I held my office REIT position through 2022 and 2023 and watched maybe 40 percent evaporate before the thesis even had a chance to be wrong, so I am genuinely asking how other people decide when the sector headwind is temporary noise versus something that should have changed the position size six months earlier.
Office in 2022 and 2023 was a real test case, because the headwind there was structural, not cyclical. A cyclical headwind (rising rates, a slow quarter of leasing) tends to hit the whole REIT market and then ease as conditions shift. A structural headwind rewrites the demand side permanently, and remote and hybrid work did exactly that to office. The distinction matters because a cyclical case argues for holding through the noise, while a structural case argues for resizing before the thesis has time to fail.
The practical signal most allocators watch is occupancy trajectory rather than share price. If the underlying buildings are filling back up, the price drop is likely a sentiment lag and the thesis is intact. If vacancy keeps rising while price falls, the market is probably reading the fundamentals correctly and you are holding a thesis that the property itself is not supporting.
One thing worth naming: position sizing before entry does a lot of work here. If office was, say, 20 percent of a REIT portfolio going into 2022, a 40 percent drop in that position hurt badly. If it was 8 percent, the same move was painful but survivable and left dry powder to rotate toward industrial or data centers as those structural tailwinds became clearer. The guide section on sector tilting covers this logic directly.
I am not sure what your full portfolio weighting looked like, and that changes the answer about what "should have" happened earlier. What percentage of your REIT exposure was in office at the time you bought in? That would help me give you something more specific to work with.
The question you are actually asking is whether the headwind was macroeconomic or structural, because those two things require completely different responses. A rate spike is temporary by definition, something you can model and wait through. A permanent shift in space demand, which is what five days a week becoming two days a week looks like, changes the terminal value of the asset class and no discount rate adjustment gets you back to the old underwriting. The signal that should have changed position size was not the price drop itself but the lease renewal data, specifically what tenants were signing for, at what term length, and at what effective rent relative to asking.
What is the current occupancy trend in the specific submarkets your REIT holds, and are leases rolling at above or below in-place rents right now?
