$35bn of refunds, and where it goes

US Customs and Border Protection said on 22 May that it has paid out roughly $35.5 billion in court-ordered tariff refunds to importers. The money has arrived, and it is already changing what companies do with their balance sheets.

Where the cash went

Corporate treasury filings show that little of the $35.5 billion is going into long-term capital spending, new domestic factories or higher pay. Most of it is being used to repair balance sheets and reward shareholders. Multinationals are putting the one-off cash towards paying down floating-rate revolving credit lines and speeding up share buybacks.

The pattern is familiar. When the American Jobs Creation Act of 2004 let multinationals bring foreign profits home at a sharply reduced tax rate, the measure was sold as a spur to domestic investment, and academic studies later found that most of the repatriated cash went to shareholders. The corporate tax cut of 2017 was followed by record buybacks in 2018. Cash that arrives in a lump, with no change in the expected return on new projects, goes wherever it can be deployed fastest.

The payments so far are only the first tranche. The total at stake in the refund litigation is around $166 billion, so roughly a fifth has been paid, and the speed of Customs' new electronic system suggests much of the rest could follow within months. Whatever companies do with the first cheques is a fair guide to what they will do with the later ones.

Why investment will not follow

Companies invest when they expect a project to earn more than it costs, and a refund of past duties changes neither side of that sum. Large importers with access to credit markets were not kept from building factories by a shortage of cash; they were held back, if at all, by doubts about demand and about trade rules. Those doubts have not lifted, since the rules are still being rewritten in the courts. No board will commit to a plant with a thirty-year life on the strength of a trade regime that shifts with each ruling.

Smaller importers are the exception worth watching. A family-owned distributor that paid duties out of its overdraft was short of cash in a way a multinational was not, and for such firms a refund can go straight into inventory, hiring or delayed equipment. Those companies file no treasury reports and make no buyback announcements, so their response will show up only slowly, in regional surveys and small-business borrowing data. They also tend to be the slowest to file claims, which means their share of the money arrives last.

Paying down the revolver

Debt repayment is the more defensible use. Revolving credit lines are priced at a spread over the Secured Overnight Financing Rate, so with the Fed's policy rate at 3.50–3.75 per cent, every dollar repaid saves interest at once and leaves the line free to draw again. That is a rational treasury decision, and it lowers the chance that a company runs short of cash in a downturn. It adds nothing to demand in the economy.

It also flatters credit metrics at a convenient moment. Lower drawn balances improve leverage ratios that rating agencies and lenders watch, and a company heading into a period of dearer energy and uncertain trade policy is sensible to keep its borrowing capacity unused. Bondholders gain a little safety; the wider economy gains nothing it can spend. Because a repaid revolver can be drawn again, the cash is not lost to the company either, merely parked where it earns the highest safe return until a better use appears.

Borrowed money, bought-back shares

The result is an odd transfer. The federal government has borrowed more to pay the refunds, and companies are promptly spending the cheques on buying back their own shares, which flatters the per-share measures on which executives are paid.

Washington does claw back a sliver. Since 2023 companies have paid a 1 per cent excise tax on the value of shares they repurchase, so a billion dollars of refund spent on buybacks returns about $10 million to the Treasury. A smaller share count lifts earnings per share without any change in profits, which is exactly the measure many bonus plans reward.

The corporate income tax recovers more. Duties were deducted as a cost when paid, so a refund is generally taxable when received, at the 21 per cent federal rate for a company with taxable profits. That still leaves most of each dollar with shareholders. Buybacks rather than higher dividends are the natural channel for it, since boards are reluctant to raise a regular payout on the strength of money they know will not recur.

A weak stimulus

Set beside the pandemic cheques sent to households, the refunds are a poor way to support demand. A family on a tight budget spends most of an extra dollar within months, whereas a company returning cash to investors hands it to people who save most of what they receive. At a time when high fuel prices are squeezing consumers, the same borrowing would have done far more for spending had it reached them directly. That was never the purpose of the refunds, which were ordered by courts rather than designed by economists, but it limits what they can do for growth.

Who gains, who pays

Shareholders and lenders are the clear winners: the first get cash back, the second get repaid early. Taxpayers finance the refunds through extra borrowing, and downstream customers who paid tariff surcharges on their invoices get nothing unless they win in court. The case for the companies is that the refunds correct an unlawful tax rather than hand out a gift, and that cash returned to investors is reinvested elsewhere in the economy. Both points are true, but neither turns a refund into new factories.

The figures that would change the call are orders for core capital goods, the capital-spending guidance in second-quarter results and the pace of buyback announcements over the summer. On the evidence of the first $35 billion, the refunds will lift buybacks and trim corporate debt this year while adding next to nothing to investment, and taxpayers will carry the extra borrowing.