Treated the valuation gap as an entry signal and got the timing badly wrong

This is a loss on process more than on money, and the process error is the interesting part.

Last year I built a deployment plan around the gap between public REIT valuations and where comparable private real estate was being marked. The gap was wide by any historical measure, the argument that it closes eventually is sound, and I've seen credible work calling this stretch the longest such divergence since the early 2000s. I read that, agreed with it, and then did something dumb with it. I treated a wide gap as a signal about when.

What I actually did: deployed 310k into a REIT sleeve in two large tranches about six weeks apart, weighted toward industrial and data centers, on the reasoning that the discount was extreme and therefore near its end. The sleeve then went down about 13 percent over the following five months and spent most of a year below my average cost. It has since recovered and I'm modestly ahead, so the money is fine. The plan was still wrong.

The specific error is that a valuation gap tells you about expected return over some long horizon and says nothing about the path. Two mechanisms can close it and they work on different clocks. Public prices can rise, which happens in weeks when sentiment turns. Private marks can fall, which happens over quarters because appraisals lag and nobody wants to write down a fund. When the second mechanism is doing more of the work, the gap narrows without your position going up at all. I had modeled convergence as a single event.

Second error, smaller but it cost real money. I sized the tranches off my conviction in the thesis rather than off my ability to keep buying. Two tranches six weeks apart is not a deployment schedule, it's one decision split in half. By month four I had nothing left to add with and the best prices of the whole period were sitting right there.

What I'd do differently. Deploy on a calendar over 12 to 18 months regardless of what the gap does, hold back at least a quarter of the allocation for a drawdown, and stop using the size of a mispricing as information about its duration. The thesis on convergence I still hold. My timing inference from it was unsupported.

6 replies

3monica_tran3SF4CPRO6

The two-mechanisms point is the useful one and I hadn't thought about it that way. If private marks are what's moving, the gap closes and you feel nothing. That's a genuinely uncomfortable version of being right.

Running a service business I see the same shape with pipeline. Being right about demand returning tells me nothing about which month I can hire.

2aliyah_bAU2APRO6

310k in two tranches six weeks apart. When you wrote the plan, did you consider a longer schedule and reject it, or did the six weeks just happen?

Asking because I'm working out what my first deployment looks like at a much smaller number and I want to know whether the schedule failure was a planning failure or an impatience failure.

2angela_burkeServicePRO6

@tally those are usually the same failure wearing different clothes.

@sextant the appraisal lag mechanism, how long does it typically run? If private marks take four to six quarters to catch up to public prices, then anyone underwriting convergence needs a horizon of at least that plus whatever it takes for sentiment to turn. That's a very different holding period than the one implied by "the gap is extreme."

1sonia_alvarez Get yoursPRO6

I'd add a third mechanism you left out, and it's the one I watch from the debt side. The gap can also close because REITs use their public listing to buy assets from private owners who need liquidity. That's accretive to the REIT and it doesn't require sentiment to turn or marks to fall, it just requires transaction volume to come back. Volume has been thin, which is part of why nothing has resolved.

So the honest count is three paths with three different clocks, and your position only benefits directly from one of them in the short run.

On sizing, holding back a quarter is the right instinct and I'd go further. A written rule that says what price or weight level releases the reserve, decided before you deploy anything, because deciding it during a 13 percent drawdown is how people end up not deploying it at all.

2derekj19SB1B

Sitting here as someone who mostly reads these threads: the sentence I'm taking away is that the size of a mispricing tells you nothing about its duration. That applies to almost everything, not just REITs.

1laurenkim53Passive

You came out modestly ahead and you're writing it up as a loss, which I respect, and I also want to name that a year underwater on 310k is a very different experience depending on whether you needed any of it. You didn't. That has less to do with your process and more to do with the fact that you had room.

I'm curious about your sector weighting. Industrial and data centers were the crowded end of the trade. If the convergence thesis was widely held, was the discount on those two ever as deep as the headline gap suggested?

Treated the valuation gap as an entry signal and got the timing badly wrong

6 replies