The freeze and the call are different provisions, and the difference is the single most useful thing on this page. They sit a few lines apart in the same regulation and they are not symmetrical.
Under 12 CFR 1026.40(f)(2) a creditor may not "terminate a plan and demand repayment of the entire outstanding balance in advance of the original term" unless one of four things is true: there is fraud or material misrepresentation by the consumer in connection with the plan; the consumer fails to meet the repayment terms for any outstanding balance; any action or inaction by the consumer adversely affects the creditor's security for the plan or the creditor's rights in that security; or federal law on credit extended by a depository institution to its executive officers requires it and the initial agreement said so. That list is short and every item is about the borrower's own conduct.
Under 1026.40(f)(3)(vi) a creditor may instead "prohibit additional extensions of credit or reduce the credit limit" during any period in which one of six things is true, and the list is far broader: the value of the dwelling declines significantly below the appraised value used for the plan; the creditor reasonably believes the consumer will be unable to meet the repayment obligations because of a material change in the consumer's financial circumstances; the consumer is in default of any material obligation under the agreement; the creditor is precluded by government action from imposing the agreed rate; the priority of the creditor's security interest is adversely affected by government action such that the security is worth less than 120 percent of the credit line; or the creditor's regulatory agency notifies it that continued advances would be an unsafe and unsound practice. Separately, 1026.40(f)(3)(i) allows a creditor to reserve the same power in the initial agreement for any period during which the maximum rate under the plan is reached.
Two of those six grounds require nothing of the borrower at all. A fall in local house prices, and a lender's reasonable belief about a change in the borrower's finances, are enough to stop the line. This is not a theoretical risk: the Federal Reserve's own compliance publication noted during the 2008 downturn that "many financial institutions have begun freezing or reducing credit limits on existing home equity lines of credit." It is the reason an undrawn line is not a substitute for cash reserves. A line is most likely to be reduced during exactly the conditions that make a household want to draw on it: falling property values and a disrupted income. Money already drawn stays on its original terms in those circumstances, because a value decline is not on the acceleration list. Money not yet drawn can disappear.
How far values have to fall is quantified, and it is a much smaller drop than "significant decline" suggests. The official interpretation of 1026.40(f)(3)(vi)(A) sets a threshold that is measured not against the home's value but against the borrower's cushion: if the value declines such that "the initial difference between the credit limit and the available equity ... is reduced by fifty percent," that is a significant decline. The Bureau's own worked example makes the scale plain. On a house appraised at $100,000 with a $50,000 first mortgage and a $30,000 credit limit, the available equity is $50,000 and the difference between it and the credit limit is $20,000; half of that is $10,000, so a fall in value from $100,000 to $90,000 is enough. A ten percent decline can therefore support a freeze, and the more of the equity a line already commits, the smaller the fall required.
A freeze is temporary in law, which is the other half of the picture. The same commentary provides that a creditor may suspend or reduce "only while one of the designated circumstances exists," and that when the circumstance ceases to exist "credit privileges must be reinstated." The creditor must either monitor the condition itself and restore access as soon as reasonably possible, or shift that duty to the borrower by giving a notice that puts them on notice to request reinstatement, which it may require in writing. It may charge only bona fide and reasonable appraisal and credit report fees actually incurred in investigating whether the condition persists, and it may not charge a fee to reinstate a line once the condition has been found not to exist. One further limit sits in the same commentary: a reduction may not take the limit below the outstanding balance where doing so would require the borrower to make a higher payment.
The rate is variable but not discretionary, which is the opposite of what "variable" suggests to most readers. 1026.40(f)(1) permits a creditor to change the annual percentage rate only where the change "is based on an index that is not under the creditor's control" and that index "is available to the general public." So a HELOC rate moves with a published benchmark plus a margin fixed by the agreement, and a lender cannot simply reprice the plan because it has decided to. That is a genuine protection, and it is also why the payment on a drawn balance rises when short-term rates rise, whatever else is happening.
Three provisions about fees and disclosures that are concrete and easy to miss because they operate before the plan exists. 1026.40(b) requires the disclosures and the brochure to be given "at the time an application is provided to the consumer," and 1026.40(e) names the brochure: "What You Should Know About Home Equity Lines of Credit," or a suitable substitute. 1026.40(h) then provides that neither the creditor nor anyone else "may impose a nonrefundable fee in connection with an application until three business days after the consumer receives the disclosures and brochure," with mailed documents deemed received three business days after mailing. And 1026.40(g) requires a creditor to "refund all fees paid by the consumer to anyone in connection with an application" if any term required to be disclosed changes before the plan is opened, other than through movement in a variable-rate index, and the consumer therefore decides not to open the plan. Taken together, those give a genuine window to read the terms and a remedy if the terms move.
The right to rescind attaches to the plan rather than to each draw. Because a line is open-end credit, the applicable provision is 12 CFR 1026.15 rather than the closed-end rescission section. 1026.15(a)(1)(i) gives each consumer whose ownership interest is subject to the security interest a right to rescind the plan when it is opened, a security interest when it is added to secure an existing plan, and an increase in the credit limit. But (a)(1)(ii) provides that the consumer does not have the right to rescind each credit extension made under the plan where the extension is made in accordance with a previously established credit limit. So opening the line, or raising the limit, comes with three business days to cancel in writing; an ordinary draw does not.