You need $40,000 and the market is down 22%. Selling turns a paper loss into a permanent one, because the shares you sell don't participate in the recovery. A funded policy is built for exactly this, and it's the clearest use case there is.
A policy loan isn't a sale of anything. The carrier lends you its own money and holds your cash value as collateral. Nothing gets liquidated, nothing gets realized for tax purposes, and no shares leave your brokerage account. Your cash value also keeps working while the loan is out, though how much depends on the contract. Some contracts use non-direct recognition. That means the carrier ignores the loan when it pays your dividend, so you get paid on your full cash value as if you never borrowed. Under direct recognition the carrier pays a different rate on the borrowed part. On an IUL it depends on the loan type you picked. Know which one you have before you count on it, and when the policy loan rate rises on a balance you're still carrying covers the rest.
What you're really doing is paying loan interest to avoid selling. So the comparison is the loan rate against what you expect the un-sold shares to do over the payback period. And it changes depending on which account the money would have come from.
From a traditional IRA or 401(k), every dollar out counts as ordinary income, taxed at the same rate as your paycheck. A withdrawal in a down year also means selling more shares to net the same cash. The policy loan wins that comparison most of the time, and the tax piece is usually bigger than the market piece.
From a regular taxable brokerage account it's closer. Selling at a loss has real value. A loss you take offsets your gains, plus up to $3,000 of regular income a year, and the rest carries forward to later years. If you're sitting on a position that's down and you have gains somewhere else, selling might be the better move even with the recovery argument pointing the other way. Run it both ways, or have your CPA run it. I'm a licensed insurance broker and not a registered advisor. Nothing here is advice to buy or sell any security.
Two things about the loan itself decide whether this works. It reduces the death benefit while it's outstanding, dollar for dollar plus the interest that's built up. So a large loan carried through years when your family still needs the coverage is a real exposure. And the interest compounds if you don't pay it, which is how a loan taken for a good reason becomes a problem eight years later. Set the payback the day you take it, using nobody bills you for a policy loan, set a schedule anyway.
The last piece is capacity. This only works if the policy holds enough cash value to cover what you need with room left over, and if the loan doesn't push you up against the loan limit. A policy in year four usually can't carry a $40,000 loan. A policy in year fifteen usually can. Which is the argument for funding it before the down year instead of during one, and it's the same argument in what waiting a year actually costs.