Two data center REITs, one balance sheet I trust, wrong pick every time

Been rotating 90k of a taxable account into REITs since spring and I've made a mess of the data center piece specifically.

What I did: bought a position in one large data center name at what I thought was a reasonable multiple, watched it run, added on the way up, then when it pulled back 14 percent I added again on the theory that the AI demand story hadn't changed. Cost basis now sits well above where I wanted it. Position is 22 percent of the REIT sleeve which is more than I planned for any single name.

What's bugging me is that the growth story and the balance sheet are pulling opposite directions. These companies are building enormously, funding it with debt and equity issuance, and every new hall they light up needs capital before it produces a dollar. The demand is clearly there. Preleasing on new capacity looks strong. But the whole model runs on external capital, and if the cost of that capital moves against them the development pipeline that justifies the multiple becomes the thing that hurts them.

Meanwhile I've got a smaller position in a second name with lower growth, lower payout ratio, and much less development exposure, trading at a discount I can actually defend on current cash flow.

Decision: do I trim the big one back to 12 percent and move the proceeds into the boring one, or do I stop trading a thesis I still believe and just stop adding? I've already paid tuition on the averaging up.

Answer

1dmitri_v6CapitalPRO6

Before you decide the size question, go read the lease documents behind the growth number. That's where these two names will actually diverge and it's not visible in the multiple.

The things I'd want to see. Weighted average lease term on the preleased capacity, and whether that's counted from signing or from delivery. Escalators, and whether they're fixed at some percentage or tied to an index with a cap. Who pays power. That last one has moved from a footnote to the main event, because a lease that passes power through at cost behaves completely differently from one where the landlord eats a chunk of it. Tenant concentration too. If a handful of hyperscalers are most of the pipeline, you've got credit risk dressed as real estate risk, and their renewal leverage at year eight is enormous.

On the development funding, look at what's already committed versus what's funded. Companies disclose remaining spend on projects under construction. Compare that to available liquidity plus the forward equity they've already priced. If remaining spend exceeds what's in hand, they have to come to market, and they'll come to market at whatever price exists then.

The answer to your trim question depends on one assumption: whether you think the AI demand you're paying for shows up as signed leases with escalators or as announcements. If the disclosures show term and escalators you can defend, 22 percent in the better business is defensible and you should stop adding rather than trim. If the pipeline is thinner on paper than in the press release, trim to 12 and don't argue with yourself about the tax hit.

4 replies

4marcus_hale41CG4A

You paid tuition on averaging up and now you're considering paying more tuition to correct it. In a taxable account. Run the actual number on the gain you'd realize before you decide anything, because I've watched people rebalance themselves into a tax bill that exceeded the concentration risk they were worried about. Confirm the treatment with your own tax person, not with a forum.

The other thing. You describe the second name as "boring" and "lower growth" and then say it trades at a discount you can defend on current cash flow. Those are the same sentence. Somebody looked at the growth and priced it. The question is whether you think they priced it wrong, and if you do, say why in one sentence. If you can't, you're moving money from a thesis to a comfort.

1chriswalton8CC1CPRO8

@warrant's point on power pass-through is the one I'd never have thought to look for. Filing that away.

Quarry, on the funding question, the private side of this is instructive. A lot of new capacity is getting built in joint ventures with institutional partners precisely so the sponsor doesn't carry the whole construction cost on balance sheet. If your big name is doing that, its reported development exposure understates the pipeline and its cost of capital picture is better than it looks. If it's building on balance sheet alone, worse. That's disclosed but it's buried.

1megan_doyle15 Get yours

I'm new so mostly listening, but the thing I don't follow. You said the pullback was 14 percent and you added because the demand story hadn't changed. Isn't the whole point that a 14 percent move in six weeks means the market changed its mind about something? Or is it normal for these to swing that much on nothing?