How to Read a Sensitivity Table Like a Lender

A pro forma gives you one answer. A sensitivity table gives you 25 or 30 of them, arranged so that the pattern between them is visible. The pattern is the point: where the numbers turn negative, how fast they change, and which input is doing the damage.

The example deal in this post shows $107 a month of cash flow. Its sensitivity table shows that a rent $125 lower leaves $11, and a price 5% higher with that same rent loses $59 a month. A lender reading the grid would see a deal with almost no margin. Most buyers reading the single pro forma would see a deal that cash flows. This post is about reading the grid the way the lender does. If you need to build one first, start with how to run a sensitivity analysis and come back.

What a sensitivity table shows in one glance

A two-variable sensitivity table takes one output, such as monthly cash flow, and recomputes it for every combination of two inputs. Everything else in the deal is held constant. The center cell is your base case. Every other cell answers "what if these two assumptions were different?"

Here is the example deal used throughout. All figures are illustrative, and the 7.0% rate is an assumption for the math.

The price is $280,000 with 25% down. Monthly cash flow is $107 and DSCR is 1.08. Now the grid: monthly cash flow across purchase price and rent, with the down payment held at 25% and taxes and insurance held fixed.

In one glance: 7 of 25 cells are negative, all of them in the lower left, and the base case sits two cells from the edge of that region in either direction.

Reading the axes: the two variables that matter for your deal

Before reading any cell, read the axes. Three questions.

Which variables? A table is only useful if its axes are the two assumptions you are least sure of. For a purchase under negotiation, price is one of them by definition. The other is usually rent, because it is an estimate until a lease is signed. For a deal with a rate lock weeks away, swap in the interest rate. For a property with a turnover problem, use vacancy. A table of two inputs you already know is decoration.

How big are the steps? The grid above moves price in $14,000 steps (5%) and rent in $125 steps (about 5%). Steps should match real uncertainty. If your rent comps span $2,300 to $2,550, a grid that runs from $2,200 to $2,700 covers the plausible range with a cell to spare. A grid with 1% steps shows 25 nearly identical numbers and teaches nothing.

Which variables can you control? Price is a decision. Rent, vacancy and rates are outcomes. When one axis is a decision and the other is an outcome, you read across the outcome axis to see your exposure, then move along the decision axis to buy yourself protection.

The grid also gives you exchange rates. Moving one column right adds about $97 a month. Moving one row up adds $70. So $125 of rent is worth roughly $19,400 of price, or about $15,500 of price per $100 of rent. If your rent estimate could be $100 high, the table says the hedge is a price roughly $15,500 lower. That is a negotiating number you can defend.

Finding the break-even frontier in the grid

The frontier is the line where the output crosses your threshold. For cash flow the threshold is zero. Trace it through the grid: it passes between $54 and -$16 in the first column, between $11 and -$59 in the second, between $38 and -$32 in the third. It runs diagonally from upper left to lower right.

Everything above and to the right of the line pays you. Everything below and to the left costs you. Once the frontier is drawn, you read three things from it.

1. Distance. How many cells separate the base case from the frontier? Here one step down or one step left stays barely positive, and a second step in either direction goes negative. The base case survives a 5% rent miss by $11 a month and does not survive a 10% miss.
2. Slope. A steep frontier means one variable dominates. A diagonal one, like this, means both matter and they trade off against each other.
3. The corridor. Mark the range of outcomes you consider plausible on each axis. If rent could land anywhere from $2,325 to $2,575 and you might pay $266,000 to $294,000, the corridor is the nine cells in the middle. Eight are positive, one is negative, and the values run from -$59 to $274. That block is your real forecast. The center cell is just its midpoint.

Zero is not the only threshold worth tracing. A lender draws the frontier at a DSCR of 1.20 or 1.25. You might draw yours at $200 a month, the minimum that justifies the work. At a $200 threshold only 8 of the 25 cells pass, and none of them is at the $280,000 price with rent under $2,575.

For the one-variable version of this idea applied to vacancy, see break-even occupancy.

The corner cases: best case is a trap, worst case is a plan

Every grid has four corners, and two of them deserve attention.

The best-case corner here is $441 a month: a price 10% under asking and rent 10% over your estimate. Both of those things happening together is unlikely, and when they do it is luck. The trap is that the eye goes there first, and the number starts to feel like upside you are entitled to. Never let a best-case cell justify a price. If the deal only meets your target in the top right, it does not meet your target.

The worst-case corner is -$226 a month, or $2,712 a year. This one is useful, because it converts directly into a reserve. Ask how long the bad condition could last and multiply. If low rent is the risk and a lease locks it in for 12 months, the cost of being wrong for one lease term is $2,712 at the worst price and $1,032 at the base price ($86 x 12). That is money you hold in cash on day one, in addition to the capital reserve already in the expense lines.

The other two corners, low price with low rent and high price with high rent, tell you about the trade. At $252,000 and $2,200 the deal makes $54. At $308,000 and $2,700 it makes $161. Paying up for a better rental can work, but only if the higher rent is real.

Three real tables read aloud

All three use the same example deal.

Table 1: DSCR by interest rate and price

Read aloud: "Lenders commonly want 1.20 to 1.25 on a loan that qualifies on the property's income. At the base case I am at 1.08, so this deal does not qualify at $280,000 and 7.0%. Only five cells reach 1.20. To reach it I need $252,000 at 7.0%, or a full point off the rate at my price. Each half point of rate costs 0.04 to 0.07 of DSCR, and each $14,000 of price costs about the same." The table has turned a vague worry about rates into a specific statement: this deal needs a bigger down payment, a lower price or a different loan.

Table 2: annual cash flow by vacancy and rent

Read aloud: "At my base rent the frontier sits just under 10% vacancy, which is about five weeks empty. A single family house is either 100% or 0% occupied, so one slow turnover puts me in the 12% or 15% row. The 20% row is a tenant who leaves and ten weeks to re-rent. Every cell in it is negative, and at base rent it costs $2,767. The worst corner is $4,675." That row sizes the reserve. Holding $4,675 to $5,000 in cash covers the worst year on the grid, and twice that covers two in a row.

Table 3: cash-on-cash return by vacancy and price

Cash invested is the 25% down payment plus example closing costs of 3%, so $78,400 at the base price.

Read aloud: "The best cell on the whole grid is 4.2%. There is no combination of price and vacancy within 10% of asking that gets this deal to a return worth the effort from cash flow alone." This is the most useful thing a table can tell you: no single assumption is the problem. The deal is priced for appreciation or for a buyer with cheaper money, and no amount of optimism on the inputs changes that.

Make the table part of every offer decision

A routine that takes ten minutes per deal:

1. Pick the two inputs you are least sure of and set steps that match your real uncertainty.
2. Draw the frontier at your own threshold and at the lender's.
3. Mark the corridor of plausible outcomes and read its worst cell before its center.
4. Size the cash reserve from the worst-case row.
5. Move along the price axis until the corridor sits on the right side of the frontier. That price is your ceiling, and you write it down before negotiating.

For the example, step 5 says the offer is near $252,000, 10% under asking, or the deal is a pass. That conclusion was invisible in the $107 base case.

In the single family calculator, the Sensitivity tab builds three grids from your inputs: cash-on-cash return by purchase price and vacancy, DSCR by purchase price and interest rate, and monthly cash flow by rent and expense changes. Alongside them it reports break-even vacancy, break-even interest rate, break-even rent decline and the maximum purchase price that still cash flows, which are the frontier points of each axis stated as single numbers. Sensitivity is a Pro feature that is included in your first full analysis, and Pro generates the grids for any saved deal in one click, so the table is there every time you reopen a deal to decide on an offer.