super interesting post ... that I don't totally follow, ha. When you say, "The US government makes a massive profit from that" ... you don't mean that literally, do you? Is it a "profit" in the sense that it's a net positive for the US economy?
It follows from my "Upside Down CAPM" framework. I'm saying that risk-free fixed income is a service from the borrower to the lender. Passive investors take cyclical risk and earn 6-7%+inflation over the long term. Savers who just want to preserve capital lend it to corporations with safe income-earning assets or governments with state capacity for stable tax income. You could say that they are paying a premium by accepting less than 6-7%+inflation. As a whole, Americans provide that service through the government, borrowing money at less than 6-7%+inflation, which allows us to defer taxation, and we can invest the untaxed capital in at-risk assets that earn 6-7%+inflation. That's our collective "profit".
We have to do that carefully. If we consume all of the difference without investing some of it, we will be like a utility company that issues a bunch of bonds and pays it all out in dividends instead of buying new electrical capacity. But, it does mean that "federal interest payments" aren't really a meaningful metric. That is just an anachronistic accounting derivative that results from our "profitable" borrowing/investing trade. The federal government issues treasuries paying, say, 4%, and some Americans invest their capital that remains available because the taxes have been deferred in something that will return 9% in the long-run. That's a net communal profit that eventually will flow to future taxes and pay off the bonds.
The metric to watch for is risk premiums. If the federal debt became unsustainable and savers started to demand a risk premium on treasuries, then we aren't engaged in the service of risk-free capital preservation any more, and the net benefit of doing the borrowing/investing trade wouldn't be so beneficial.
On the other hand, if high federal interest expenses are due to high interest rates with no risk premium, that is likely because capital allocators are more interested in making at-risk investments and aren't willing to "pay" a large "premium" to save capital at a guaranteed interest rate. That is usually associated with strong real economic growth (like in the 1990s).
If savers are willing to "pay" a large "premium" and are willing to take a very low interest rate on treasuries, then the the difference between the growth of private investments and the interest paid on federal debt is high (what I call our "profit"). If they want to invest in risky assets for growth, then the federal "profit" on treasuries is lower, but economic growth should be higher and future tax revenues (including taxes on wages and everything else) will be robust.
It's counterintuitive. I hope that helps clear it up a little.
There's an opposite force here too, however. USA stocks pay a lot less dividends than foreign stocks. Since dividends is counted in current account, and selling stock isn't even counted in the CA, this would actually mute the USA CA deficit, as domestic investors would make relatively more dividends from foreign stocks than foreign investors would make from USA stocks.
I can't imagine it's big enough to offset the profit gap you point out (buybacks are ~$300B and foreign investment is 15-20% of the stock market), but it might explain part of it.
Meanwhile row 26 of table 4 here (https://www.bea.gov/sites/default/files/2025-03/trans424.pdf) shows $296B of direct equity investment dividends + retained earnings in 2024. That's only 1.87% or an implied 53x P/E multiple. How does that make sense?
I think it may have to do with reinvestment. I think the measure of profit only captures repatriated profits because they are measuring it as transactions cross borders. Not 100% sure, but I think that’s what might be going on.
possible, not sure either. But if so, does that weaken the broader point of the piece? Let's say foreigners are earning $2.5T in income (but only repatriating $1.5T of it). Does that mean that the capital flow into the US is unsustainable?
I think it just means the reported numbers already net out the reinvested profits in both directions that I discussed. The gross flows are lower but the net numbers are the same.
good point, it cuts both ways. For what it's worth, I asked chatGPT o3 and it mentioned a few other possibilities, but the only one that made sense to me: "Part of the economic return leaves the U.S. through other channels. Affiliates routinely pay royalties for technology and branding, service fees, and intra-group interest. These are booked in the current-account services line, not as direct-investment income, so they are missing from your numerator."
I'm worried that Scott Sumner's endorsement will drive up traffic on your blog and I'll have to tone down the lunacy of some of my comments. There are so few safe spaces on the internet anymore....
Someday, we will succumb together on the town square under a barrage of rotten tomatoes. Until then, we have the comment section. That is my pledge to you.
super interesting post ... that I don't totally follow, ha. When you say, "The US government makes a massive profit from that" ... you don't mean that literally, do you? Is it a "profit" in the sense that it's a net positive for the US economy?
It follows from my "Upside Down CAPM" framework. I'm saying that risk-free fixed income is a service from the borrower to the lender. Passive investors take cyclical risk and earn 6-7%+inflation over the long term. Savers who just want to preserve capital lend it to corporations with safe income-earning assets or governments with state capacity for stable tax income. You could say that they are paying a premium by accepting less than 6-7%+inflation. As a whole, Americans provide that service through the government, borrowing money at less than 6-7%+inflation, which allows us to defer taxation, and we can invest the untaxed capital in at-risk assets that earn 6-7%+inflation. That's our collective "profit".
We have to do that carefully. If we consume all of the difference without investing some of it, we will be like a utility company that issues a bunch of bonds and pays it all out in dividends instead of buying new electrical capacity. But, it does mean that "federal interest payments" aren't really a meaningful metric. That is just an anachronistic accounting derivative that results from our "profitable" borrowing/investing trade. The federal government issues treasuries paying, say, 4%, and some Americans invest their capital that remains available because the taxes have been deferred in something that will return 9% in the long-run. That's a net communal profit that eventually will flow to future taxes and pay off the bonds.
The metric to watch for is risk premiums. If the federal debt became unsustainable and savers started to demand a risk premium on treasuries, then we aren't engaged in the service of risk-free capital preservation any more, and the net benefit of doing the borrowing/investing trade wouldn't be so beneficial.
On the other hand, if high federal interest expenses are due to high interest rates with no risk premium, that is likely because capital allocators are more interested in making at-risk investments and aren't willing to "pay" a large "premium" to save capital at a guaranteed interest rate. That is usually associated with strong real economic growth (like in the 1990s).
If savers are willing to "pay" a large "premium" and are willing to take a very low interest rate on treasuries, then the the difference between the growth of private investments and the interest paid on federal debt is high (what I call our "profit"). If they want to invest in risky assets for growth, then the federal "profit" on treasuries is lower, but economic growth should be higher and future tax revenues (including taxes on wages and everything else) will be robust.
It's counterintuitive. I hope that helps clear it up a little.
amazing ... thanks for the thorough response!
There's an opposite force here too, however. USA stocks pay a lot less dividends than foreign stocks. Since dividends is counted in current account, and selling stock isn't even counted in the CA, this would actually mute the USA CA deficit, as domestic investors would make relatively more dividends from foreign stocks than foreign investors would make from USA stocks.
I can't imagine it's big enough to offset the profit gap you point out (buybacks are ~$300B and foreign investment is 15-20% of the stock market), but it might explain part of it.
Interesting piece, thank you. Do you understand why the return on investments by foreigners into the USA is so low?
Example: row 40 of table 1 here (https://www.bea.gov/sites/default/files/2025-03/intinv424.pdf) shows $15.8T of direct equity investments.
Meanwhile row 26 of table 4 here (https://www.bea.gov/sites/default/files/2025-03/trans424.pdf) shows $296B of direct equity investment dividends + retained earnings in 2024. That's only 1.87% or an implied 53x P/E multiple. How does that make sense?
I think it may have to do with reinvestment. I think the measure of profit only captures repatriated profits because they are measuring it as transactions cross borders. Not 100% sure, but I think that’s what might be going on.
possible, not sure either. But if so, does that weaken the broader point of the piece? Let's say foreigners are earning $2.5T in income (but only repatriating $1.5T of it). Does that mean that the capital flow into the US is unsustainable?
I think it just means the reported numbers already net out the reinvested profits in both directions that I discussed. The gross flows are lower but the net numbers are the same.
good point, it cuts both ways. For what it's worth, I asked chatGPT o3 and it mentioned a few other possibilities, but the only one that made sense to me: "Part of the economic return leaves the U.S. through other channels. Affiliates routinely pay royalties for technology and branding, service fees, and intra-group interest. These are booked in the current-account services line, not as direct-investment income, so they are missing from your numerator."
I'm worried that Scott Sumner's endorsement will drive up traffic on your blog and I'll have to tone down the lunacy of some of my comments. There are so few safe spaces on the internet anymore....
Someday, we will succumb together on the town square under a barrage of rotten tomatoes. Until then, we have the comment section. That is my pledge to you.