Why the all cash column tends to win the self directed IRA leverage comparison, and what the leveraged version actually buys
Run a side by side for a small rental inside a Roth, all cash against non-recourse leverage, and the all cash column usually wins by more than people expect at first pass. Say a 200k property, all cash, generates 1,600 in rent and roughly 1,050 net after taxes, insurance, management, and a maintenance reserve. That is 12,600 a year, no UBIT since rental income from a debt-free asset is exempt, no 990-T, no preparer fee. Call it 6.3% on the cash, compounding tax free. Now the leveraged version of the same property: 35% down, 70k in, 130k borrowed. Debt service eats most of the cash flow, leaving something like 250 a month, 3,000 a year on 70k, about 4.3% cash on cash, plus UDFI owed on the debt-financed share and an annual preparer fee. The upside is 130k of remaining account cash available for a second property. That second property is the part that is genuinely hard to model cleanly. Two properties at 70k down each versus one at 200k means more total asset base and more appreciation exposure, but also two of every recurring cost, two sets of custodian per-asset fees, and a UDFI calculation on both. The reserve requirement is where the leveraged version's advantage tends to erode fastest: holding 12 months of expenses per property in account cash across two properties frequently eats most of the capital efficiency the leverage was supposed to create. The honest answer for most investors comparing this inside a Roth is that leverage buys optionality and scale, not better current yield, and that trade only pays off if the second property's appreciation and eventual paydown outweigh the UDFI drag and doubled fixed costs over the hold.
