The all cash version keeps winning. What am I missing?

I've been running a side by side for a small rental inside a Roth, all cash versus non-recourse, and the all cash column keeps winning by more than I expected. I want someone to poke holes because I don't trust a result that clean.

Simple version. $200k, all cash, $1,600 rent, call it $1,050 net after taxes, insurance, management, and a maintenance reserve. That's $12,600 a year, no UBIT because rental income is exempt, no 990-T, no preparer fee. Roughly 6.3% on the cash, compounding tax free.

Leveraged version. Same property, 35% down, $70k in, $130k borrowed. Debt service eats most of the cash flow, I get maybe $250 a month, so $3,000 a year on $70k. That's 4.3% cash on cash, and then I owe UDFI on the debt-financed share plus a return preparer every year. The upside is I have $130k of remaining account cash to buy a second property.

Which is the part I can't model well. Two properties at $70k down each versus one at $200k. The two-property version has more total asset base and more appreciation exposure, and it also has two of every problem, two sets of custodian per-asset fees, and a UDFI calculation on both.

Where I keep going in circles is the reserve requirement. If I hold 12 months of expenses per property in account cash, the leveraged version's second property eats the reserve advantage I thought I was getting. So what exactly is the leveraged version buying me?

5 replies

1gregpowell66AG4BPRO8

Your $1,050 net on $1,600 rent is optimistic if the maintenance reserve inside it is a percentage rather than a dollar figure tied to the actual building. What's the roof age and what's the mechanical situation? On a 1960s house I'd want $250 a month set aside on its own before I called anything net.

2wendy_park48CapitalPRO7

The leveraged version is buying you appreciation exposure on $400k of property instead of $200k, and that's the entire case. Everything else about it is worse. Whether that trade is good depends on an appreciation assumption you haven't stated, which means your model isn't actually comparing the two things yet.

The other item missing from your columns is the exit. If there's debt on the property within the look-back period before you sell, part of the gain gets pulled into the debt-financed rules rather than sitting in the general exemption for gains, and how that lands depends on the debt history. That's a CPA question, and it belongs in the leveraged column as a cost you can't size yet.

Also your non-recourse loan will have a term and a balloon, and refinancing inside an IRA when rates have moved is a problem with fewer solutions than it has personally. Read the maturity provisions and confirm the terms in writing with the lender before you treat the loan as a fixed input.

1derekj19SB1B

Naive question. If the all cash version is that much simpler, why does anyone do the leveraged version at all?

1dpatel_reCS4A

beacon, because of fathom's point. Two properties appreciating beats one appreciating if appreciation is a meaningful part of your return, and inside a Roth that gain is the whole prize. I hadn't put an appreciation number in the model because I don't like forecasting it, which is exactly why the all cash column looked clean.

girder, the reserve was a percentage. You're right that it shouldn't be. Redoing that.

fathom, the balloon is the thing I'm going to sit with. A maturity I have to solve inside an account with a contribution cap and no personal guarantee available is a real constraint and I had it as a footnote.

The all cash version keeps winning. What am I missing?

5 replies