Stocks vs bonds: ownership against lending
A stock makes you an owner of a company. A bond makes you a lender to one. Owners are paid last, out of whatever is left, with no cap on the upside and nothing promised underneath. Lenders are paid on contract, ahead of owners, and no more than the contract says. Volatility, return and the queue all follow from that.
| Stocks | Bonds | |
|---|---|---|
| What you actually hold | A fraction of the company itself, including everything left over once all other claims are settled. | A claim on the issuer for money, on stated terms. You own the debt, not the business behind it. |
| What you are promised | Nothing. A dividend is declared at the board's discretion and can be trimmed or stopped in a bad year. | Interest on set dates and principal at maturity. Missing either is a default, not a management decision. |
| Where you rank if the issuer fails | Last. Common shareholders are paid only after every creditor is made whole, which in a real insolvency often means nothing. | Ahead of shareholders, though ranked among themselves: secured, then senior unsecured, then subordinated. The costs of the insolvency and, in most jurisdictions, certain wage and tax claims come ahead of all of them. |
| Shape of the return | No ceiling. The floor is losing everything you put in, and one holding can multiply many times over or go to nothing. | Close to the yield you bought at, if you hold to maturity and the issuer pays, with the rate coupons are reinvested at the loose end. The floor is the same as equity's, since a default can recover nothing, but recovery is more often partial. |
| Typical volatility | High. Broad equity indexes swing by tens of percent in a year, and market-wide falls of roughly half have happened more than once. | Much lower for short, high-quality bonds, a small fraction of equity's variability. It climbs with maturity and with credit risk, and a long government bond can fall hard in a repricing. |
| What moves the price | Expected future profits, and the rate at which those profits are discounted back to today. | Prevailing interest rates, mostly, plus the market's view of whether the issuer will pay. |
| How inflation lands | Mixed. An unexpected jump usually hurts, though the claim sits on real assets and revenues that can eventually reprice. | Directly, on a fixed coupon: it buys less each year, and the rate rise that tends to accompany inflation marks the price down. Index-linked bonds are built for this case and behave differently. |
| Income | Dividends, discretionary and variable. Many companies pay none, retaining the cash or returning it by buying back shares, and which of those dominates differs by market. | Coupons, contractual and known in advance for a fixed-rate bond. That predictability is most of the appeal. |
| Tax in a taxable account, in the United States | Under current rules, qualified dividends and gains on holdings sold after more than a year are taxed at long-term capital gains rates, and a direct holder chooses when to realise a gain. | Interest is generally taxed as ordinary income in the year it is received. Interest on municipal bonds can be exempt from federal tax. Other countries tax the two quite differently. |
| What each is typically used for | Money that will not be needed for a long time, where the point is to grow faster than prices do. | Money with a date attached, income that is contractual rather than discretionary, or ballast driven by something other than company profits. |
| What it risks | Permanent loss if the business fails, and long stretches of poor returns you have to sit through to collect the average. | Purchasing power lost quietly to inflation, a price fall when rates rise, and default at the weaker end of credit. |
Ownership against lending
A share of stock is a piece of a company. A bond is a loan to one. Every other difference between them falls out of that.
An owner holds no promise. There is no date on which anyone owes you anything, no coupon, and no obligation to pay a dividend in any particular year. What you hold is a claim on whatever is left once suppliers, employees, tax authorities and lenders have been paid, which is why equity is called the residual claim. Residual cuts both ways. In a good decade the leftover is enormous. In a bad one it is nothing.
A lender holds a contract. The issuer owes interest on stated dates and the principal on a stated date, and failing to pay is a default rather than a decision anyone gets to make. That promise is also what caps the return. However well the company does, the bondholder still receives the coupon and the principal back and no more. The upside was sold in exchange for the certainty.
So these are not two flavours of the same product. They are opposite sides of the same balance sheet. Issuing stock sells a share of the future. Issuing a bond creates a liability the company must service whatever the future turns out to be.
The order of the queue
The sharpest way to see the difference is to ask what happens when the issuer runs out of money.
Claims are settled in order. The costs of the insolvency itself come off the top, and in most jurisdictions certain employee and tax claims rank high. Then secured creditors, who have a claim on the specific assets pledged to them. Then unsecured creditors, which includes most corporate bondholders, ranked among themselves: senior bonds ahead of subordinated ones. Preferred shares sit below all of that. Common shareholders come last, and the residual they are entitled to is frequently zero. In the United States this ordering is known as the absolute priority rule, and while real reorganisations are negotiated and sometimes depart from it, the ranking is the starting point. The detail differs by jurisdiction; the direction does not.
Two consequences matter, and the first is often stated wrongly. Both claims can end at zero, so it is not the case that a bondholder is protected from total loss. What the bondholder has is position: paid first out of whatever is recovered, with equity entitled to nothing until every creditor is whole. Senior secured debt has historically recovered a good part of its value, while subordinated unsecured debt in a bad insolvency can recover close to none, so the difference is in the distribution of outcomes rather than in the worst one.
Second, the two claims disagree about risk. A shareholder gains when the firm takes a bet that might pay off spectacularly, because the upside belongs to equity and the downside stops at zero. A bondholder receives none of that upside and carries the added chance of not being repaid, which is why bond contracts carry covenants and why lenders care how much debt already sits ahead of them.
Government bonds shift the picture, though not always in the direction the label suggests. A government borrowing in a currency it issues can always produce the currency, so its risk moves from default towards inflation and the exchange rate. A government borrowing in a currency it does not issue, including a member of a currency union, sits closer to a company: defaults and restructurings on that kind of debt are not rare, and holders have taken deep write-downs within living memory. Government bond is a label, not a risk category.
Volatility and return, and what the averages hide
Over long stretches of United States history, stocks have returned more than government bonds by something on the order of a few percentage points a year. That gap is the risk premium, the payment for holding the claim that is settled last and moves the most. Other long-running markets show a premium too, smaller in several of them, so the size of the gap depends on which country and which century you measure.
The volatility earning it is not an abstraction. Broad equity indexes have annual variability several times that of high-quality bonds, and market-wide drawdowns of roughly half have occurred more than once inside a single investing lifetime.
The broad United States market has recovered from each of those, which is the fact usually quoted, and it is worth being careful with. It is one country's record, and it is the country whose market did best over the period being measured. Japan's main index took more than three decades to regain its 1989 peak. A few markets have been closed to their investors outright and returned nothing at all. Where recovery came it still took years, and money with a date attached does not get to wait.
Two things hide inside the average. One is that the premium is measured across a very long sample, and shorter windows have delivered nothing: there have been stretches longer than a decade in which stocks trailed bonds outright. The other is that bonds are not one risk. A short government bill barely moves. A thirty-year government bond has a long duration, meaning its price is highly sensitive to rates, and can lose a fifth of its value or more in a year when rates rise sharply. A high-yield corporate bond behaves in a downturn a great deal like the stock of the same issuer.
What counts at the end is the real rate of return, after inflation, which the real return calculator works out for any pair of nominal and inflation rates.
What a mix changes, and what it does not
If stocks return more over long periods, bonds can look like a drag on the total. They are not a second copy of the same holding, though, because the two are driven by different things, and that is what makes the arithmetic of a mix different from the arithmetic of its parts.
An equity price mostly reflects expected profits. A high-quality bond price mostly reflects interest rates. Those inputs are related, since the rate used to discount future profits is built on the same rates that price the bond, but they are not the same input, so the two do not move together reliably. A portfolio's variability depends on how its parts move relative to each other, not on the average of their separate variabilities. That is the mechanism behind diversification, and it is why a mix can be steadier than the weighted average of its parts.
The limit deserves saying plainly. The correlation between stocks and bonds is not a constant. Through much of the low-inflation decades it was negative, which made government bonds an effective shock absorber. In an inflation shock the two fall together, because the same rise in rates marks bonds down and compresses what investors will pay for future profits. Anyone who treated that negative correlation as a law has since learned otherwise.
The second point is not statistical. A mix decides what a holder can sit through in a bad year, and a portfolio sold at the bottom earns only what it earned up to that moment. Bonds also attach money to dates in a way an expected return cannot: a maturity is a date, while an expected return is an average that may or may not have arrived by the date that matters. What proportion of each fits a particular person depends on the horizon, on what the money is for, and on the variability that person will actually sit through.
Common questions
Are bonds safer than stocks?
Safer against some risks and not others, so the question needs a horizon and a definition of safe. Short, high-quality bonds are far safer against a large fall over a short period, and one held to maturity has a known nominal outcome provided the issuer pays. They are worse against inflation, since a fixed coupon buys less each year, and long-dated bonds can fall hard when rates rise. Credit quality matters as much as the label: a bond from a weak issuer can behave like equity, and it tends to do so in exactly the periods it was being counted on not to.
If stocks beat bonds over long periods, why hold bonds at all?
Because the long period is longer than most goals, and the average conceals the path. Stretches longer than a decade in which stocks trailed bonds have happened, and the deep falls arrive without notice, so money needed on a date has to be somewhere it will still be on that date. The long-run figures also come mostly from the markets that did best, which flatters them. There is a behavioural half too: a holding sold in a panic pays only what it had earned by then, so the mix someone can sit through matters alongside the mix that looks best on a long chart.
Does a bond fund behave like an individual bond?
Not exactly, and the gap catches people out. An individual bond has a maturity date, so holding it returns the principal on that date if the issuer pays, whatever the price did in between. A typical bond fund holds a rolling range of maturities and never matures. When rates rise its price falls, though the fund is then reinvesting at the new higher yields, so a holder who stays put is repaid through income over roughly the fund's duration. The two are not the same promise, and the difference matters most for money with a specific date attached.
Keep reading
This page is educational material, not financial advice. The figures come from the formula shown and assume the inputs you enter hold for the whole term. Your own rate, fees, taxes and timing will differ, so treat the output as arithmetic to check a decision against, not as a recommendation.