Are Retirees Putting Too Much Into Bonds?

For decades, retirement investing has followed the same conventional wisdom.

When you are young, you can afford to take risk, so you own more stocks. As retirement approaches, you gradually move money into bonds. Once you retire, preserving what you have becomes more important than growing it.

There is a lot of common sense behind that approach. If the stock market falls 40% while you are still working, you may have years to recover, and you'll may even emerge better-than-ever if you keep contributing to your investments throughout that downturn. But if it falls 40% just after you retire and you are simultaneously withdrawing money to live on, the consequences can be much more serious.

But a recent investment paper raises an interesting question:

What if retirees are actually becoming too conservative with their asset allocations?

In Beyond the Status Quo: A Critical Assessment of Lifecycle Investment Advice, researchers Aizhan Anarkulova, Scott Cederburg and Michael O'Doherty challenge two of the most established ideas in retirement investing: that investors should diversify between stocks and bonds, and that they should steadily reduce their stock exposure as they age. The most recent SSRN version of the paper was revised in July 2025.

Their conclusion is surprisingly aggressive.

Using a broad historical dataset covering 39 developed countries, the researchers find that their optimal long-term portfolio for investors around the world is approximately:

33% domestic stocks

67% international stocks

0% bonds

0% Treasury bills

The precise domestic-versus-international mix is not the most important part of the finding, particularly now when the US represents something like 2/3s of global market capitalization.

Their larger conclusion is that, when investors have access to broad international diversification and are investing over very long periods, bonds contribute much less to retirement success than conventional retirement advice assumes.

Retirement Is Still a Long-Term Investment Problem

One reason retirees naturally become more conservative is that retirement feels like the end. Mission accomplished!

But financially, the journey has only just begun.

Someone retiring at 65 may need their portfolio to continue supporting them into their 90s. 

That means there are two competing risks.

The first is one most people are aware of: owning too many stocks and suffering a major market decline at the wrong time.

The second receives less attention: holding too little in growth assets and slowly losing financial flexibility over decades.

We tend to associate safety with stability. If one portfolio fluctuates less than another, it feels safer.

But lower volatility and greater financial safety are not necessarily the same thing when you take your whole life into account.

A portfolio that earns a higher return over time can build a larger cushion. That additional wealth can help absorb future market declines, inflation, unexpectedly high expenses, a longer-than-expected life, or increased spending later in retirement.

In other words, growth itself can provide a form of protection.

The researchers found this effect to be substantial.

In their simulations, a retired household following a traditional 4% withdrawal strategy exhausted its investment portfolio 7.0% of the time when using the globally diversified all-equity strategy.

The corresponding failure rate was 16.9% for a traditional 60% stock / 40% bond portfolio and 19.7% for the representative target-date fund they studied.

How's that for counterintuitive?

The portfolio that looked riskiest from the standpoint of short-term volatility was considerably less likely to run out of money.

The authors are careful about what that means. They do not argue that an all-stock portfolio is safe. Stocks can suffer enormous declines. Their argument is that the alternatives we normally label "safe" may carry other, less visible, risks. Life is risky, and there are no guarantees.

What About a Market Crash Right After Retirement?

This is the strongest argument for owning fixed income in retirement.

Imagine retiring with $1 million and withdrawing $40,000 per year.

If the market immediately falls 40%, you do not merely have to wait for stocks to recover. You are also selling investments during the decline to pay your bills. Those withdrawals leave less money invested as the recovery happens.

This is known as sequence-of-returns risk.

Interestingly, the researchers' model recognizes the problem.

When they allow the optimal portfolio to change with age, it remains almost entirely invested in stocks throughout life, with one notable exception.

At age 65, the optimal allocation falls to approximately 73% stocks and 27% Treasury bills.

But the shift is temporary.

By age 68, the stock allocation is already above 90% again. The researchers specifically identify the temporary Treasury bill position as protection against the sequence risk created by taking fixed, inflation-adjusted withdrawals from the portfolio.

Even more interestingly, when they model retirement spending as 4% of the portfolio's current value rather than requiring the same inflation-adjusted dollar withdrawal every year, the temporary fixed-income allocation disappears.

The more powerful tool for addressing sequence of return risk isn't bonds, it's the ability to decrease spending temporarily when your portfolio is down. And when you're willing to make small spending cuts, the optimal portfolio remains all equity throughout retirement.

But I would add some nuance here. This doesn't doesn't mean retirees should blindly invest everything in stocks. Rather the appropriate investment portfolio depends partly on how flexible the rest of the retirement plan is.

Your Social Security Is Part of Your Retirement Portfolio, Too

The households in this study don't live entirely from their investment accounts. They also receive Social Security. This is an important detail.

Economically, Social Security has many characteristics we are trying to create when we add fixed income to a retirement plan.

It provides dependable income, lasts for life, and adjusts with inflation.

You can't sell your future Social Security benefits to pay for a new roof, so they are not technically interchangeable with bonds or cash. But from the standpoint of supporting ordinary retirement spending, Social Security functions a lot like an inflation-adjusted income-producing asset that sits outside your investment accounts.

This is important for thinking through your asset allocation.

Consider two hypothetical retired couples:

Both spend $80,000 per year and both have $1.5 million invested.

The first couple receives $60,000 per year from Social Security and pensions. They only need about $20,000 from their investment portfolio.

The second receives $25,000 of guaranteed income and needs approximately $55,000 from the portfolio.

They have identical investment balances.

But they do not have identical capacity to tolerate investment risk.

The first household already has most of its basic lifestyle supported by dependable lifetime income. Their investment portfolio has considerably more freedom to fluctuate.

For the second household, portfolio withdrawals represent much more of their budget. A severe market decline could have a much greater impact on their ability to maintain spending.

In other words, you are in a very different situation when only 10% of your retirement income depends on the market, vs 50%. For the first investor, market fluctuations and accompanying spending cuts change almost nothing about their day-to-day lives. But the second household is much more sensitive to this volatility.

This also gives some context to the paper's failure statistics.

Also, when the researchers say that a household "runs out of money," they mean the investment portfolio reaches zero before the last surviving spouse dies. Social Security does not disappear. Those households continue to receive Social Security. In real life, a couple may also have a house which they can sell or borrow against to help cover long term care and other expenses near the end of life.

The Price of Being Too Conservative

The question with bonds (and any other asset class) is what you are giving up and what you are receiving in return.

Suppose two retirees begin with similar resources. One maintains a relatively aggressive portfolio and experiences stronger long-term growth. The other keeps considerably more in cash and bonds.

The second retiree may enjoy a smoother ride.

But after 10 or 15 years, the difference in accumulated wealth can be substantial.

That extra wealth isn't merely an impressive account balance.

It can mean:

• More ability to travel or help children and grandchildren.

• More room for unexpected home or health expenses.

• More capacity to withstand the next market downturn.

• More flexibility to spend without constantly worrying about whether the portfolio will last.

This is why I think retirement investing becomes distorted when we define risk exclusively as "How much can my account fall?"

A 30% market decline is a risk.

But reaching age 80 with considerably less purchasing power than you otherwise could have had is a risk too. 

A portfolio designed primarily to prevent short-term losses has not grown enough to comfortably support the life you want. This risk is much harder to reckon with and correct. 

But There Is a Problem With the "Optimal" Portfolio

There is an enormous difference between finding an optimal portfolio in a research model and actually owning it.

Imagine retiring with $2 million.

You invest almost all of it in stocks.

Then a major bear market arrives and your account falls to $1.3 million.

The research paper can tell you to stay invested, and intellectually you may understand that markets offer some of their strongest returns when recovering from a drawdown. 

But you do not experience life emotionally through a spreadsheet, research paper, or analyst projections. And what you do with those feelings will have far more to do with your optimal portfolio strategy than any theoretical knowledge you may have. 

An otherwise optimal strategy that causes someone to panic and sell during a market crash is not optimal for that person.

The authors acknowledge this problem themselves. They note that large stock-market drawdowns can cause considerable psychological pain and could lead investors to abandon the strategy. You are always making some kind of trade-off.

That is why I don't think the appropriate takeaway from this paper is that retirees should own 100% stocks.

Putting myself in a retiree's shoes, I don't think that's an approach I would be comfortable with. 

I would want enough money outside the stock market so that I could continue living relatively normally through a serious downturn and not be forced to sell large amounts of stock. I don't want to be on an international travel spree one year, and then spend the next year eating ramen at home. 

For me, keeping roughly two years of portfolio spending needs in cash takes the worst of the edge off. 

That is still aggressive by conventional retirement standards, and it would not carry someone through every bear market, but it gives a small security blanket during major crashes without creating too much drag on returns. So that's how I personally would balance that tradeoff with my own money.

After all, holding five, seven or ten years of spending leaves less money participating in long-term growth, and that creates long-term uncertainty. 

There is no allocation that eliminates both risks.

Beliefs Matter Too

There is another useful lesson buried in the research.

The authors' base-case portfolio allocates about two-thirds of its equity exposure internationally. That may look extreme to many American investors.

The researchers specifically examine that objection.

When they change the model based on an investor's belief that the United States will continue to have unusually favorable returns, the optimal allocation shifts toward U.S. stocks.

At a 50% belief that U.S. stock returns will remain exceptional vs the rest of the world, the model produces approximately 60% U.S. stocks and 40% international stocks. At complete conviction/certainty, it chooses 100% U.S. stocks. 

I don't think 100% certainty really exists. But I do notice that the US has an outsized presence in markets, trade, and global affairs. The US economy is much larger and more diverse than many other countries, and so I think the idea of allocating 2/3 of capital abroad seems inverted. Some international exposure makes sense, for reasons of lower correlations, diversification, and various tail-risks. But when you overdo it, you give up more than you gain.  

Those questions cannot be answered entirely by historical data. We just don't know what the future will look like. But if you're going to stick with the process, you need to build your approach around what you believe in.  

A Better Question Than "How Much Should I Have in Bonds?"

I think this paper ultimately raises a better question than whether a retiree should own 20%, 40% or 60% in bonds.

The more useful questions are:

• How much of my spending is already covered by Social Security, pensions or other dependable income?

• How much actually needs to come from my portfolio each year?

• How much money do I need outside the stock market to avoid selling stocks during a severe decline?

• How flexible could my spending be if markets performed poorly?

• How much volatility can I realistically live with without abandoning the strategy?

• And how much long-term growth am I sacrificing in exchange for greater stability today?

For a retiree whose Social Security and pension cover most essential expenses, the answers may support a surprisingly high allocation to stocks.

For someone who is heavily dependent on portfolio withdrawals, has little spending flexibility, or simply knows that a 40% stock-market decline would be intolerable, more fixed income may be entirely appropriate.

Age alone doesn't tell us the answer. The broader retirement income plan does.

Though I think it is a well-researched piece, this paper hasn't convinced me that retirees should put every dollar into stocks.

It has, however, made me much less comfortable with the assumption that moving steadily into bonds automatically makes retirement safer.

Yes, a retirement portfolio has to survive bad markets.

But it also has to support decades of spending, keep pace with inflation, provide room for the unexpected and, ideally, give retirees greater freedom.

The real work is figuring out how much of each form of safety you actually need.