Credit, rates, inflation and liquidity are four separate risks that a single word - safe - is routinely used to dismiss. The tax treatment of interest is the one most holders never price at all.
12 min read · Level 2
Fixed income is the asset class where the account matters more than the selection, and it is the one where investors spend the least time on the account. The reason is structural rather than behavioural: a bond hands you taxable income every period whether you want it or not, so a taxable holder loses the deferral every single year, permanently, on every holding - and no amount of skill in picking issues recovers it. In the illustration above, moving the same 5% return from annual taxation to deferred taxation at the same rate was worth 1,675.20 on 10,000 over twenty years, which is more than most people would ever gain by choosing better bonds. The uncomfortable corollary is that a great deal of the effort that goes into fixed income - comparing yields to two decimal places across similar issues - is spent on the smaller variable, while the larger one is settled once, silently, by whichever account happened to have room when the purchase was made. This does not hold everywhere; several countries tax interest and gains alike, and the question is genuinely national. But it is a question worth asking before the first bond is bought, because the cost of getting it wrong accrues quietly for as long as the holding exists.
Fixed income carries four distinct risks. They have different causes, different remedies and almost nothing in common, and the single word safe is used to wave away all four at once.
Credit risk - the issuer does not pay. This is the one everybody names, and for high-quality bonds it is rarely the one that costs the money. When it does arrive it is not usually total: in the wind-up from the first lesson of this subject, the senior unsecured bondholders recovered 78.95 cents in the dollar in one scenario and 31.58 cents in another, while the shareholders received nothing in both. What you finally get in a default is the recovery rate, and it depends on seniority, on security, and on what the assets actually fetch.
Interest-rate risk - prevailing yields move and your fixed coupon reprices. Nothing about the issuer changes. From the pricing lesson: the same 4% coupon repriced against a two-point rise in yields fell 3.6668% at two years to maturity, 8.4247% at five, 14.7202% at ten and 27.5297% at thirty. This is the risk that surprises holders of impeccable government bonds, precisely because they were watching the credit.
Inflation risk - the coupon and the principal are nominal. A 1,000 par value returned in five years buys, at 3% annual inflation, 1,000 / 1.03^5 = 862.61 of today's goods. That is 13.74% of purchasing power gone, on an instrument that performed exactly as contracted. And the yield itself has to be deflated: a 4% nominal yield against 3% inflation is a real yield of 1.04 / 1.03 - 1 = 0.9709%, not the 1.00% that subtraction suggests. Push inflation to 5% and the same bond delivers 1.04 / 1.05 - 1 = -0.9524% in real terms - a contractual, on-time, guaranteed-by-the-issuer loss of purchasing power.
Liquidity risk - you can sell, but only at somebody's price. Most corporate and municipal bonds trade by dealer quote rather than on a public order book. A bond quoted 97.25 bid and 98.50 offered has a 1.25 point spread, which is 1.2771% of the 97.875 mid. On a 4.50% coupon that is 0.2838 years of income, or about three and a half months of coupon, consumed by a single round trip. And the gap is not constant: it widens exactly when a lot of holders want to sell at the same time, which is when you are most likely to be one of them.
Only one of those four is about whether the borrower pays. A holder who checks the credit and stops has examined a quarter of the problem.
Because the costs and the tax treatment then decide how much of the surviving return actually reaches you - and in fixed income both are unusually easy to overlook.
The costs first. The dealing spread above is the largest and the least visible, because it is inside the price rather than shown as a fee. Nothing on the confirmation says 1.2771%. Fund holders pay an ongoing charge instead, which at least is disclosed, and pay the same dealing spreads indirectly through the fund's own trading. On top of that, individual bonds bought through a broker frequently carry a mark-up embedded in the quoted price. Ask for the bid and the offer on the specific issue; the gap is the number, and asking is the only way to see it.
Now the tax, which is where fixed income differs most sharply from equity, and which is stated here for two named countries because tax rules are national, not universal.
The structural point is that a bond's return arrives mostly as interest, and interest is generally taxed in the year it is received, at whatever rate applies to income. A capital gain, by contrast, is generally taxed only when the asset is sold. That difference in timing is worth money even before any difference in rate.
In the United States, interest on most corporate bonds is taxed as ordinary income; interest on US Treasury securities is subject to federal income tax but exempt from state and local income tax; and interest on most municipal bonds is exempt from federal income tax, and often from the issuing state's income tax for that state's residents. In the United Kingdom, bond interest is taxed as savings income, with a personal savings allowance that shelters an amount depending on the taxpayer's band; conventional UK government gilts are exempt from capital gains tax on disposal, as are corporate bonds that qualify as qualifying corporate bonds, so a gain on price is treated differently from the coupon. Other countries differ in every respect, several tax interest and gains identically, and all of these rules change with legislation. Nothing above is advice about your position; find out how interest is treated where you actually live, from your own tax authority or a qualified adviser, before you decide what a yield is worth to you.
Now the arithmetic, with clearly illustrative rates so it cannot go stale into being wrong. Take 10,000 for 20 years, and a 5% annual return in both cases.
A bond paying 5% interest, taxed each year at an illustrative 30% ordinary-income rate Net annual rate: 5% x (1 - 0.30) = 3.50% 10,000 x 1.035^20 = 19,897.89
The same 5%, arriving instead as an unrealised gain and taxed once at the end at 30% Gross: 10,000 x 1.05^20 = 26,532.98 Gain: 16,532.98 Tax at 30%: 4,959.89 Net: 21,573.08
The same again, taxed once at the end at an illustrative 20% capital-gains rate Tax at 20%: 3,306.60 Net: 23,226.38
Decompose the gap, because it is made of two separate effects that people conflate.
Deferral alone (same 30% rate, paid late instead of annually): 21,573.08 - 19,897.89 = 1,675.20 Lower rate alone (20% instead of 30%, both deferred): 23,226.38 - 21,573.08 = 1,653.30 Total: 23,226.38 - 19,897.89 = 3,328.49, or 16.73% more
The first of those two effects is tax drag in its purest form: nothing was mispriced and no rate was punitive, the money was simply taxed earlier and therefore had less of itself left to compound. Those two effects are within 22 of each other on these figures, which is the point worth taking away. Everybody notices the rate. Almost nobody prices the timing - and here the timing was worth as much as the rate.
The annualised after-tax figures are 3.5000% for the bond and 4.3035% for the deferred, lower-taxed alternative. That is not an argument that equities are better than bonds; they do different jobs and the next lesson is about which. It is an argument that a bond's headline yield overstates what a taxable holder keeps by more than most people assume, and that the account the bond sits in is therefore a first-order decision rather than an administrative one.
This is educational information, not tax or accounting advice. Rules differ by country and change — get advice from a qualified professional in your jurisdiction before acting.
Educational, not personalised financial advice. Nothing in the Investor Academy is a recommendation to buy, sell or hold anything — what is right for you depends on your own circumstances and the rules where you live.
Sources: U.S. SEC - Investor.gov; U.S. IRS - official tax information; UK Government - Capital Gains Tax guidance; FINRA - investor information; OECD - tax policy and statistics
Put the four risks and the two costs on one bond and see what is left.
Bond C from the previous lesson: 5% coupon, seven years remaining, bought at 945, yield to maturity 5.9847%. A taxable holder in a jurisdiction applying an illustrative 30% rate to interest.
Stated yield to maturity 5.9847% Less tax on the interest component at an illustrative 30% (the coupon is 5.2910% of the 945 paid; 30% of that) -1.5873% Approximate after-tax running position 4.3974%
That figure is deliberately approximate, and the reason is instructive: the 55 of pull-to-par is not interest and may be taxed as a gain, at a different rate and only on disposal or maturity, and in some jurisdictions a discount of that kind is taxed as income under specific rules instead. Which of those applies to you is a question with a definite answer in your own country and no general answer at all.
Now deflate it. At 3% inflation, a 4.3974% after-tax nominal return is a real return of 1.043974 / 1.03 - 1 = 1.3567% a year. At 5% inflation it is 1.043974 / 1.05 - 1 = -0.5739%.
And subtract one round trip. If the holder buys and later sells rather than holding to maturity, a 1.2771% dealing spread on a seven-year hold is roughly 0.18 percentage points a year of the remaining 1.36 - about a seventh of it.
So the sequence runs: a 5.98% yield, a 4.40% after-tax yield, a 1.36% real after-tax yield, and something nearer 1.17% if it is traded rather than held. Every one of those subtractions is legitimate, none of them involves anything going wrong, and the issuer paid every penny it contracted to pay.
None of this makes the bond a poor holding. It may be exactly the right instrument for the job it is being asked to do. What it makes indefensible is judging it by the 5.98%.
The first mistake is auditing the credit and calling that risk management. Three of the four risks above have nothing to do with the borrower, and for high-quality bonds they dominate.
The second is subtracting inflation from the yield. The correct operation is division: 1.04 / 1.03 - 1 = 0.9709%, not 1.00%. The gap looks trivial at these levels and grows quickly at higher ones, and the habit matters more than any single figure.
The third is assuming the dealing cost is what the broker charged you. On an individual bond the significant cost is inside the price. If nobody has shown you the bid and the offer on your specific issue, you have not seen it.
The fourth is putting the income-producing assets in the taxable account and the growth assets in the sheltered one. In a jurisdiction that taxes interest annually as income and gains only on disposal, that ordering hands away the deferral shown above, every year, for as long as both are held.
The fifth is treating a high real yield and a high nominal yield as the same news. A 9% coupon in a 12% inflation environment is a contractual loss of purchasing power wearing an attractive number, and the number is the reason people buy it.
Write the four down against the specific holding: credit, rates, inflation, liquidity. A short-dated government bond carries almost no credit risk and real inflation risk. A long-dated one carries enormous rate risk with the same credit. A thinly traded corporate carries all four. The word safe applied to any of them without saying which risk it refers to is not a description.
Divide one plus the nominal yield by one plus your inflation assumption, and subtract one. State the inflation assumption explicitly so it can be argued with. If the answer is negative, that is not a reason never to hold the bond - it may still be the right instrument for a dated liability - but it is something you should have decided knowingly rather than discovered later.
Request both the bid and the offer on the specific issue, take the gap as a percentage of the mid, and convert it into months of coupon. On the quote above that is 1.2771%, or about three and a half months of income at a 4.50% coupon. Doing this once will change which issues you are willing to hold in small size.
Not how gains are taxed - how interest is taxed, in the year it arrives, at your marginal rate. In several countries that treatment is materially worse than the treatment of a capital gain, and in several it is not. This is a question with a specific answer for your country and your circumstances, obtainable from your own tax authority, and it changes the ranking of instruments rather than just adjusting a number.
Where a jurisdiction offers a tax-advantaged wrapper and you hold both bonds and growth assets, the asset producing annually taxed income generally has the stronger claim on the shelter, because it is the one losing the deferral every single year. The rules that make this true or false are national, and the calculation is worth doing rather than assuming.
Maturities have moved a year closer, your own dates may have changed, and a holding that was correctly matched two years ago may now be too short or too long. This is the review that catches the mismatch while there is still time to fix it cheaply, rather than at the moment the money is needed.
Before buying any fixed income holding, once a year across everything you hold, and whenever you are comparing a bond's yield with the expected return on anything else - the comparison is only meaningful after tax and after inflation.
Do not apply any tax figure in this lesson to your own position; every rate here is illustrative and no country's system works this way. And do not let a tax consideration stop you from selling a holding that should be sold - the charge is a cost of a correct decision, not a reason to avoid making one.
In extended periods where inflation runs above prevailing bond yields, holders of high-quality fixed income experience something that has no bad news in it anywhere. Every coupon arrives on time. No issuer defaults. Statements show a stable nominal value and a steady income. And the purchasing power of the whole position falls, year after year, by the gap between the yield and the inflation rate - compounding, in exactly the way returns compound, in the opposite direction. Because nothing appears to go wrong, there is no event that prompts a review, which is why this particular loss is usually noticed in retrospect rather than while it is happening.
A holding can perform perfectly and still cost you. Track the real yield on your fixed income at least annually, and treat a persistently negative one as information rather than as noise.
Create a free account to read the rest and apply it to your own investing.