IRR for Rentals: When It Beats Cash-on-Cash and When It Lies

Cash-on-cash return describes one year. IRR describes the whole hold: every dollar in, every dollar out, and the year each one moves. When two rentals differ mainly in rent growth, appreciation or exit timing, cash-on-cash cannot see the difference and IRR can.

The example in this post puts $100,000 into each of two rentals that both earn an 8.0% cash-on-cash return in year one. After a 10-year hold, one finishes at a 15.2% IRR and the other at 20.0%. Then a third deal with half the starting cash flow beats the first one anyway. The last part of the post is about why you should distrust that result until you have tested the assumptions under it.

What IRR measures that cash-on-cash cannot

Cash-on-cash is annual pre-tax cash flow divided by total cash invested. It is the right tool for one question: what does this property pay me next year on the money I put in? The mechanics are in the cash-on-cash return guide.

It leaves out most of what happens over a hold.

IRR, the internal rate of return, is the single annualized rate that accounts for the full schedule of cash flows. Formally it is the discount rate at which the present value of everything the deal pays you equals what you invested, so that net present value is zero. That is as much math as this post needs.

The time-of-money intuition without the calculus

A dollar that arrives sooner is worth more, because you can put it to work while you wait for the later one. Discounting puts a number on that. At a 10% annual discount rate, $10,000 received one year from now is worth $9,091 today. The same $10,000 received ten years from now is worth $3,855 today.

IRR turns the question around. Instead of choosing a discount rate and computing a value, you take the deal's actual cash flows and ask which rate makes them worth exactly what you paid. A useful way to picture it: IRR is the interest rate on a savings account that could reproduce the deal, accepting your deposit on day one and paying out each of the deal's cash flows on schedule, with a zero balance after the last one.

Two one-line examples:

• Invest $100,000 and receive $118,000 one year later. The IRR is 18%.
• Invest $100,000, receive nothing in year one and $139,240 at the end of year two. The IRR is also 18%, because $100,000 compounding at 18% for two years is $139,240.

With more than a couple of cash flows there is no formula to rearrange. The rate is found by trial and error, which is what a spreadsheet's IRR function and any calculator do behind the scenes: guess a rate, check the present value, adjust, repeat.

Worked example: two deals with equal cash-on-cash, different IRR

This is an example with invented properties and an example interest rate. Both deals share the same structure:

• Purchase price $400,000, with $100,000 down. Closing costs are left out to keep the arithmetic visible.
• Loan of $300,000 at an example 7% rate over 30 years. Annual debt service is $23,951.
• Year-one NOI of $31,951, so year-one cash flow is $8,000 and cash-on-cash is 8.0%.
• Sale at the end of year 10 with 6% selling costs. The loan balance at that point is $257,437.
• Income taxes are ignored on both sides.

The only difference is growth. In Deal A, a steady market, NOI and property value both grow 2% a year. In Deal B, a growth market, both grow 4% a year.

To reproduce the IRR in a spreadsheet, enter the year 0 through year 10 cash flows in a column, add the net sale proceeds to the year 10 figure ($215,138 for Deal A, $320,660 for Deal B) and run the IRR function on the column.

On day one these deals were identical by cash-on-cash. Over ten years Deal B returned $131,981 more on the same $100,000.

The realistic version: growth costs you cash flow

Markets do not hand out 4% growth at the same price as 2% growth. Buyers bid up the growth market, so the same $400,000 buys less income there. Call this Deal C: same price, same loan, same 4% growth, but year-one NOI of $27,951. Year-one cash flow is $4,000 and cash-on-cash is 4.0%, half of Deal A's.

Deal C's yearly cash flows run $4,000, $5,118, $6,281, $7,490, $8,748, $10,056, $11,416, $12,831, $14,302 and $15,832. It pays less than Deal A in each of the first seven years and $14,271 less in operating cash flow over the whole hold. It still wins on IRR, 16.9% to 15.2%, because the sale is $98,230 larger.

The assumptions doing the work

That win rests on one input. In Deal C, 76% of all cash returned arrives on the day of sale, against 65% for Deal A. The result depends on a 4% growth rate holding for ten straight years and on a buyer paying the projected price at the end.

Drop Deal C's growth from 4% to 3%, for both NOI and value, and its IRR falls to 14.5%, below Deal A. The two deals tie at roughly 3.3% growth. One point of a growth assumption, which nobody can forecast over a decade, flips the ranking. Deal C also starts with a 1.17 DSCR, so a rent dip or a vacancy stretch in the early years hurts it far more than Deal A at 1.33. IRR rewards Deal C for the upside and says nothing about that thinner margin.

Where IRR misleads: reinvestment, short holds, and refinance spikes

Reinvestment

A 15.2% IRR does not mean your $100,000 compounded at 15.2% for ten years. That is only true if every interim cash flow was reinvested at 15.2% as well. Deal A turned $100,000 into $311,250 of total cash. If the annual cash flows sat in a checking account, your wealth grew at about 12.0% a year, since $100,000 compounding at 12.0% for ten years is roughly $311,250. The more of a deal's return that arrives early, the wider this gap gets. Always read IRR next to the equity multiple.

Short holds

IRR ignores scale and duration. The one-year deal from earlier, $100,000 in and $118,000 out, has an 18% IRR and beats Deal A's 15.2%. It also produced $18,000 of profit against Deal A's $211,250, and it left you holding cash that has to find a new home, with a fresh round of purchase costs and another search. A high IRR on a short hold is a statement about speed. It tells you little about how much wealth the deal created.

Refinance spikes

A cash-out refinance returns capital early, and IRR loves early capital. Take Deal A and refinance at the end of year 3, when the example value is $424,483 and the loan balance is $290,181. A new loan of $318,181, about 75% of value, puts $28,000 in your pocket. Ignore refinance costs and keep the 7% rate. Debt service rises from $23,951 to $25,402, and the payoff at sale rises to $290,014.

The IRR goes up, from 15.2% to 16.3%. Total cash returned goes down, from $311,250 to $296,516, and the property carries more debt with thinner coverage for seven years. Nothing about the property improved. Pushed to the limit, a refinance that returns all of your capital makes IRR undefined, which is where "infinite return" claims come from. Whenever a projection shows a refinance, look at the IRR with and without it.

Which metric for which decision

No single number covers all five rows. A deal worth buying should look acceptable on the first row and good on the second, with the last row telling you how much faith the projection requires.

Model the full hold, not just year one

IRR changes with the exit year, and the pattern is worth knowing. Here is the example again with the sale moved to different years, everything else unchanged.

Early exits are crushed by selling costs. Late exits fade because equity piles up in the property and earns less per dollar than it did when the loan was large. Deal C peaks around year 7 and Deal A around year 10, in this example.

Practical next steps:

1. Project cash flow year by year for the full hold, including the sale, selling costs and loan payoff. The single family calculator produces the year-by-year figures.
2. Compute IRR at your base growth rate, then again with growth one point lower. If the ranking of your candidate deals flips, the decision rests on a forecast and you should weight year-one DSCR more heavily.
3. Write down the share of total cash that arrives at sale. As a rule of thumb, once that share passes 70% the IRR is mostly a bet on the exit.
4. Compare exit years before you commit to a plan. The 30-year projections in Pro compute IRR for each hold period, so you can see whether year 5 or year 15 is the better exit for a specific property.