What Role Should Bonds Play in a Portfolio?

What Ramesh is actually buying

Ramesh has been told to hold more in bonds as he approaches retirement, and the reason given was that bonds reduce risk. That is true, and it is too vague to act on — it does not tell him which bonds, how many, or what he should expect them to do at the moment it matters.

The sharper version is that bonds do three separable things, and which one he needs determines what he should own.

The first and most important is being spendable money during a bad market. If Ramesh has to withdraw during a fall, selling equities at depressed prices does permanent damage to a portfolio that no longer has decades in which to recover. Bonds are the pot he draws from instead, leaving the equities alone to recover. This job requires the bonds to hold their value at precisely the moment equities do not, which turns out to be a demanding requirement rather than an automatic one.

The second is reducing how much the portfolio moves overall — the conventional justification, and it matters mainly because a smoother path is one an investor is more likely to stay invested in.

The third is providing a source of return, which is real and the least of the three for anybody with a long horizon and no withdrawals. Over long periods equities have generally out-returned bonds, so holding a great many of them costs expected return.

Ranking those three against his own situation gives Ramesh his allocation, and for somebody four years from drawing an income the first one dominates the other two.

Why quality matters more than yield here

If the first job is the point, then what matters is how the bonds behave during an equity crisis rather than what they yield in ordinary conditions.

Government bonds tend to hold up or gain when equities fall in a growth scare, because money moves toward the safest available asset and rates often fall at the same time. That is exactly the behaviour Ramesh is buying.

Lower-quality corporate credit does not reliably do this. Weaker borrowers become more likely to default in the same conditions that hurt equities, so their prices tend to fall alongside them. Holding those as ballast means owning something that fails at its job precisely when the job comes up, which is the most expensive way to be wrong about an asset class.

This is the argument for keeping the defensive part of a portfolio in high-quality bonds even though they yield less. You are not buying yield; you are buying a specific behaviour in a specific scenario, and the extra yield available on weaker credit is the payment for giving that behaviour up.

How much

There is no universal number, and anybody offering one without knowing your situation is guessing. Four things actually decide it.

Whether you are withdrawing is the largest factor by some distance. Somebody drawing an income needs enough in bonds and cash to cover several years of withdrawals, so that no fall forces an equity sale. Somebody still contributing needs far less, because their contributions are buying during the fall rather than selling into it — which is the clearest single difference between Ramesh now and Ramesh fifteen years ago.

Your horizon matters, since a longer one can carry more equity because there is time for a fall to recover. Whether you will actually hold on matters just as much, because a portfolio abandoned in a crash returns less than a milder one kept throughout; if a large equity allocation would cause you to sell at the bottom, the theoretically optimal allocation is not optimal for you. And what else you have counts, since a secure pension, rental income or other stable resources can substitute for bonds in the role of spendable income.

Age-based rules of thumb exist and are widely quoted. They are a starting point at best, because they ignore whether you are drawing income, what else you own and how you behave — all of which matter more than your age does.

What bonds do not do

Two things are worth being blunt about, because the shorthand misleads.

They are not risk-free. A bond fund can fall, sometimes substantially, when rates rise, and long government bonds can move a great deal. "Defensive" describes their role in a portfolio rather than an absence of movement, and Ramesh will see red numbers on the bond side of his statement at some point.

And they do not always rise when equities fall. The negative relationship is a tendency rather than a law, and it is weakest in exactly the scenario people worry about most — an inflation shock, where rising rates hurt bonds and equities at the same time. A portfolio built on the assumption that bonds always cushion equities contains an untested assumption inside it.

That is not an argument against holding them. It is an argument for holding cash as well, since cash is the only asset that reliably keeps its nominal value in every scenario, and for not being surprised on the day both fall together.

Structuring the defensive part

Match duration to when the money is needed, as the horizon article sets out. Ballast for a long-term portfolio can carry duration; money that might be spent within a year should not.

Keep credit quality high in the part meant to be defensive, and take risk in equities where you are paid for it with unlimited upside rather than in credit where the upside is capped at getting your money back. Hold some genuine cash alongside the bonds, because cash is the asset that does not fall in the one scenario where bonds and equities fall together.

And rebalance, which is where the defensive allocation earns its keep beyond mere safety. After equities fall, restoring the intended mix means buying equities cheaply using the part that held up. That mechanism only works if the defensive part actually held up, which brings the argument back to quality.

What to take away

The main job of bonds is not to smooth a line on a chart. It is to be the money you can spend when equities are down, so that you are never a forced seller of the asset that needs time.

That job dictates the design: high credit quality, duration matched to when you might need it, and some cash alongside for the case where both fall together. How much depends far more on whether you are withdrawing, and on whether you would genuinely hold your equities through a crash, than on how old you happen to be.