Money Stuff: Save Havens May Not Be

For as long as I can remember, US Treasury bonds have been the world’s preeminent safe-haven asset. When scary stuff happens in financial markets, investors sell their risk assets and buy Treasuries. This has been true to a paradoxical extent. When scary stuff happens to the creditworthiness of the US government, investors sell their risk assets and buy Treasuries. When ratings agencies downgrade the US, people buy Treasuries. When it looks like the US Treasury might run into its debt ceiling and default on its debts, people buy Treasuries. When Treasuries get riskier, investors flee from risk, to Treasuries.

This is extremely ingrained in market expectations, but it is not a law of nature. It could change. The US government could find a way to make Treasuries risky in a way that causes investors to flee from Treasuries. I do not have a well-worked-out mechanism for how it could do that. Evidently saying “hey FYI we might default on Treasuries next week” does not work. Perhaps literally explicitly saying in so many words “hey we do not want Treasuries to be safe haven assets anymore, and we are going to enact wrenching far-reaching economic policy to make that happen” would work? I don’t know. We might find out.

Yesterday I quoted a speech by the chairman of President Donald Trump’s Council of Economic Advisers, Steve Miran, uh, saying that? A sample:

The U.S. provides the dollar and Treasury securities, reserve assets which make possible the global trading and financial system which has supported the greatest era of prosperity mankind has ever known. ...

The reserve function of the dollar has caused persistent currency distortions and contributed, along with other countries’ unfair barriers to trade, to unsustainable trade deficits.  These trade deficits have decimated our manufacturing sector and many working-class families and their communities, to facilitate non-Americans trading with each other. …

That trade entails savings housed in dollar securities, often Treasurys. As a result of all this, Americans have been paying for peace and prosperity not just for themselves, but for non-Americans too. ...

Reserve status matters and, because demand for the dollar has been insatiable, it has been too strong for international flows to balance, even over five decades.

Miran’s speech justifies Trump’s blanket tariffs as a way to correct these imbalances, by increasing US manufacturing and exporting, and by reducing global demand for Treasuries. You do not have to believe that, on any number of levels. You can feel free to believe that Trump will walk back the tariffs, that the tariffs will not have the desired effects, or that the desired effects are bad. But one way to read the tariff policy is that the status of US Treasury securities as the global reserve asset is undesirable, and that the tariffs will fix that. 

Anyway here’s Bloomberg’s Richard Henderson:

A lackluster run for haven assets during one of the most acute market selloffs in years has left investors seeking new forms of protection.

US Treasuries plunged Wednesday as Donald Trump’s reciprocal tariffs took effect, sending the benchmark 10-year yield up by more than 10 basis points to its highest level since February. Gold rose but remains down for the week, during which global equities touched a one-year low. The dollar also has weakened.

Just as Trump’s latest tariff war pushes global trade into uncharted territory, financial markets also are scrambling for answers to questions about the role of assets that typically shield investors during crises. While some observers point to the likes of German bunds and Japan’s currency as potential new shelters, the candidates also face risks from liquidity to their own economic and monetary policy outlooks.

And the Financial Times adds:

Treasuries sold off on Wednesday as President Donald Trump’s tariffs took effect, deepening investor concern about the “safe haven” status of US sovereign debt.

The 10-year US Treasury yield jumped to 4.51 per cent before falling back to 4.45 per cent — up 0.19 percentage points on the day — while the 30-year yield briefly rose above 5 per cent. The 10-year yield has risen from less than 3.9 per cent earlier this week.

The move dragged government borrowing costs around the world higher, with yields in the UK and Japan climbing sharply.

The moves offer a new challenge to the Trump administration, which had previously cited lowering Treasury yields as a policy aim, and could mark a loss of investor confidence in the world’s largest sovereign debt market.

“The sell-off may be signalling a regime shift whereby US Treasuries are no longer the global fixed-income safe haven,” said Ben Wiltshire, a rates strategist at Citi.

Yes lowering Treasury yields is also a policy aim of the Trump administration [1] ; not everything needs to be consistent.

One consequence of US Treasuries being the classic safe haven asset is that you can borrow a lot of money against Treasuries. In particular, hedge funds apparently do the basis trade — buy Treasuries, fund them in repo markets and sell Treasury futures — at leverage ratios of 50 or 100 to 1. In times of market dislocation, you know. If you have 100 to 1 leverage and your position moves against you by 1%, you have blown up. The basis trade is not supposed to move against you much. “One model,” I wrote recently, “is that some trades want to be done with a lot of leverage”; buying Treasuries and selling more-or-less-precisely offsetting Treasury futures is one of them. People worry.

We talked about this yesterday, and my view at the time was (1) people are worried, (2) those worries seem reasonable but (3) there is not exactly clear evidence of big blowups and dislocations yet. I guess I am still there, but here’s some new stuff:

• Bloomberg’s  Edward Bolingbroke and Michael Mackenzie note that “the upheaval from President Donald Trump’s tariffs is accelerating the collapse of a popular hedge-fund bet that Treasuries would perform better than interest-rate swaps,” but add that “the trade had been losing momentum since February, in part on waning expectations for an imminent move by the Trump administration to loosen bank regulations and allow lenders to keep more Treasuries on their balance sheets.” The theory of the basis trade is that, for many market participants, it is cheaper or easier to get synthetic leverage by owning Treasury futures than it is to get real leverage by owning Treasury bonds and borrowing against them. Big hedge funds can own Treasury bonds and borrow against them cheaply, so they manufacture the futures, owning the bonds and selling the futures to investors who are more constrained. Classically the constrained investors are long-only asset managers who want to make leveraged interest-rate bets. But banks are also constrained; holding Treasuries on their balance sheets is expensive. If it became cheaper, there would be more demand for Treasuries and less demand for swaps, which would make existing basis trades — long Treasuries, short swaps — more valuable. People expected that to happen due to deregulation, but now they expect it less, so the basis trade was less attractive even before the impact of tariffs.
• Bloomberg’s Tracy Alloway has a good explainer of the basis trade in historical context, also noting that in recent months it has been “in effect a deregulation and duration trade.” She adds: “So far, the deleveraging looks okay-ish.”
• Liz Capo McCormick and Mackenzie write that “there’s little concrete evidence of dealers cutting off financing or hedge funds getting caught wrong-footed thus far,” but “basis trade deleveraging has played at least some role in pushing long-end yields higher in recent days.”
• Bloomberg also reports: “The Bank of England said hedge funds have faced ‘significant’ margin calls from their prime brokers as they navigated extreme market volatility in the aftermath of US President Donald Trump’s tariff announcements and warned that the risk of ‘further sharp corrections’ remains high. While the central bank’s Financial Policy Committee found that so far those firms had been able to meet margin calls, it warned that the overall global risk environment has deteriorated, according to minutes from meetings it held on April 4 and April 8.”
• At FT Alphaville, Robin Wigglesworth notes that “the basis trade has become such a major pillar of support for the Treasury market, at a time when the US government’s borrowing costs have already ballooned,” and adds that “so far it doesn’t seem like any basis trade liquidation is having a major disruptive effect on the Treasury market.”

So I think the overall view is that there has been some deleveraging, but no huge dislocations. Scott Bessent agrees:

Treasury Secretary Scott Bessent played down a selloff in US Treasuries, saying that there was nothing systemic at play, and also served warning against China not to attempt to devalue its exchange rate in retaliation for American tariff hikes.

“There’s one of these deleveraging convulsions that’s going on right now in the markets,” Bessent said on Fox Business, adding that he’d witnessed those very often in his hedge-fund career. “It’s in the fixed-income market. There are some very large leverage players who are experiencing losses, that are having to deleverage.” ...

“I believe that there is nothing systemic about this — I think that it is an uncomfortable but normal deleveraging that’s going on in the bond market,” Bessent said.

It would be a little weird if an economic move of this magnitude doesn’t cause any financial blow-ups, but so far so, uh, uncomfortable but normal.

I have written a couple of times that there will eventually be lawsuits against Trump’s tariffs. The legal basis for the tariffs is pretty slim: These tariffs are based on a law that has never been used to impose tariffs before, and that requires a declaration of a “national emergency” to deal with an “unusual and extraordinary threat,” which is a weird way to characterize the last 50 years of global trade. (How can everything that has happened in the last 50 years be unusual?) It seems plausible that a court might reject the tariffs (not legal advice), and with trillions of dollars at stake surely someone will sue. In fact we talked this week about a small company, supported by a “powerful legal group backed by conservative funding,” that did sue.

On the other hand, if you sue to block the tariffs Donald Trump will probably post mean things about you on social media, and possibly find other ways to punish you, so. The Wall Street Journal reports:

Businesses are contemplating a risky strategy to fight tariffs: suing President Trump.

Amid the scramble to beat back the tariffs, the Chamber of Commerce and other top industry groups are discussing whether to file a lawsuit, according to multiple people familiar with the conversations. …

Trump-aligned lobbyists have warned that speaking out publicly against the president could only cause blowback—and make him dig in further on his controversial tariff strategy. Suing him would provoke an even harsher response, they have said.

The talks highlight the lack of options available for many businesses that oppose tariffs. On Capitol Hill, lawmakers are sympathetic to business concerns but this week there is little appetite to publicly buck the president. …

“Lawyers seem to be in consensus that this is illegal,” said Consumer Technology Association CEO Gary Shapiro, who declined to comment on the possibility of his group joining a lawsuit. “There will be lawsuits. And Congress will be forced to act,” he said. 

And Bloomberg reports that the Retail Industry Leaders Association decided not to bring a lawsuit over the tariffs, “even though the group’s research indicated a legal case had a good chance of succeeding on the merits,” in part because of “the potential challenge of finding law firms willing to bring suits over tariffs in light of Trump’s attacks on some of the biggest names in the legal profession.”

I guess I would summarize the position as “we all agree that the president is doing something that is both illegal and bad for the country, but we are afraid to say that publicly.” Seems bad!

Broadly speaking the way that quantitative hedge funds work is that they use machine learning models to find stocks that will go up. Sometimes the way this works is that a researcher will have some idea for some intuitive signal — “maybe a new chief executive officer’s golf handicap is correlated to the medium-term performance of the company’s stock” or whatever — and test to see if it works, and if it does then the signal goes into the fund’s trading model. Sometimes, though, the machine learning model just finds some extremely complicated pattern in the data that seems to predict prices, and that no human can understand or even hold in their mind. “If these 600 data points look like this, then these 12 stocks go up 53% of the time,” that sort of thing. Why? They just do.

There are reasons not to rely on signals like that — if you can’t explain them, how real are they? — but there are also reasons to prefer them. I am fond of quoting something Robert Mercer of Renaissance Technologies said to Sebastian Mallaby: ““The signals that we have been trading without interruption for fifteen years make no sense. Otherwise someone else would have found them.”

As artificial intelligence models get more complicated and capable, and as AI agents are increasingly able to take actions in the real world, a funny yet plausible sort of signal would be: “If we manipulate this stock, it will go up.” Like if you are training your AI model by rewarding it for finding stocks that go up, and if you don’t really understand how it predicts stocks that will go up, eventually it might get it into its little AI head that it can make the stocks go up by doing some sort of subtle market manipulation, and that this is what you want. And then maybe it will do it.

Is it what you want? I dunno, maybe. One model might be “everyone would do market manipulation if (1) it worked and (2) you didn’t get in trouble for it.” Perhaps the AI will find a market manipulation that works. (This is hard.) And perhaps:

• It will be so subtle and complex that no one will notice it; or
• If a regulator does notice it, the regulator will come to you and say “your AI is doing market manipulation,” and you will (accurately!) say “what, I had no idea, I don’t really know what the AI gets up to, sorry about that,” and the regulator will be like “that’s fine AI is pretty complicated” and you won’t get in trouble. Not legal advice!

Anyway:

The Bank of England plans to closely monitor the use of artificial intelligence by banks and hedge funds over concerns that the technology could trigger a market crash or manipulation without humans even knowing about it.

The central bank’s Financial Policy Committee warned that the technology could destabilize markets or act in other adverse ways in a new report on AI published Wednesday. It added that AI was making such rapid headway among hedge funds and other trading firms that humans may soon not understand what the models are doing.

Some firms are already experimenting with autonomous neural networks that “may not be well understood by risk managers at the firm” and could result in “unpredictable behavior,” the bank said in the report.

“Models with sufficient autonomy could act in ways that are detrimental to the overall stability or integrity of markets, for example by ignoring regulatory or legal guardrails such as market abuse regulations,” according to the report. “Human managers would also need to manage such regulatory risks.”

Obviously one point here is that, from the regulators’ perspective, you shouldn’t be able to disclaim responsibility if your AI does market manipulation. Who is responsible, if not the human managers?

Here is the report, which makes other interesting points. One that I like:

Greater use of AI to inform trading and investment decisions could help increase market efficiency. But it could also lead market participants inadvertently to take actions collectively in such a way that reduces stability. For instance, the potential future use of more advanced AI-based trading strategies could lead to firms taking increasingly correlated positions and acting in a similar way during a stress, thereby amplifying shocks. Such market instability can then affect the availability and cost of funding for the real economy.

One simple model that I have of hedge funds is that they engage in scientific research to find the stocks that will go up. This is a rigorous, truth-seeking, somewhat collaborative enterprise done by highly qualified people, and so, just as in real science, you should expect them to be good at finding the correct answers. Then all the hedge funds will buy all the good stocks (the ones that will go up) and avoid the bad ones (the ones that will go down). And then, because this model is only approximately true — there is no such thing as a “stock that will go up” as an objective fact of nature — it will lead to increased risk, because all the hedge funds will own the same stocks, and if something goes wrong at one fund there will be contagion to the other ones.

This is of course an approximate and jocular model, and you would get roughly similar results if you replaced “the hedge fund analysts engage in rigorous, truth-seeking, high-quality scientific research to find the good stocks” with “the hedge fund analysts all have the same background and biases and all converge on the same stocks out of groupthink.” (Presumably good performance by big hedge funds would support the first model.)

Anyway if you replace “hedge fund analysts” with “AI models” you get similar issues. If the AI’s training incorporates human biases and groupthink, you will get herding and “increasingly correlated positions” in ways that seem straightforwardly bad. But if AI is just really really good at picking the right stocks, everyone will pick the same right stocks, and then what will happen in times of stress?

I wrote in 2021 that “it’s only a slight overstatement to say that, every year, every publicly traded oil company asks its shareholders to vote on a proposal” calling for it to write a report about climate change. For a long time, the way corporate governance worked in the US was that activist shareholders asked all the oil companies to write reports about climate change, and the oil companies groused, and there was a shareholder vote, and it was all a bit theatrical.

When I wrote that, I was referring specifically to ExxonMobil Corp., which did get those proposals every year, but in 2021 it faced a different form of climate-flavored activism: A small fund, Engine No. 1 LLC, was running a proxy fight, criticizing some of Exxon’s business and climate decisions and asking shareholders to elect some of its candidates to Exxon’s board of directors. Unlike the usual report stuff, this would be a binding vote: If Engine No. 1 won, its directors would join the board. It did, and they did. 

The Engine No. 1 thing was climate-flavored, but only just. Engine No. 1 is economically motivated, and its goals in the proxy fight were not, like, “let’s shut down Exxon’s fossil fuel business.” In fact, since 2021, Exxon became rather more aggressive in fighting back against the climate-report stuff. We talked last year about a lawsuit that Exxon brought against two other investment firms who had submitted another nonbinding shareholder proposal asking it “to go beyond current plans, further accelerating the pace of emission reductions in the medium-term.” Exxon more or less succeeded with that lawsuit, and the shareholders withdrew the proposal.

More broadly, between 2021 and 2024, the overall environment for climate-related shareholder activism has shifted. In 2021, Engine No. 1 moved beyond the typical nonbinding shareholder proposals, using climate-related arguments to enlist big shareholders in a binding proxy fight to change the direction of Exxon. It looked like climate-related shareholder activism was becoming more powerful, more effective, more able to shake up boardrooms. In 2024, Exxon swatted away a climate-related nonbinding shareholder proposal with extreme force. It looked like climate-related shareholder activism might be dead.

Anyway yesterday Bloomberg’s Kevin Crowley and Saijel Kishan reported:

Exxon Mobil Corp. is facing no shareholder proposals this proxy season for the first time in at least 25 years.

The absence of requests follows a year after the oil company sued two climate-focused investors to remove what it described as their “extreme agenda.” It also comes as the US Securities and Exchange Commission released guidelines making it easier for corporations to block votes on shareholder resolutions at their annual meetings.

Exxon said in a statement late Monday that it received only one proposal this year and the SEC agreed it should be discarded because “it tried to micromanage the company.” In 2024, investors voted on four resolutions, including one that linked Exxon’s executive pay to cutting greenhouse gas emissions and another related to plastics production.

We talked about those new SEC guidelines earlier this year and I guess they are working. My impression was always that these nonbinding proposals were mostly theatrical, so it is not exactly a huge substantive development that Exxon doesn’t have any of them this year. But they were symbolic, and their absence does seem symbolically important.

Trump’s Tariffs Put Fed Chair Powell in a ‘No-Win Situation.’ Big Banks’ Record Trading Streak Overshadowed by Tariff Upheaval. Dalio Sees Once-a-Lifetime Collapse in Economic, Political Order. US consumers face sharp price increases from Donald Trump’s tariffs, say analysts. Dimon Says Recession Is the ‘Likely Outcome’ of Tariff Upheaval. Big Bank CEOs Reckon With Their Lack of Influence on Trump. The Dip Buyers Braving the Market During Stocks Carnage. Quant hedge fund Renaissance suffers steep losses in tariff tumult. Bain Capital clinches $2.6bn deal to buy Blackstone-backed HealthEdge. BlackRock Backs Adani’s $750 Million Bond to Fund An Acquisition. How Elon Musk Rescued X From the Brink. Musk Skips Deposition in Investor Lawsuit Over Twitter Buyout. Market turmoil strains bro-sphere’s bromance with Donald Trump. MrBeast attacks Trump’s ‘brutal’ tariffs, claims it’ll be ‘way cheaper’ to make his Feastables chocolate bars overseas. A Masters Tradition Unlike Any Other: Gifting Tickets to Government Officials. The Skin on Mysterious Medieval Books Concealed a Shaggy Surprise. “Statistically, the worst baseball game of all time.”

If you'd like to get Money Stuff in handy email form, right in your inbox, please subscribe at this link. Or you can subscribe to Money Stuff and other great Bloomberg newsletters here. Thanks!