ZAG vs VAB vs XBB 2026: Which Canadian Bond ETF Should You Actually Own?

ZAG, VAB, and XBB each hold roughly 1,000 to 1,500 Canadian government and investment-grade corporate bonds, charge almost identical fees, and move in near-perfect lockstep with the Government of Canada yield curve. So which one wins for 2026? The honest answer is "whichever one your other ETFs already use" - but there is one real differentiator (tax-loss harvesting) and one real tie-breaker (liquidity) worth knowing before you buy.

What each ETF actually holds

ZAG (BMO Aggregate Bond Index ETF) tracks the FTSE Canada Universe Bond Index, which is the standard Canadian investment-grade benchmark: roughly 1,500 bonds covering federal government debt (~35%), provincial government debt (~35%), and investment-grade corporate debt (~30%). All maturities above one year are included, so the fund's duration lands around 7.4 years. Trades on the TSX in CAD, pays distributions monthly, launched 2010, and is the largest Canadian bond ETF by assets at roughly $12 billion.

VAB (Vanguard Canadian Aggregate Bond Index ETF) tracks the Bloomberg Global Aggregate Canadian Float Adjusted Bond Index. In practice the holdings are 95%+ the same as ZAG - same mix of federal, provincial, and investment-grade corporate issuers, same roughly 7.4-year duration - but the index methodology is different. Bloomberg adjusts for float (bonds actually available for trading), which gives VAB a very slightly lower weight to tightly-held issues. Trades on the TSX in CAD, pays monthly, launched 2011, roughly $7 billion in AUM.

XBB (iShares Core Canadian Universe Bond Index ETF) tracks the same FTSE Canada Universe Bond Index as ZAG. Launched in 2000, XBB is the elder statesman of the three: it was the first broad Canadian bond ETF on the TSX, and it still carries the oldest cost structure at 0.10% MER (one basis point above its newer rivals). Monthly distributions, around $6 billion in AUM.

Head-to-head: fees, structure, and yield

The yield gap between the three is almost entirely driven by the one-basis-point MER drag on XBB, not by any real difference in the underlying bonds. On a $100,000 bond allocation held for 20 years, the 1 bp difference compounds to roughly $200 to $300 total - well inside the noise of monthly distribution timing and tracking error. Compared to the typical bank-run bond mutual fund at 1.0%+ MER, every one of these three saves you more than 90 bps per year, which on the same $100K position compounds to over $20,000 across two decades.

Index methodology: where ZAG, VAB, and XBB actually differ

ZAG and XBB share the FTSE Canada Universe Bond Index, which includes every CAD-denominated investment-grade bond issued by Canadian federal, provincial, and municipal governments plus Canadian investment-grade corporations. Minimum issue size is CAD $100 million, minimum credit rating is BBB, and all maturities above one year qualify. Monthly rebalance, no float adjustment.

VAB's Bloomberg Global Aggregate Canadian Float Adjusted Bond Index applies the same investment-grade and maturity screens but reweights by the portion of each issue actually available to trade in the secondary market. In practice this cuts the weight on tightly-held provincial issues by a few basis points and lifts the weight on actively-traded federal and corporate issues. The result is a portfolio that is marginally easier for Vanguard to replicate without tracking error, which is why VAB has consistently matched its index to within 2-3 bps per year.

Both methodologies arrive at the same 7.4-year duration and the same 70/30 government/corporate split. The difference matters for only two things: tracking error (negligible, favors VAB) and tax-loss harvesting pair eligibility (meaningful, see below).

The VAB vs ZAG tax-loss harvesting trick

Bonds have had two bruising calendar years in the last decade (2013 taper tantrum, 2022 inflation shock), each producing double-digit paper losses in a non-registered account. If you held your bond sleeve in a taxable account during 2022, you could have crystallized those losses to offset capital gains elsewhere - but only if you had a legitimate alternative to park the proceeds in during the 30-day superficial loss window.

Here is where the index methodology matters: ZAG tracks FTSE Canada Universe, VAB tracks Bloomberg Canadian Aggregate. Different index providers, different methodologies, different issuers (BMO vs Vanguard) - the CRA generally accepts these as not "identical property" under the superficial loss rules. XBB tracks the same FTSE Canada Universe index as ZAG, so an XBB-for-ZAG swap carries real superficial-loss risk even though the issuers differ. The clean tax-loss harvesting pairs are:

• ZAG and VAB - different index, different issuer. The textbook pair.
• XBB and VAB - different index, different issuer. Also clean.
• ZAG and XBB - same index. Avoid for superficial loss purposes.

If you expect to actively tax-loss harvest your bond sleeve in a non-registered account, hold the pair you plan to swap between (typically ZAG and VAB) from day one so the sell-and-buy mechanics work cleanly whenever bonds dip.

Account placement playbook

• RRSP: First choice for any of the three. Bond interest is fully taxable at your marginal rate, and the RRSP shelters it entirely while you compound. Pick whichever matches your equity issuer.
• TFSA: Fine if you have run out of RRSP room, but recognize that bonds inside a TFSA burn your highest-value tax shelter on your lowest-returning asset class. Most planners prioritize equities in TFSA and bonds in RRSP when both have room.
• FHSA: Any of the three work for the short-fuse savings portion of an FHSA, but given the 15-year lifetime limit, consider shorter-duration bond ETFs (ZSB, VSB) if your home purchase is within 3-5 years.
• Non-registered: ZAG, VAB, or XBB is a last resort - the monthly interest distributions are fully taxable. Prefer a discount bond ETF (ZDB, HBB) that minimizes distributions and compounds returns inside the fund as capital gains, which are only triggered on sale.
• RESP, LIRA, RRIF: All three are fully eligible. For a RRIF specifically, the monthly distribution calendar helps smooth the required minimum withdrawals.

When each one wins

What about ZDB or HBB for taxable accounts?

If you are forced to hold bonds in a non-registered account (room already filled in RRSP and TFSA), ZAG, VAB, and XBB all have a tax drag problem. The monthly interest distributions are 100% taxable at your marginal rate - no dividend tax credit, no capital gains inclusion rate. In a top Ontario or Quebec bracket that is 50%+ of the distribution lost to tax every year.

ZDB (BMO Discount Bond Index ETF) solves this by targeting bonds trading at a discount to par - most of the return comes from price appreciation to par at maturity rather than from coupon payments. Over a full market cycle ZDB delivers a similar total return to ZAG but with roughly half the taxable distribution yield. The MER is 0.10%, one basis point above ZAG, which is a tiny price for the tax deferral.

HBB (Global X Canadian Select Universe Bond Index ETF) goes further by using a total-return swap structure. HBB pays zero annual distributions - all return is deferred and triggered only when you sell, at which point it is treated as a capital gain (50% inclusion rate). The MER is 0.09%, same as ZAG, but counterparty risk via the swap counterparty is real and the structure can be unwound by the issuer if regulations change.

How does this fit into your target weighting?

A typical Canadian investor's bond allocation ranges from 0% (VEQT-style all equity) to 40% (VBAL-style balanced) depending on age and risk tolerance. The right weight is less about picking ZAG vs VAB and more about picking the right risk level in the first place. Once you have set the bond percentage, any of these three gets you there. Our rebalancer tool flags when your bond weight drifts more than 5% from target and shows you exactly how much of your next contribution should flow to bonds to pull you back without triggering a taxable sale.

Frequently asked questions