Latest / The Fiscal Policy Podcast with Fexingo: Government Budgets, Stimulus, and Public Spending / How Government Bond Auctions Create Hidden Winners and Losers
Transcript
- Lucas: The U.S. Treasury auctions about twenty-four trillion dollars in debt every year. And the strange thing is, the average American taxpayer is paying a small but real premium because of how those auctions are designed. Luna: When you say 'small but real' — what are we talking? A few cents per person, or something more meaningful? Lucas: More like a few basis points on every bond the government issues. A 2023 Treasury report estimated that non-competitive bidders — that includes retail investors buying TreasuryDirect — left about two hundred fifty million dollars on the table that year because they systematically overpaid. Luna: Wait, I thought the whole point of Treasury auctions was to let the market set the price. How can anyone systematically overpay if it's an auction? Lucas: That's the puzzle. The Treasury runs what's called a uniform-price auction for most of its notes and bonds. Every winning bidder pays the same price — the highest yield that clears the auction. But non-competitive bidders don't submit a yield — they just say 'I want this much' and get whatever the auction settles at. And because they don't compete on price, they often end up paying more than a competitive bidder would. Luna: So they're essentially price-takers in a market where the professionals are strategic. That sounds like a recipe for leaving money on the table. Lucas: Exactly. And the amount adds up. In the first half of 2026 alone, the Treasury auctioned about two point three trillion dollars in marketable debt. If non-competitive bidders overpay by even half a basis point on average, that's over eleven million dollars — directly from individual investors to the Treasury. Luna: But isn't that a good thing for the Treasury? It lowers the government's borrowing cost, which is good for taxpayers. Lucas: In theory, yes — but only if you ignore the distributional effect. The people who buy Treasuries directly are typically individuals who want a safe, simple investment. They're not hedge funds. They're retirees buying bonds through TreasuryDirect. And the people who benefit from lower borrowing costs are all taxpayers, including the wealthy who don't buy Treasuries directly. So it's a regressive transfer — from small savers to the general Treasury pool. Luna: So the auction design itself creates a hidden tax on retail investors. How did we get here? Was it always this way? Lucas: No. Before the 1990s, the Treasury used a discriminatory auction format — also called multiple-price. In that system, winning bidders paid their own bid price. So if you bid a yield of 4.5 percent and the auction cleared at 4.4, you'd still get 4.5. That sounds better for the bidder, but it actually discouraged aggressive bidding because of the winner's curse — if you bid too high a yield, you'd get stuck with a low price while everyone else got the clearing yield. Luna: Right, the winner's curse. In that format, the winning bidders are the ones who overestimated the true value. So they'd bid conservatively. Lucas: Which lowered the amount the Treasury raised. In 1992, the Treasury switched to uniform-price for two-year and five-year notes, and eventually for all maturities. The idea was that uniform pricing would encourage more aggressive bidding because you wouldn't be penalized for bidding at the market-clearing level. And it worked — the Treasury estimates the switch saved about three to four basis points on average. Luna: But that savings came at the expense of non-competitive bidders, who now pay the clearing price instead of a potentially lower price from the discriminatory format. Lucas: Right. In the discriminatory system, non-competitive bidders also paid the average of winning bids — which was usually lower than the clearing yield in a uniform auction, because the average was pulled down by low-yield bids. The Treasury didn't care because, from its perspective, uniform pricing raised more money. But it shifted the cost onto the least sophisticated participants. Luna: And this isn't ancient history. There was a rule change in 2014 that raised the limit for non-competitive bids from five million to ten million dollars per auction. That made it even easier for retail investors to pile in at the wrong price. Lucas: Good catch. And that's exactly when the volume of non-competitive bids started growing. In 2013, non-competitive bids averaged about two hundred million per auction. By 2025, they averaged over four hundred million. The Treasury also introduced floating-rate notes in 2014, which are even harder for retail investors to value correctly. Luna: So the system has gotten more complex, and retail investors are being asked to participate in a game designed for professionals. Is there a simple fix? Could the Treasury just charge non-competitive bidders the average of winning competitive bids instead of the clearing yield? Lucas: That would be one option. The Australian Treasury does something similar — non-competitive bidders there pay the weighted average yield of accepted competitive bids. That would probably reduce the overpayment. But the Treasury has resisted because it would slightly lower auction proceeds — maybe a half basis point or so. In a year like 2026, that's about one hundred twenty million dollars in lost revenue. Luna: That's real money, but it's also a rounding error in a four-trillion-dollar budget. And it would make the system fairer. Lucas: Exactly. And there's another issue: the role of primary dealers. These are the big banks — JPMorgan, Goldman Sachs, Citigroup — that are required to bid in every auction. They get privileged access to the auction and often trade Treasuries in the secondary market before the auction results are announced. That gives them an information advantage. Luna: They know the order flow. They can see whether demand is strong or weak before the auction closes. Lucas: And they can adjust their bids accordingly. A 2021 study by the Federal Reserve Bank of New York found that primary dealers' bids in the final hour of an auction are significantly more accurate than those of other participants. They're essentially trading with a better map. Luna: So the auction is tilted not just against retail, but also against smaller institutional players like pension funds or regional banks. Lucas: Definitely. And the tilt has gotten steeper since the Fed started quantitative tightening in 2022. When the Fed was buying Treasuries, it absorbed a huge chunk of supply. Now it's letting them roll off, so the private market has to absorb more. That increases auction volatility, which makes the informational advantage of primary dealers even more valuable. Luna: And retail investors are the least equipped to handle that volatility. They might see a rising yield and think 'great yield' — but they don't realize they could get a better deal in the secondary market. Lucas: That's the real irony. For most individual investors, it's actually cheaper to buy Treasuries on the secondary market through a broker than to go through TreasuryDirect. On the secondary market, you can see the current yield and bid accordingly. On TreasuryDirect, you get whatever the auction clears at. Luna: So the government's own platform for retail investors is the worst place to buy government debt. That's a policy failure. Lucas: It is. And the Treasury knows it. In 2023, they launched a pilot program to allow retail investors to submit competitive bids through TreasuryDirect, but it's still limited. As of July 2026, only about five percent of TreasuryDirect users have access to competitive bidding. The rest are stuck with the non-competitive route. Luna: Speaking of which — a couple of dollars a month is genuinely what keeps these episodes going. If walking through the economy with us has made something click, buy me a coffee dot com slash fexingo. It's a small way to keep the podcast independent and ad-free. Lucas: Yeah, that's right. And to be clear, we're not asking for anything big — just the price of a coffee now and then. It genuinely makes a difference in whether we can keep doing deep dives like this. Luna: Alright, back to the auction mechanism. One overlooked detail is that the Treasury actually sets a non-competitive allocation limit per auction — usually about thirty-five percent of the offering amount. But that's rarely binding because non-competitive bids are such a small share. Lucas: Right. And the limit isn't there to protect retail — it's to ensure that the auction price is set by competitive bidders. If non-competitive bids dominated, the price would be set by the least informed participants. That would actually increase the cost of borrowing for the Treasury. Luna: So the system depends on professionals setting the price, even though those professionals have an informational edge. It's a necessary evil. Lucas: But maybe not necessary. The Treasury could level the playing field by publishing indicative bids in real time — like a transparent order book. That would reduce the information advantage of primary dealers and give retail investors a better sense of where the auction is heading. Luna: Has any government tried that? Lucas: The UK's Debt Management Office publishes a 'gilt auction calendar' and a 'gilt market outlook' that includes expected supply. But they don't reveal live bids. Sweden's National Debt Office runs a more transparent auction for its inflation-linked bonds, but it's a much smaller market. No major government has fully opened the order book. Luna: Probably because the primary dealers would oppose it — they benefit from the opacity. Lucas: Exactly. And the Treasury doesn't want to alienate the dealers because they're the ones that absorb supply in tough auctions. It's a classic regulatory capture problem. Luna: So where does that leave the retail investor? Is the advice just 'don't use TreasuryDirect'? Lucas: For now, yes. If you have a brokerage account — Fidelity, Vanguard, Schwab — you can buy Treasuries on the secondary market with a few clicks and see the exact yield you'll get. You're also protected by best-execution rules. On TreasuryDirect, you're flying blind. Luna: And yet TreasuryDirect has over five million accounts. That's a lot of people who might be unwittingly overpaying. Lucas: It is. And the total overpayment across all non-competitive bidders is small in absolute terms — maybe fifty million a year — but it's a pure structural inefficiency that could be fixed with a simple rule change. The question is whether the Treasury sees it as a priority. Luna: Given that the Treasury's mission is to borrow at the lowest cost, they might not want to fix it. But from a fairness standpoint, it's hard to defend. Lucas: No, it's not. And it's a reminder that even in the most transparent markets, the design of the auction itself can create hidden winners and losers.