Three ways to call a REIT cheap, and why they can disagree about the same name

Question

Pricing a public REIT by forward FFO multiple, discount to consensus NAV, and implied cap rate against private comps can produce three different verdicts on the same name. Say a forward FFO multiple sits around 14x against a five year average nearer 19x. Reads cheap on that basis alone. A discount to consensus NAV might show roughly 12 percent below fair value, reading cheap but less dramatically, since NAV itself is an analyst estimate built on a chosen cap rate assumption. And an implied cap rate from the current share price, say around 6.1 percent, compared against private sales of comparable product printing closer to 5.4 percent by broker reports, reads cheap on that spread, though private comps and appraisal-based marks both tend to lag public pricing. So the three metrics can rank the same name differently depending on which anchor gets trusted, and running the exercise across several names often reshuffles the ranking entirely. The FFO multiple compares a company to its own history, which only holds up if that history reflects a fair pricing regime. The implied cap rate compares public pricing to a private market that may itself be the mispriced side of the equation. Which metric to lean on tends to depend on hold period. A two year hold favors the metric most likely to mean-revert quickly, often the FFO multiple relative to history. A ten year hold can better tolerate leaning on NAV or implied cap rate, since the underlying real estate value has more time to be recognized regardless of which market is temporarily right.

hannah_west29 · Answer

The first and third metrics here are the same argument counted twice. Consensus NAV is mostly NOI divided by a cap rate the analyst picked, so a 12 percent discount to NAV and a 70 basis point implied cap spread to private comps are two views of one number, and if the analyst's cap rate is stale, both move together. Averaging them feels like confirmation and isn't. Debt is worth adding to the picture. Implied cap rate calculations usually take debt at book value, and a company holding fixed-rate debt struck years ago at low coupons has debt worth less than book in today's market, which flatters the enterprise value it gets divided by. Two names at the same implied cap can have very different maturity walls behind them. Knowing what share of debt reprices in the next 36 months matters more than a fourth valuation metric. AFFO is the more reliable measure here, since capital spending is real and the FFO add-back hides it. Building AFFO by hand is worth the effort, and because companies define recurring capex differently, the resulting figures shouldn't be compared across names without checking those definitions first.

Rena Expert · Answer

The three metrics disagree because each one holds a different thing constant, and which one you deploy on depends on the assumption you are willing to underwrite.

The FFO multiple against its own history assumes the discount rate regime of that history is the one you are returning to. If the cost of capital has permanently reset higher, a 19x average is a memory, not a target, and the 14x is not cheap. That is the weakest of the three anchors on its own.

Discount to consensus NAV and implied cap versus private comps rest on the same input, as @hannah_west29 says. Both assume the private mark is closer to clearing value than the screen price. Appraisal-based valuations adjust slowly, so during a repricing the public market usually moves first and the gap you measure is partly a timing artifact. The gap between listed REIT valuations and private real estate marks has been unusually wide and unusually persistent by historical standards, and the argument for it closing is a mean-reversion argument rather than a dated one, which is why it can stay open past anyone's patience.

Hold period changes the weighting. Over ten years, cash flow growth and per-share capital allocation dominate the entry multiple, so AFFO growth and the debt maturity profile matter more than which anchor said cheap. Over two years you are underwriting the re-rating itself, which means the multiple gap and rate path are doing the work, and that is a timing bet with a wide distribution.

One mechanical point on replacement cost, since it is on your poll. It sets a ceiling on new supply rather than a floor on existing asset value, and it is the reason falling construction starts support pricing in a sector.

raj_mehta14 · Answer

@hannah_west29 the debt point is worth following closely. If the fixed-rate debt on the books is worth less than face in current markets, that value accrues to the equity holder in a real sense, but only cleanly if the company can actually buy the debt back at a discount, and most don't. Worth tracking whether a given REIT has a stated buyback or tender program for its own debt before assuming that gap shows up in the equity.

aliyah_b · Answer

I'm still working out what my first direct deal costs, so most of this is above me. But the replacement cost line landed. If building new costs more than buying the shares of somebody who already owns the finished thing, that's a comparison I can hold in my head without a spreadsheet.

stephen_ok9 · Answer

I vote implied cap against private comps, and the reason is that it is the only one of the three that compares the listed market to something outside itself. Multiple versus own history is a closed loop.

Where I'd push back on my own answer is dispersion. That 70 basis point spread @gregpowell66 found on one industrial name is not the sector's spread, it's that portfolio's spread against whatever comps happened to trade. In sectors where private transaction volume has been thin, the comp set is small enough that I wouldn't size a position on it. Data centers are the worst for this, since so little of the good stuff trades openly and the assets aren't interchangeable the way infill warehouse space nearly is.

Three ways to call a REIT cheap, and why they can disagree about the same name

5 replies