How do you weight for the sector dispersion the index number hides?

Question

The two-decade figure of roughly 9 percent annualized on the all-equity index keeps getting used as the case for owning REITs broadly. My problem with using it as a planning number is that the index composition has changed enormously over that period. Data centers and towers were barely a category and now they're a large slice. Office was a foundational sector and is now bifurcated between trophy assets that lease and everything else that doesn't.

So the historical index return is partly the return of a portfolio that no longer exists. If I buy the index today I'm buying today's weights, which are heavy in exactly the sectors that the structural story favors, which also means I'm buying them after that story became consensus. The supply picture supports it: new deliveries forecast down sharply, 20 to 70 percent depending on sector, across apartments, industrial, storage and senior housing. That's a genuine pricing-power argument for owners of standing assets.

What I can't resolve is whether to hold the index and accept whatever weights the market gives me, or to tilt toward the favored sectors and take concentration risk on the theory that sector selection is the decisive variable. The tilt version means fewer names, more dependence on individual management teams, and a real chance I'm paying up for the tailwind. The index version means I'm holding office at whatever weight the market says.

How are people who've thought about this actually splitting it?

Rena Expert · Accepted Answer

Two things you're treating as one problem. Sector exposure and manager exposure aren't the same risk, and you can separate them. Sector-focused REIT ETFs exist for data centers, industrial and healthcare, which gets you a tilt without betting on one management team. They cost more than a broad REIT index fund and they're narrow enough that a single tenant problem in a top holding shows up in your return, so read the concentration disclosure before assuming you've diversified anything.

On the composition point, you're right that the index return you're quoting is a chain of different portfolios. That cuts against the tilt argument too. The index reweighted into towers and data centers because those companies grew, and it will reweight out of whatever fails next without you making a decision. That reallocation is a lot of what produced the long-run number. A static tilt held for a decade doesn't get that.

Where I'd push back hardest is on the supply argument. Declining new deliveries support rent growth for existing owners, and that's real. It also tells you that development economics stopped working, which means construction costs and financing costs are high enough to stop projects. The same conditions raise the cost of capital for the REIT that wants to buy or build, and REITs distribute most of their taxable income so they depend on external capital more than most equities do. A supply squeeze and an expensive-capital environment are the same environment.

The risk you haven't named is tenant concentration in the favored sectors. Data center demand looks structural in aggregate and rests on a small number of very large tenants signing long leases. If those tenants slow their own capital spending, the sector's growth rate changes faster than the supply picture does.

mikeoconnell7 · Answer

The sector ETF answer is the practical one. I looked at a healthcare-focused one and the top five holdings were 55 percent of the fund, which is worth knowing before you call it diversified.

raj_mehta14 · Answer

I'm mostly lost on this thread but the point about supply falling because development stopped penciling is the first thing that's made the whole picture make sense to me. Cheap shares and expensive construction are the same fact seen twice.

aaron_fitz10 · Answer

@quill what's your holding period? If it's twenty years the reweighting argument seems strong. If you're expecting the valuation gap to close in two or three years then you're making a timing bet either way and the tilt is just a bigger version of it.

brandonlee36 · Answer

Practical middle: broad index as the base, a smaller sleeve for the tilt, and a rule about when the sleeve gets trimmed. Without the rule I've noticed the sleeve grows every time the story gets more exciting.

laurenkim53 · Answer

Nobody's mentioned senior housing much. Aging demographics is about as slow and predictable a driver as exists, and the operating side is genuinely hard, staffing especially. Different risk shape than data centers.

devon_pierce · Answer

@hollow that's the one I keep coming back to as well. The demand looks certain and the margins depend entirely on labor costs, which nobody forecasts well.

paulgrant17 · Answer

The tenant concentration point about data centers is the one that's going to keep me up. I had my mind mostly made up before reading this thread and now I'm less sure than I was. A structural tailwind that runs through five customers isn't the same kind of structural as an aging population.

How do you weight for the sector dispersion the index number hides?

8 replies