Suppose you could make a double or nothing bet with your retirement account. That’s not likely to make a lot of sense, given the possibility of losing the bet. But suppose you could bet half your retirement account on some proposition, and bet the other half on the opposite proposition. Does that sound like a pointless exercise? An intriguing article by UNC law professor Gregg Polsky explains how such an arrangement could be used to cut the tax cost of a Roth conversion.
Imagine that you’re converting a $200,000 IRA to a Roth and paying tax at 35%. In a normal conversion the tax cost would be $70,000. We’re going to imagine an abnormal conversion, however. You divide the IRA into two accounts and convert each one into a separate Roth. One makes an investment that will double in value under certain conditions and otherwise become worthless. The other makes an investment that performs in the opposite manner. After the dust clears, you have one Roth that made a winning bet, making it worth $200,000, and another that became worthless due to a losing bet. Do you see what’s coming next?
Recharacterize the conversion that went to the Roth that made the losing bet. Undoing this conversion eliminates $100,000 of conversion income. You pay tax on only $100,000, reducing your conversion cost from $70,000 to $35,000 — yet you still end up with $200,000 in your Roth.
Polsky doesn’t recommend that you rush out and perform a transaction exactly like this one. For one thing, it isn’t clear that there are suitable investment vehicles. What’s more, there are legal doctrines the IRS might use to attack the result, including the recently codified economic substance doctrine. Yet the idea may provide the conceptual base for a defensible arrangement that at least partially accomplishes the tax reduction goal. For those interested in a thoughtful exploration of sophisticated Roth conversion planning, his article makes a good read.
- You can download the article from a link on this page.