Traded my REIT sleeve 11 times in seven months, and spreads and taxes took it
The thing I liked about public REITs was that I could act. Direct property takes months to buy and months to sell, and a REIT takes a click. So I built a 40k sleeve and used the liquidity, which is exactly what I thought the point was.
Over seven months I made 11 round trips. Rotated out of residential into industrial, out of industrial into data centers, back into residential when the rate story changed, trimmed and re-added a healthcare name twice. Every single trade had a reason I could have explained to you at the time.
Result: the sleeve is up about 1.9 percent over seven months. A broad REIT index fund over the same period was up around 6. So I underperformed the thing I could have bought once and forgotten by roughly 4 points, on 40k, which is about 1,600 of decisions I paid for.
Where it went wrong, step by step. Two of the 11 trades were good. Six were roughly neutral before costs. Three sold something that then ran without me, and in two of those three I bought back in higher. The bid-ask spread on a couple of the smaller names was wider than I'd noticed, maybe a third of a percent each way, and eleven round trips multiplies that. And every gain I did take was short-term, which is taxed differently from long-term, something I should have confirmed with an accountant before I started rather than after.
The deeper error is that I treated liquidity as an instruction. Being able to sell in a day doesn't mean selling in a day is useful. The liquidity is worth having for two reasons, deploying into a trough when you find one and rebalancing back to a target weight. Neither of those is 11 trades in seven months.
What I'd do differently: set target sector weights, rebalance on a schedule or a threshold, and require myself to write the thesis down before a trade and keep the note. I suspect most of my reasons wouldn't have survived being written.