years the portfolio values of Option A and Option B are identical (both have $970k left at the end of 10 years).

In other words, if you are investing for the long-run, it doesn’t matter if you have negative returns first and then positive returns last or vice versa (provided that those returns have the same compound annual IRR)—as long as you aren’t also taking withdrawals from the portfolio.

The sequence of returns doesn’t matter in this case. Having negative returns followed by positive returns, or vice versa, results in the same outcome.

However, if you are making withdrawals, then the sequence of returns very much matters.

And again, that’s because if you have negative returns first while also making withdrawals you are actively depleting the portfolio such that by the time you get to the positive returns, you might not have enough left in your portfolio to overcome the early losses—especially since you’ll continue to be making those withdrawals in the future as well.

Getting a 20% return is great if I have a $1 million in my portfolio since it means I have a $200k gain. But if my portfolio is depleted and I only have a $100 left in my portfolio, that same 20% gain means I only made $20.

To better visualize this, let’s look at our Option A and Option B returns again.

Only this time, let’s assume that at the end of each year, we are taking an $80,000 withdrawal from each portfolio.

What are the end of year values at the end of each year (after the withdrawal)?

Here’s a quick math example showing the values at the end of the first year for both

Option A and Option B:

Option A: 1M*(1.10) - 80,000=$1.02M

Option B: 1M*(0.90) - 80,000 = $0.82M

The table below shows the values for the subsequent years as well.

Table 6: Effect on Portfolio Value of Changing Sequence of Returns with Withdrawals

In the above table we see that in Option A the portfolio value is barely growing in the early years even though the highest positive returns are in the early years. That’s because the early year portfolio growth is only slightly larger than the withdrawals being taken out. However in the later years, the portfolio starts to deplete as the yearly returns turn negative. At the end of 10 years, the remaining portfolio value is only $330,000.

Option B however shows a different story.

The early year returns of Option B are highly negative. This combined with the large $80,000 withdrawal at the end of each year severely depletes the portfolio. At the end of year 5 the portfolio value is only $310,000. Even though years 6-10 have very high positive returns, it’s not enough to build the portfolio back up—especially since $80,000 is still being withdrawn at the end of the year. At the end of year 5 this $80,000 withdrawal is almost 25% of the entire portfolio value ($310,000). Keep in mind that at the start that $80,000 represented only 8% of the $1M portfolio value the client had. And now that withdrawal percentage is 25%. This is a huge problem because the portfolio is only earning 8%.

The client is withdrawing more than 3 times what the portfolio is earning.

At this point it’s only a matter of time until the portfolio runs out of money. And this happens at the end of year 10.

This is the danger of having a sequence of negative returns early in retirement combined with withdrawals that are a large percentage of the remaining portfolio value. The portfolio ends up in a death spiral that eventually results in it in running out of money.

From a retirement standpoint this means that the client will have completely drained their portfolio after 10 years of drawing income from the portfolio.

This problem is known as sequence of returns risk.

This is really misnamed because as we saw in Tables 4 and 5, there really is no risk if positive returns come first or negative returns come first—as long as there are no withdrawals.

If the client starts taking withdrawals while they are also having negative returns, however, the story changes.

So it really should be called “Withdrawals and early negative returns risk”.

But that’s besides the point.

The point I’m trying to raise here is that the choice to shift the portfolio away from stocks and towards bonds as clients near retirement is to reduce this “withdrawals and early negative return risk”.

Doing so reduces the risk of running out of money and having less after-tax wealth over the long-run.

Of course, if clients never needed to make withdrawals then they would be better off investing in the higher returning stock portfolio for the long-run.

This should hopefully get you asking the question:

“Are there more efficient ways for me to protect myself against this withdrawal and early negative return risk than simply by shifting the portfolio away from stocks and towards bonds”

And the answer to this question is a definitive “yes.”

But that “yes” involves the use of smart financial planning that provides guarantees to the portfolio that reduce this withdrawal and early negative return risk in more effective ways than just shifting the portfolio away from stocks and towards bonds.

And that’s exactly what life insurance and annuity products provide.

I’ll be covering this in more depth in later articles, but at its core life insurance products invest in underlying bonds while providing a “wrapper” around those bonds that provides higher yields, reduced risk and lower tax liabilities that both increase the chance of retirement success as well as the after-tax wealth in retirement as opposed to just investing in bonds directly.