How Much Home Equity Should You Leave Untouched? The Safety Cushion Math for 2026

Start With Your Timeline, Then Work Backward

The first question I ask about any borrowing decision is the one I ask about a rate lock: what's your timeline? If you're opening a home equity line of credit you plan to hold for fifteen years in a home you'll never sell, you're in a different position than someone who might relocate for work in three. The loan documents look identical. The risk doesn't.

Within any five-year window, there are three exits you might need from a home equity position. You might sell the house. You might refinance, either to fold the line into a first mortgage or because the first mortgage itself needs new terms. Or you might keep the line open as a reserve and simply need it to still be there if your local market dips. Each exit puts a different demand on the equity you didn't borrow. That untouched slice is the subject of this article, and I'm going to treat it as something you size on purpose rather than whatever happens to be left over after the lender's calculation.

The scale of the decision is large. ICE Mortgage Technology's most recent Mortgage Monitor puts total mortgage-holder equity at a record $18 trillion, with $11.7 trillion of what ICE calls tappable equity. That's an enormous pool of borrowing capacity, and every dollar of it comes with the same question attached: how much of it should stay where it is?

What "Tappable Equity" Measures, and Whose Question It Answers

ICE defines tappable equity as the amount you could borrow while still keeping a 20% equity cushion in the home, which is the same thing as combined debt at or below 80% of value. That 80% line is the industry's working definition of prudent, and it's the figure behind most lender ceilings for a HELOC or a cash-out refinance.

What that 20% cushion is for shapes everything that follows. The cushion protects the lender. Earlier in my career I worked in a bank treasury department, and the daily exercise was asking what each position would be worth if collateral prices fell and the borrower stopped paying. A 20% cushion means the lender can absorb a meaningful price decline, foreclosure costs, and a discounted sale, and still recover its principal. It's a loss-severity calculation made from the lender's seat, and it's the same calculation any lender runs before setting a ceiling.

Nothing in that calculation asks whether you could sell the house next year and walk away with cash, or whether you'd still qualify to refinance, or whether the line would still be open when you reached for it. Those are your questions, and the lender's ceiling was never built to answer them. I'd like you to bring your own number to the conversation about a line amount, and the rest of this piece is about how to compute it.

The Freeze Rule Written Into Every HELOC

If your plan is to keep the line open as a reserve, one provision of Regulation Z, the CFPB rule that governs home equity plans, deserves your full attention. Section 1026.40(f)(3)(vi)(A) permits a HELOC lender to prohibit additional extensions of credit or reduce the credit limit when the value of the home securing the plan "declines significantly below the dwelling's appraised value for purposes of the plan." A price decline alone is sufficient grounds under the rule, with or without a missed payment.

The rule leaves "significantly" undefined, so the Office of the Comptroller of the Currency's consumer guidance fills the gap with the standard regulators work from: if the price decline erases 50% of the equity cushion that existed when the line was set up, the decline counts as significant. The OCC's own worked example uses a $100,000 home with a $50,000 first mortgage and a $30,000 HELOC limit. The cushion is $20,000, so half of it is $10,000, and a drop from $100,000 to $90,000, a 10% decline, reaches the threshold.

Run that math forward and something useful falls out. The cushion is the home's value minus the combined debt. Half the cushion, expressed as a share of value, is half of your untouched equity percentage. Said differently, the price drop that reaches the regulator's own definition of significant is roughly your untouched equity divided by two. Leave 20% untouched and a 10% drop gets you there. At 30% untouched the required drop is 15%, and at 40% it's 20%.

That's my derivation from the OCC's example, and I'd label it as such. What I find striking is where it lands if you've borrowed to the ceiling. ICE's 80% CLTV line, the one that defines tappable equity across the whole $11.7 trillion, sits exactly at the point where a 10% price decline meets the regulator's example of a freeze-worthy drop. The stress test I'm about to run uses 10% for that reason. The tripwire is built into the ceiling itself.

One Home, Three Cushions, One Bad Year

Take a $400,000 home at the moment you set up a line. I'll run it at three combined loan-to-value levels, where combined debt means the first mortgage plus the full HELOC limit, since the limit is what the lender is exposed to and what the freeze standard measures against.

At 60% CLTV, combined debt is $240,000 and untouched equity is $160,000, or 40% of value; at 70% the figures are $280,000 and $120,000 (30%). At 80%, which is ICE's tappable ceiling, debt is $320,000 and the cushion is $80,000, or 20%.

See How Much Cash You Qualify For

Our AI calculates your top personalized loan options in minutes.

Calculate My Options

Now apply the stress test. Prices in your area fall 10%, so the home is worth $360,000. I'm holding the debt fixed, which is realistic if your HELOC has an interest-only draw period and your first mortgage is only a few years old, since neither balance moves much in the short run. The most recent FHFA House Price Index shows national prices still up 2.1% year over year, so I'm describing a local correction rather than a forecast, and I'll come back to what the FHFA data says about how local those corrections are.

Read the last column against the drop we just applied. If you set up at 60% or 70%, you've absorbed the decline with the trigger still 10 and 5 points away, respectively. If you set up at 80%, you're standing on it. Same house and the same market move; the only variable is how much equity you chose to leave alone at the start.

Stress-Testing the Three Exits

The table tells you where each cushion stands after the drop. What it can't tell you is what the drop costs you, because the cost depends on which exit you need.

Exit One: Selling Into a Softer Market

Selling punishes a thin cushion most severely, because a sale converts equity into cash only after transaction costs come out. For illustration, I'll assume total seller-side costs, meaning agent commissions, title and closing costs, and prorated items, of roughly 8% to 10% of the sale price. That's a modeling assumption for this exercise, and your own figure will depend on your market and how you sell. On a $360,000 sale, it comes to $28,800 to $36,000.

Subtract that from the remaining equity in each scenario. The 60% setup nets $84,000 to $91,200 and leaves the closing table with real money, and the 70% setup nets $44,000 to $51,200, thinner but still a cushion that funded the move. The 80% setup nets $4,000 to $11,200.

Using a 9% midpoint for selling costs, the break-even price decline for the 80% setup is about 12.1%, roughly 2 percentage points past the stress test. A slightly deeper local correction or a single repair credit after inspection, and you're writing a check at closing to sell your own house. By way of comparison, the 70% setup under the same 9% assumption can absorb about a 23% decline before reaching break-even, and the 60% setup about 34%.

The FHFA data is what keeps this from being abstract. In the most recent quarterly House Price Index, national prices rose 2.1% year over year, but four states posted year-over-year declines, with New Mexico the largest at 1.2%, and 24 of the 100 largest metro areas declined, led by Everett, Washington, at 3.7%. None of those is a 10% move. What they tell you is that a national average can be positive while a quarter of the largest metros are falling, and if you're the one selling in a falling metro, the national figure does nothing for you at the closing table.

Nobody designs a hull to sit at the waterline. Freeboard is the part of the boat that does nothing until the water moves, and once it moves, it's the part that keeps you dry. Untouched equity is your freeboard.

Exit Two: Refinancing on Someone Else's Schedule

The refinance exit is the one that looks most natural on paper, and it depends most heavily on timing you don't control. The typical plan runs like this: draw on the line at a variable rate, wait for the rate cycle to turn, then refinance the balance into a fixed first mortgage. In the most recent weekly Freddie Mac survey the 30-year fixed averaged around 7.4%, and if that figure falls meaningfully over your holding period, folding a HELOC into a new first mortgage may look attractive.

The dependency is the appraisal. A refinance requires one when you apply, and the new loan is sized against that appraised value. Rates and home prices move on their own schedules, and there's no rule that says they'll cooperate with yours. If rates fall during a stretch when your local market has softened, the 80% setup from the example is applying at an 88.9% CLTV. If the refinance lender draws its line at the same 80% CLTV that ICE uses to define tappable equity, you'd need to bring roughly $32,000 to closing to get combined debt down to $288,000 on a $360,000 value before the refinance could proceed.

The 70% setup faces a smaller gap: at $280,000 of debt against $360,000, the CLTV is 77.8%, under that 80% line with about $8,000 of room. The 60% setup at 66.7% has no gap at all and may leave room to structure the new loan differently if that suits you. The refinance exit rewards the cushion twice, once by keeping the door open and again by leaving you options once you're through it.

Exit Three: Keeping the Line Open Through a Dip

Doing nothing is the third exit. You open the line, draw little or none of it, and treat it as a reserve against a job loss, a medical bill, or a roof. The tax treatment of a draw depends on its use. Under IRS Publication 936, interest on a home equity loan or HELOC is deductible only when the borrowed funds are used to buy, build, or substantially improve the home that secures the debt, and only within the $750,000 aggregate cap on acquisition debt ($375,000 if you're married filing separately), so a draw for a reserve, a debt consolidation, or tuition carries its full interest cost with no deduction softening it. Your tax situation is your own; AmeriSave doesn't give tax advice, and a tax professional is the right person to apply these rules to your return. I'm renovating a historic home myself, and I understand the appeal of a line that's ready for whatever turns up behind the plaster. A reserve is only a reserve if it's there when you reach for it, though, and the freeze rule decides that.

See Your Top Loan Options In Minutes

Get Started

Go back to the formula. If you set the line up at 80% CLTV, a 10% local price decline reaches the OCC's own example of a significant decline, and your lender has the authority under Regulation Z to suspend further draws or reduce your limit. Think about when regional declines of that size tend to occur. They tend to arrive with local economic stress, which is also when job losses arrive. The reserve you set up for a bad year is structured so that a bad year can take it away. In risk management we call that correlated exposure, and in roughly three decades in mortgage finance I've watched it undo plans that looked sensible on the day they were signed.

This is where frequency and magnitude earns its keep. How often does a 10% local decline happen? In the most recent FHFA release the largest metro decline was 3.7%, so in the current data a drop that size isn't happening anywhere among the 100 largest metros. How big a deal is it when it does happen? If the line was your reserve and the freeze lands during the same downturn that cost you income, the magnitude is severe. A low-frequency, high-magnitude event is exactly the kind you insure against, and the premium here is simply leaving more equity unborrowed. At 70% CLTV the freeze standard needs a 15% decline, and at 60% it needs 20%. Neither cushion makes the line untouchable, since the rule permits a freeze and your lender's own policy governs whether it acts, but each moves the tripwire further from the declines the FHFA data shows actually occurring.

Sizing the Cushion: A Decision Rule

Everything above reduces to one rule: size the equity you leave untouched to the exit you're most likely to need within five years. The lender's ceiling answers the lender's question. Yours is different, and you get to answer it before you sign.

Applied to the three exits, the rule looks like this.

If selling is a live possibility, the cushion has to cover a plausible local decline plus the full cost of the sale. In the worked example, 30% untouched equity (70% CLTV) still nets $44,000 to $51,200 after a 10% drop and 8% to 10% selling costs, while 20% untouched nets close to zero. If your metro is already among the ones the FHFA shows declining, I'd want 40%.

When refinancing is the exit, work backward from the CLTV ceiling you'd face at the new lender and the decline you want to survive. Untouched equity of 30% leaves you under an 80% refinance line after a 10% drop. Untouched equity of 20% puts you above it and turns a refinance into a cash-in transaction.

For a line held as a reserve, use the derived formula. Decide the local decline you want to absorb without reaching the regulator's significant-decline example, then double it. Absorbing 15% means leaving 30% untouched, and absorbing 20% means leaving 40%.

In all three cases the answer lands well inside the lender's ceiling, and that's the point. The ceiling is the maximum the lender can tolerate. The cushion is the minimum you should demand for yourself. Between them is a range, and the exit you're most likely to need tells you where in that range to stop.

When you sit down with an AmeriSave loan officer to size a line, bring the exit and the cushion to the conversation rather than the maximum. In terms of the paperwork, a smaller line is the same application with a different number on it. Five years from now, it can be the difference between selling on your own terms and writing a check at closing to sell.

The Bottom Line

The equity you don't borrow looks like idle money, and every LTV table you'll ever see is built to make it feel that way. I'd ask you to see it as freeboard instead. That untouched slice is what stands between you and three outcomes worth avoiding: writing a check at closing to sell, refinancing with cash brought to closing, or reaching for a reserve that's been frozen. Each is a mechanism you can read directly in the regulation and the OCC's example, or check with simple arithmetic on a $400,000 house.

The decisions that build wealth in a home are few. Buying at a well-negotiated price is one. Choosing how much equity to leave in place when you borrow against it is another, and it's made in a single conversation about a single number.

Ask yourself what you'd regret two years from now: a line that was a little smaller than the maximum, or a cushion that turned out to be a little thinner than the exit you needed. Bring your timeline.