Agree Realty

Agree Realty does what every net-lease landlord does. It buys a building, leases it to the retailer who trades in it, and the retailer pays the rent, the property taxes, the insurance and the maintenance. Agree's job is to own the building and bank the cheque.

We described that model at length in this morning's analysis of NNN REIT and will not repeat it. What matters here is the one decision that separates the two companies, and it is a decision about credit.

Here is what that looks like when you open the buildings, set against the company we published this morning.

Two companies doing the same thing to two different kinds of tenant. Agree's rent comes from grocers, home-improvement chains and Walmart. NNN's comes from car washes, tyre shops and convenience stores. Agree's tenants have credit ratings; NNN's mostly do not. Agree's leases are two and a half years shorter. And Agree costs a quarter more per dollar of cash flow.

The rest of this analysis is about whether that is the right price.

One further structural difference is worth stating early, because it is where Agree's growth actually comes from. Most net-lease REITs have a single source of new buildings: they buy them. Agree has three.

Total investment in the second quarter was a record $502 million across 102 properties, and full-year guidance was raised from $1.4–1.6bn to $1.6–1.8 billion. For scale: NNN's guidance for the same year is $750 million, from a portfolio a third larger by count.

We set this out in full this morning, so here it is in short form — because the numbers on Agree's screen are, if anything, more alarming.

Agree grew revenue nearly eight-fold in nine years and its earnings per share went backwards. Anyone reading the accounting line would conclude the company destroyed value for a decade. On the measure the industry actually uses, cash flow per share is growing at close to eight per cent. The gap is depreciation, and it is the reason we say the same thing in every REIT analysis we write.

One honest divergence we will not hide. Adjusted funds from operations per share is growing 7.7%. But operating cash flow per share, taken straight from the cash flow statement, was $4.58 in 2022 and $4.53 in 2025 — essentially flat for three years while the portfolio grew enormously. AFFO is the industry standard and it is what management, analysts and we use. But the two measures disagree, and when two cash measures of the same company diverge for three years, an investor should know it. The likely explanations are straight-line rent accounting and working capital timing; we flag it rather than resolve it.

Here is the comparison in full. We have published NNN today; Realty Income was published in June. All three are triple-net landlords, and they are priced very differently.

★★ Read the "yield plus growth" row and the naive conclusion inverts. A landlord's return to a shareholder is, roughly, the dividend you collect plus the rate at which it grows. Agree gives you 4.30% growing at 7.7%. NNN gives you 5.35% growing at 3.8%. Twelve per cent against nine. The company with the higher multiple offers the higher expected return, and it is not close.

This is what a multiple is for. It is not a price tag; it is a statement about what the market expects to happen next. The market charges a quarter more for Agree because Agree is deploying $1.8 billion a year into buildings occupied by Walmart, at a 7.0% initial yield, and compounding the result. It charges less for NNN because NNN deploys $750 million and grows at three per cent — by choice, and for reasons we thought were sound this morning and still do.

We are not going to leave that argument unqualified, because the qualification is important and it is structural.

Agree's growth is funded by selling shares. In the first half of 2026 it raised $686 million of forward equity through its at-the-market programme, and has $1.1 billion outstanding. Its reported leverage is 5.2× net debt to recurring EBITDA and only becomes 3.7× once that equity is actually issued.

Nothing about that is improper — pre-selling equity forward at a known price is prudent, and Agree does it well. But it means the engine has a dependency that NNN's does not. If Agree's share price falls far enough, issuing equity becomes dilutive rather than accretive, and the $1.8 billion of annual deployment slows or stops. NNN can fund roughly $550 million a year from rent it has already collected, whatever its share price does.

★ So the honest formulation is this: Agree offers a materially higher expected return, and that return is more contingent. NNN offers a lower one that is more certain. Those are two defensible answers to two different questions, and an investor should know which question they are asking before choosing.

The pro forma leverage figure deserves a moment's scepticism, and then acceptance. A company reporting 5.2× and asking you to think of it as 3.7× is asking for credit for money it has not yet received. In Agree's case the money is contracted at a fixed price under forward sale agreements, so the adjustment is reasonable. But note which direction the adjustment runs, and note that it depends on those agreements settling as written.

Cash of $21.2 million on a $9.8 billion balance sheet is also worth seeing plainly. Agree does not hold cash; it holds a revolver and a forward equity book. That is efficient and it is normal for the sector. It is also why the liquidity figure matters more here than at a company that simply keeps money in the bank.

In the second quarter Agree bought at a 7.0% cap rate and sold at 7.1%. That means it disposed of buildings on slightly better terms for the buyer than it obtained on the ones it acquired.

The amounts are small — 14 properties for $30.3 million against $451.5 million bought — and a single quarter proves nothing. But it is worth contrasting with NNN, which sold income-producing assets at cap rates 170 basis points below its acquisition rate in the same quarter: selling dear, buying cheap. Recycling capital at a negative spread is not value creation, however small the sums. We would watch whether that pattern persists.

That is a materially better tenant list than NNN's, and it is the reason two-thirds of the rent carries an investment-grade rating. Walmart does not go bankrupt. Neither, on any reasonable horizon, do Tractor Supply or O'Reilly.

★ But be careful about what that buys you, because it is less than it appears. A credit rating protects you against a tenant failing. It does not protect you against a tenant declining to renew — and Agree's leases run 7.7 years against NNN's 10.1. You are trading two and a half years of contracted rent for a better balance sheet behind it.

Which is the better trade depends entirely on what you fear. In a recession, Agree's portfolio is safer. Investment-grade retailers keep paying; independent car-wash operators may not. Over twenty years, NNN's contracted term is worth more, and NNN's answer to the credit question — that it receives store-level financial statements covering 84% of its rent, so it watches the till rather than the rating — is a genuinely good one.

The bankruptcy record supports Agree's approach so far. Credit loss was 28 basis points in 2025 — twenty-eight hundredths of one per cent of rent — against a bankruptcy wave that took down Big Lots and Rite Aid. Guidance for 2026 is 25 to 50 basis points, and the first-half loss was Big Lots. That is what a portfolio of investment-grade retailers is supposed to do in a bad year, and it did it.

Agree pays monthly — $0.267 a share, or $3.204 a year, a yield of 4.30%. The most recent increase, in April 2026, took the annual rate up 4.3%.

What would force a cut. Nothing visible. A 70% payout, 99.8% occupancy, two-thirds investment-grade tenants and 28 basis points of credit loss through the worst retail bankruptcy year in a decade. Agree would need a simultaneous failure across its largest tenants — Walmart, Tractor Supply, Dollar General — to threaten the payment.

★ The more useful point is that at Agree the dividend is the smaller half of the return. 4.30% of income and 7.7% of growth. At NNN it is the larger half: 5.35% and 3.8%. If you need the cheque itself to be large — because you are living on it — NNN pays you a quarter more today. If you want the total to be larger over a decade, Agree is the better instrument. These are genuinely different products and the yield alone will mislead you about which is which.

★ We are usually suspicious when the sell side is this unanimous, and we will say so plainly. Twenty-two buys, nine holds and no sells is a consensus, and consensus is not analysis. But the underlying reasons here are checkable rather than atmospheric: AFFO per share genuinely grew 7.7%, occupancy genuinely is 99.8%, credit loss genuinely was 28 basis points, and guidance genuinely was raised twice. We reach the same conclusion by a different route and note the crowding.

So what does 16.3× assume? It assumes Agree keeps deploying $1.6–1.8 billion a year at cap rates around 7%, keeps growing cash flow per share at something like 7%, keeps occupancy near 100%, and keeps being able to issue equity on terms that make the arithmetic work. Three of those four are entirely within management's control. The fourth is not.

That is the whole bet, and it is a reasonable one at this price rather than an obvious one.