This week the Biden Administration briefly considered a “billionaire’s tax” on unrealized capital gains of the very wealthy. Something like this is clearly necessary. The very wealthy are the people it is crucial to tax, because one of the most important (badly lapsed) functions of taxation is to compress the wealth and income distribution. Civilized society does not survive outrageous dispersion of outcomes. But for now, the very wealthy are effectively immune to income taxation, because they can simply borrow against appreciating assets rather than sell them and realize what the tax code recognizes as income. Eventually loans must be repaid, and so some assets sold, but “step-up basis” means taxes deferred this way ’til death are never paid upon wealth that accumulates behind the veil of an asset.

Taxing unrealized gains is the most straightforward solution. Jesse Eisinger argues we shouldn’t consider it that unwieldy. But… it’s kind of unwieldy, and kind of ugly. In order to tax unrealized gains, we need to be able to quantify them. When zillionaires hold their wealth as publicly traded common stock, we can just mark it to market. But if we were to impose an unrealized gains tax, the rich would likely flee from price-transparent shares to more bespoke, less liquid assets. Who can know how much their value has appreciated? Taxing unrealized gains sets up an endless war over valuations between the wealthiest people in the world and the tax authority, exactly the kind of struggle tax authorities have historically surrendered.

Rather than an unrealized valuation tax, I’ve long been partial to the idea of treating use as collateral as a realization event. When Mr. Burns takes a loan, pledging assets marked in the contract at a value of 1.56 zillion dollars, tax the appreciation inherent in that 1.56-zillion-dollar valuation as income, and then set the cost basis of those assets to 1.56 zillion (so that if Burns does sell the assets some time in the future, he is not taxed twice on the same appreciation). I thought this was a novel idea, but the infinitely knowledgeable Carlos Mucha disabused me of that. We’ve been doing it in some contexts since the 1970s.

Still, there’d be a lot of wiggle room by which zillionaires could squirm free. Lenders might compete to offer low collateral marks. More straightforwardly, since zillionaires’ personal spending, however extravagant, cumulates only to a fraction of their overall wealth, they are creditworthy. They can simply ask their bankers for unsecured loans. No use of assets as collateral, no realization, no tax.

So here’s a simpler, more bulletproof idea. When a wealthy person takes a loan, they are electing that means of generating cashflow against the alternative of selling some assets and paying the capital gains tax. Let’s make those two ways to generate cashflow equivalent for tax purposes. How would we do that? Well, we’d tax loan proceeds as income, and increase the cost basis on the borrower’s asset portfolio by the amount of a loan.

Let’s try a concrete example. Mr. Burns owns shares of ZeroCarbon, Inc. worth 10 zillion dollars. He needs one zillion dollars to deck out his dungeon. Suppose he sells 12.5% of his holdings in ZCI. (He acquired the shares years ago, for basically nothing, when he founded the firm.) He realizes 1.25 zillion from the sale, but pays Uncle Sam 20%, or 0.25 zill, raising the cool $1Z. If, some time later, he decides to sell the rest, he’d realize $8.75Z more in gains, but pay $1.75Z in taxes, netting $7Z more, or a total of $8Z, as the tax man takes 20%.

Now suppose he borrows. Under our proposal, he pays 20% on the loan proceeds, so he has to borrow the same $1.25Z and pay the same $0.25Z if he want $1Z in cash flow. But we add $1.25Z to the cost basis of his remaining portfolio. If, after a while, he decides to sell out and retire to Cuba, he sells all $10Z worth, pays back the loan of $1.25Z, but only pays tax on ($10Z – $1.25Z =) $8.75Z, leaving a tax bill of $1.75Z. Whether he borrows or sells, he pays the same $0.25Z on the initial cash flow generation, and $1.75Z on the eventual liquidation. Borrowing is, for tax purposes, the same as selling (although Mr. Burns’ forward-looking risk and income are affected by the choice). [*]

In practice, we don’t have to worry about what portfolio assets we add basis to. Any personal loan would be taxed as income, but the borrower would be granted a tax asset that could be deducted from the proceeds of any future capital gain. These kinds of assets exist already. For example, if you take an overall loss on your investments in a year, you acquire a “carryforward”, a right to deduct the amount of that loss from the gains in future years. A loan would simply be treated as an event that generates both income and generic capital-gains basis.

This would close the billionaire’s loophole. But what would it do to regular people? When Marge takes a mortgage to buy a home, should she have to inflate the value of the loan to cover an income tax hit, and console herself with an offset against future capital gains that she doesn’t anticipate realizing any time soon? No, probably not. But that is easy to fix. Up to a lifetime limit of $X, a household should be able to elect the current status quo — borrowings go untaxed, but no tax asset is issued. Beyond $X of borrowing, the new tax accounting would become mandatory.

How big would $X have to be to hold normal people harmless under this new scheme? Not so high. Add the price of a middle-aged (not “starter”) home and a bunch of college debt. We’ll presume transactional credit card debt (ie paid off quickly and without interest) is excluded, and assume like $100K of revolving hell. For most of the country, that takes you to maybe $600K. (Note that you don’t need to worry about moves and new mortgages, because the homeowner can plough the proceeds from her last home into the new home to avoid unnecessary borrowing.) If, my dear unrepresentative reader, you think of people who go to expensive schools and get graduate degrees and become homeowners in high-priced metros, then we add those items up and we get to maybe $1.5M. So we could make the lifetime limit of untaxed borrowings like $2M, and 99% of not-super-rich people would lose nothing from this change. (We’d have to decide how much to raise the threshold over time to accommodate inflation, or resist raising it to help counter leveraged tuition and home-price appreciation.) For some people of ordinary means, this change would be a tax cut, because we would gain an option to be taxed at low rates on borrowed cash in low-income years, but deduct from future capital gains in high income years.

But for the super rich? They’d now be taxed on what we think of as income, on the money they take from the empires they build in order to spend and live. A $2M exemption would be a rounding error to them.

This proposal wouldn’t raise as much as taxing unrealized gains, if we imagine unrealized gains can be effectively taxed. Zillionaires typically borrow to spend far less than their assets appreciate. As long as step-up basis remains in place, the difference between what tycoons borrow over their lifetimes and the value of their assets at death would remain untaxed. However, at least the funds they take for personal use would be taxed in a clean, hard-to-contest way. And we can counter step-up basis a bit by insisting that as long as basis steps up with death, tax assets accrued from loans must disappear with it. A better tax regime would keep basis at actual cost through death and let tax assets pass to heirs. But if we are fudging in favor of rich families with the step up, there’s no reason why they should also be able to add past tax payments to that fictional appreciated basis. Whinings about families having to sell the farm to cover the tax bill would not apply. The taxes would already have been paid. Tax assets would simply disappear, nothing would have to be liquidiated to raise new cash.

Treat loans as income plus future basis. It’s a simple, clean way of taxing people who currently live extraordinarily well and contribute effectively nothing for it. I think it’s worth considering.