Funds & ETFs11 min read

BND vs AGG vs BNDX vs SGOV vs TLT: The Best Bond ETFs, Compared

These five are all called bond ETFs. One of them barely moves; another swings like a stock. The label tells you almost nothing.

By Pavel Penev, MScFounder, TradeWize · 10+ years trading the markets

Bond ETFs are where sensible people go to be boring, so it is a little startling to discover how wide the category is. The five funds below are all described as bond ETFs, all hold high-quality debt, and all sit on the same shelf at every broker. One of them moves about as much as a savings account. Another has swung around with roughly the volatility of a stock fund for the last three years. Between those two extremes sits a 150-fold difference in how hard the same interest-rate move hits you — and nothing in the name warns you.

The short answer

BND and AGG are the default core and are functionally the same fund — same 0.03% fee, 4.60% versus 4.62% yield, 5.8 versus 5.73 years of duration. Buy whichever your broker offers free and stop thinking about it. SGOV is where you park money you need soon: 3.57% with almost no rate sensitivity. BNDX adds international bonds, currency-hedged, for diversification rather than income. TLT is the long end — the biggest yield here at 5.10% and by far the roughest ride.

The five, side by side

BND vs AGG vs BNDX vs SGOV vs TLT
FundFeeAnnual yieldDurationHoldingsWhat it's for
BND0.03%4.60%5.80 yrs11,476The default US core
AGG0.03%4.62%5.73 yrs13,288The same job, different badge
BNDX0.07%3.47%6.60 yrs6,808International, currency-hedged
SGOV0.09%3.57%0.10 yrsT-billsMoney you need back soon
TLT0.15%5.10%15.07 yrs46The long end, for better or worse

Figures from each issuer's own fund page in July 2026 and linked at the end. Yields move daily; duration is the number that actually distinguishes these funds, and it moves slowly.

The duration column is the one to read. It is, roughly, how many percent the fund's price falls if interest rates rise by one percent — so SGOV's 0.10 and TLT's 15.07 are not a difference of degree, they are different products with a shared surname. If that number is new to you, we have a whole piece on it; the short version is that duration, not credit quality, is what made 2022 so unpleasant for people who thought they had bought the safe thing.

BND vs AGG: stop trying to rank them

This is the most-asked question in the category and it has a deeply unsatisfying answer. BND tracks the Bloomberg U.S. Aggregate Float Adjusted Index; AGG tracks the Bloomberg U.S. Aggregate Bond Index. Both charge 0.03%. Their yields are 4.60% and 4.62%. Their durations are 5.80 and 5.73 years. BND holds 11,476 bonds and AGG holds 13,288, which sounds like a real gap until you remember that both numbers are enormous and the extra bonds are the same kind of bonds.

BND vs AGG, line by line
BND vs AGG: there is no meaningful differenceBoth track the US investment-grade bond market. Every ranking of these two is noise.BNDAGGAnnual fee0.03%0.03%sameAnnual yield4.60%4.62%sameDuration5.80 yrs5.73 yrssameBonds held11,47613,288Pick whichever your broker offers commission-free. Owning both is owning one twice.

Four numbers, three of them effectively identical. Any article confidently ranking one above the other is generating content, not information.

So the honest tie-breaker is administrative: whichever one trades commission-free at your broker, or whichever you already own. What you should not do is buy both under the impression you have diversified. You have bought the US investment-grade bond market twice and paid two lots of spread for the privilege.

SGOV: the one that refuses to be exciting

SGOV holds Treasury bills maturing within three months, which gives it an effective duration of 0.10 years — mathematically almost a rounding error. Rates can lurch and it barely notices, because everything it owns matures before the news has finished being news. It yields 3.57% at 0.09%, and it has quietly gathered nearly $100bn from people who wanted somewhere to put money that was not a bank account.

The catch is the mirror image of the appeal: if rates fall, SGOV does not rally, it just starts paying less. It is a car park, not an investment — genuinely the right answer for a house deposit two years out, and the wrong answer for money you were hoping would grow.

Learn it by doing

Reading about it is one thing — it clicks when you do it. Practise this hands-on in a free, interactive lesson (Stage 19: Beyond Stocks).

Try the free lesson →

BNDX: the one that isn't there for the yield

BNDX holds international investment-grade bonds with the currency exposure hedged back to dollars, and its numbers look odd until you understand the job. It carries more duration than the US core — 6.60 years against BND's 5.80 — and pays less: 3.47% versus 4.60%. More interest-rate risk for less income is not, on the face of it, a compelling pitch.

It is a diversification holding. Foreign central banks do not move in lockstep with the Federal Reserve, so BNDX's returns do not arrive on the same schedule as BND's, and the currency hedging is what stops the whole thing being swamped by exchange-rate noise. Whether that is worth 0.07% and a lower yield is a real argument with real people on both sides — Vanguard puts it in its own target-date funds, and plenty of sensible investors skip it entirely.

What the extra income actually costs
What the extra income actually costs youAnnual yield3.0%3.5%4.0%4.5%5.0%5.5%0y4y8y12y16yDuration — how far it falls when rates riseSGOVBNDXAGGBNDTLT
Cash-likeThe US coreInternational (hedged)The long end

Yield rises left to right with duration, mostly. SGOV sits alone in the cheap corner — more income than BNDX for a sixty-sixth of the rate sensitivity — while TLT buys its 5.10% with fifteen years of it.

TLT: government bonds that behave like stocks

TLT holds 46 Treasury bonds with more than twenty years left to run. Credit risk is not the concern; the US government is going to pay. Duration is the concern, and at 15.07 years it is roughly three times the core's. Its three-year standard deviation is 13.74%, which is equity-like volatility from an instrument most people file under safe.

Put the two ends together and the trade is stark: TLT pays about 1.5 percentage points more than SGOV, and takes around 150 times the rate sensitivity to do it. That is not an argument against TLT — if you want a position that profits when rates fall, this is the cleanest way to hold one, and it is a legitimate thing to want. It is an argument against buying it because it topped a list sorted by yield.

So which bond ETF should you buy?

Wrong question, slightly. A bond fund is a tool, and the three jobs people want one for pull in different directions. Pick the job first:

Try it: which bond fund is for you?

A bond fund is a tool, and the three jobs want different tools. Say what the money is for and when you need it.

What is this money for?
10 years
The fit
Total US bond market
e.g. BND or AGG
Yield
4.60% / 4.62%
Duration
~5.8 yrs

The default ballast: thousands of investment-grade bonds at 0.03%. It is the boring middle of the market, which is the entire point of ballast.

What it costs you: About 5.8 years of duration, so a 1% rise in rates knocks roughly 5.8% off the price before income. 2022 is what that feels like.

Yields and durations are the issuer figures sourced at the end of this article and move constantly. This matches a job to a category of fund; it is not a recommendation to buy any particular one, and it cannot know your tax situation or anything else about you.

You may not need one yet

Bonds steady the ride and they also lower the destination. If you are decades from needing the money, contributing monthly, and confident you will not sell in a crash, plenty of reasonable investors hold no bonds at all until they are closer to spending the money. The ballast is worth having when a fall in your portfolio would change what you actually do — and if you are honest that a 30% drop would have you reaching for the sell button, that is exactly when it earns its keep.

Is BND or AGG better?

Neither, in any way you could act on. Both charge 0.03%, both track the US investment-grade bond market, their yields are 4.60% and 4.62%, and their durations are 5.80 and 5.73 years. Buy whichever trades commission-free at your broker. The one genuine mistake is buying both and believing you have diversified — you have bought the same market twice.

Why did my bond fund lose money when bonds are supposed to be safe?

Because safe means the issuer pays you back, not that the price holds steady. When interest rates rise, existing bonds paying the old lower rate are worth less, and a fund's price falls roughly in line with its duration. A fund with 5.8 years of duration drops about 5.8% for a one-point rise before income. Nobody has to default for you to lose money, which is what caught people out in 2022.

Is SGOV better than a savings account?

It is a different thing with different trade-offs. SGOV holds Treasury bills, currently yields 3.57%, and its price barely moves — but it is a fund you buy through a broker, its yield floats with short-term rates, and it carries no deposit insurance the way a bank account does in most countries. It is popular for money that is earmarked and near-term. Compare its current yield against what your bank actually pays, not against what banks advertise.

Should I buy TLT if I think rates will fall?

That is the trade TLT expresses most cleanly, and it is also a directional bet on interest rates rather than a ballast holding. Fifteen years of duration cuts both ways: the same maths that delivers a large gain if rates fall delivers a large loss if they rise, and its 13.74% three-year standard deviation is what that looks like in practice. TradeWize does not tell anyone which way rates are going — just be clear which of the two things you are doing.

Do I need international bonds like BNDX?

It is genuinely contested. The case for is diversification: foreign rate cycles are not synchronised with the US, and BNDX hedges the currency so you get the bond exposure without the exchange-rate swings. The case against is that it takes more duration for less yield than the US core, and that the diversification benefit within high-quality bonds is modest. Vanguard includes it in its own target-date funds; plenty of thoughtful investors skip it.

How much of my portfolio should be in bonds?

That depends on when you need the money and how you behave when markets fall, not on a formula. The classic starting point is the 60/40 portfolio, and the honest test is behavioural: bonds earn their place if holding them is what stops you selling your stocks at the bottom. If you are far from needing the money and certain you would sit tight, a lower allocation is defensible.

Duration is the whole game — learn it properly

The number that decides which of these funds suits you gets a full piece of its own, with a dial you can drag to see what a rate move does to a bond fund. TradeWize's free track then covers bonds, portfolio construction and the simulator to practise in. No card required.

Written by

Pavel Penev, MSc

MSc Investment & Finance, Queen Mary University of London · 10+ years trading the markets

Pavel founded TradeWize after years of trading and an MSc in Investment & Finance from Queen Mary University of London. He writes these guides to teach the decisions, not just the theory.

More about TradeWize →

Terms in this article

Keep reading