Programming note: Money Stuff is off for the next two weeks, Elon permitting. We’ll be back August 25.

Historically one way to save for retirement is that you work at a company for a few decades, and the company pays you a salary, and then you retire, and the company keeps paying you a pension. The company has an obligation to pay you a certain amount every month, and it funds that obligation by putting money aside and investing it to make sure it has enough money to pay you. If it invests the money really well, it will have more than it needs to pay you, and it can keep the extra. If it invests the money badly, it will have less than it needs, and will need to kick in more money to pay you what it owes you. That outcome is bad, not only for the company (which has to kick in more money), but also for you, because it exposes you to credit risk. There is a law in the US, called ERISA (the Employee Retirement Income Security Act), which requires companies to manage their pension funds prudently, as fiduciaries for their retirees, so they don’t lose all the money.

Another way to save for retirement is that you work at a company for a few decades, and the company pays you a salary, and you save a portion of that salary yourself and invest it, and then you retire, and the company pays you nothing — you don’t get a pension — but you spend down the money that you have invested. If you invested really well, you will have more than you need to live on. If you invested badly, you will have less than you need. But this is your problem, not your employer’s. If you want, you can invest all of your money in meme stocks or sports gambling to get a wide dispersion of possible outcomes, some of them quite bad.

But the paradigmatic way to save for retirement, in America in 2025, is a third, in-between option, the 401(k) plan. In this method, you don’t get a pension; your company doesn’t keep paying you a fixed amount after you retire. Instead, as in the second option, you put aside some money every year to invest. (And the company might put some extra money into your investments.) But, unlike in the second option, you can’t generally invest the money however you like; you can’t put it into sports gambling or your brother-in-law’s vending-machine business. Instead, your company gives you some menu of investing options, and you choose among them. [1] And the company is a fiduciary and has a duty to give you a prudent menu of investment options. And if one of those options is “if you want we will take your 401(k) assets to Vegas and put them all on red,” and you choose that one and the ball lands on black and you lose all of your retirement savings, you will definitely sue your employer and you’ll definitely win. (Not legal advice.)

That is: Individuals are responsible for their own retirement savings, but not entirely. Their employer has some fiduciary duties to guide their investing and not let them go too crazy. This, too, is part of ERISA, [2] and it seems sort of historically contingent. In the olden days, companies offered pensions and had to manage them prudently, so they developed expertise in prudent retirement investing. Modern retirement saving involves individual workers making investment decisions, but the companies still have some expertise at prudent retirement investing, while the individual workers sometimes, you know, enjoy a little sports betting. So the companies have some paternalistic role in guiding the workers’ investment decisions and trying to keep them from losing all their money. And if the companies don’t take that role seriously, they’ll get sued.

The main point of all of this is of course “they’ll get sued.” There are shifting norms about what sorts of investments are prudent and thus can be offered in 401(k)s. In the 19th century — long before 401(k) plans or ERISA — courts sometimes found that investing in any common stocks (as opposed to, say, government bonds or debt secured by mortgages) was imprudent, so a fiduciary could be sued for doing it. In the 1970s, when index funds were first being marketed, some employers stayed away, worrying that they would violate their fiduciary duties by investing in all of the stocks without doing prudent due diligence on each of them.

By 2015, though, I was writing about “a congealing regulatory view” — specifically at the US Department of Labor, which is in charge of ERISA — “that index funds are Good and should be encouraged, and that active management is Bad and should be discouraged.” The logic was:

1. Actively managed long-only stock funds tend not to beat the index consistently, and
2. They charge much higher fees than index funds, so
3. It’s probably more prudent to index than to hire active managers.

This is not a universal view, and companies still regularly offer actively managed mutual funds in their 401(k)s, but there is some pressure to offer index funds and a lot of pressure to keep fees down. Everyone understands that investments have risks, and if a fund on the 401(k) menu loses money that is not necessarily the employer’s fault. But if a fund on the 401(k) menu charges twice as much in fees as some other nearly identical fund, that does seem imprudent, and you can get sued for that.

More recently, environmental, social and governance investing (ESG) became a 401(k) battleground. We talked in January about how American Airlines Group Inc. got in trouble for doing ESG with its 401(k). American didn’t even offer ESG funds in its 401(k) plan — just straightforward low-cost index funds — but those funds were managed by BlackRock Inc., which used to talk a lot about ESG, and a judge in Texas found that ESG was not a prudent thing to do with employees’ money.

But these days the main things that people talk about, when they talk about what should or shouldn’t be in your 401(k), are:

1. Private equity,
2. Private credit and
3. Crypto.

And the essential question is not whether you should be allowed to invest in those things, but rather whether you should be able to sue your company for letting you invest in those things, if they lose money. Because the 401(k) system involves both individual investment choices and employer paternalism, and because you can sue your employer if it imprudently lets you invest in something that loses money, employers tend to offer only investment options that are clearly “prudent” under current norms. Index funds, in 2025, are clearly prudent. ESG funds, in 2025, are a little dicey. But what about privates and crypto? Are those prudent options for your employer to put on the 401(k) menu?

I mean the obvious answer is “last year they were not prudent, but the norms changed and now they are.” [3]  That’s not, like, a financial answer. It’s not that private equity and crypto were volatile and charged high fees in 2024, but now there have been structural changes such that they are safe and cheap. It’s that in 2024 the US federal government was pretty skeptical of crypto and of high-fee opaque investments in 401(k)s, and in 2025 a different federal government loves that stuff. Now that shift is getting formalized:

President Donald Trump will sign an executive order Thursday that aims to allow private equity, real estate, cryptocurrency and other alternative assets in 401(k)s, a major victory for industries looking to tap some of the roughly $12.5 trillion held in those retirement accounts.

The order will direct the Labor Department to reevaluate guidance around alternative asset investments in retirement plans subject to the Employee Retirement Income Security Act of 1974, according to a person familiar with the plans who requested anonymity before the order was formalized. The department will also be tasked with clarifying the government’s position on the fiduciary responsibilities associated with offering asset allocation funds that include alternative holdings. ...

Top officials in Washington have been weighing a directive for months that would ease the legal concerns that have long kept alternative assets out of most worker defined-contribution plans. Retirement portfolios are mostly concentrated in stocks and bonds in part because corporate plan administrators are reluctant to venture into illiquid and complex products. ...

Alternative and traditional asset managers are eager to grab a slice of the defined-contribution market, which they see as the next frontier of growth. Many institutional investors like US pension funds and endowments have reached internal limits of what they can put into private equity amid a broader slowdown in dealmaking and a lack of distributions to clients.

From first principles, it makes sense to put illiquid private assets in 401(k)s: The whole point of a 401(k) really should be that you put the money in and don’t take it out until you retire, so you shouldn’t need liquidity, and if illiquid assets pay a premium you should be the one to earn it. Also, private markets are the new public markets: As Byrne Hobart points out, “the set of companies owned by PE plus public markets looks a lot like the set of companies that would all have been public thirty years ago,” so if it is prudent for 401(k)s to own stocks then it’s arguably prudent for them to own private equity too. 

From second principles, “the financial industry wants to sell stuff to individual investors because it can’t sell any more of that stuff to institutional investors, and because the fees are high,” is really the worst imaginable advertisement for an investment. When the most sophisticated asset managers are desperate to sell you stuff, you should consider that perhaps you should not be buying the stuff, though of course this is not investment advice. [4] When crypto managers are desperate to sell you stuff, that also might be some sort of signal.

That said, I am not sure how much this is a story about the rise of private assets (and crypto), and how much it is a story about the decline of 401(k) paternalism. In general, outside of 401(k)s, Americans have far more, and far spicier, investment options than they did a few years ago. You could always buy stocks, but now you can buy meme stocks, which are sillier. You can buy zero-day options. You can get all sorts of private credit. You can buy 10x levered perpetual crypto futures on Coinbase. You can, and I really cannot stress this enough, bet on sports in your brokerage account. You can’t yet buy tokenized shares of OpenAI, but give that like a month.

Led by crypto and meme stocks and legalized sports gambling, the general norms about what sorts of investments regular people make, and should be allowed to make, and might enjoy making, have shifted. There is broader acceptance that of course you should be able to make all sorts of wild bets in your investment accounts, that a little excitement is a good feature of financial markets, that it is a good function of the financial system to offer you some entertaining bets. And if that’s true, then the 401(k) system is going to offer you some of those bets too.

By the way: You can still buy low-cost broad public stock index funds! [5]  You can do that in your individual brokerage account — it’s the button next to “Bet On Sports” — and you will probably be able to do it in your 401(k) for the foreseeable future. But your 401(k) plan might offer you a few additional, zanier options. And if you choose those and they don’t work out, that’s your problem.

An important stylized fact is that sell-side research analysts can move stocks. When a well-known analyst at a big firm upgrades a stock from Hold to Buy, the stock will probably go up. It is possible that there is a fundamental explanation for this — the analyst is very good at identifying which stocks will go up, so her upgrades correctly predict future cash flows — but there is also, obviously, a social-influence explanation. People follow her recommendations, so when she says that a stock is a Buy, people buy it, so it goes up. Her recommendations are, over some time horizon, self-fulfilling. 

And therefore sophisticated investors who do their own analysis also pay attention to sell-side research analysts. And getting early access to research notes is insider trading, even though the analyst is not actually an “insider.” Her recommendations are market-moving in their own right.

Two things that I have mused about in recent weeks are (1) whether people are asking ChatGPT for stock recommendations and (2) whether it’s giving them all similar recommendations. If the answers are yes and yes — if people increasingly are getting their financial advice from a handful of generative artificial intelligence models, and if the models give people the same advice — then the models’ recommendations should be, over some time horizon, self-fulfilling. If everyone asks ChatGPT what stocks to buy, and ChatGPT tells them all “OpenDoor,” then OpenDoor will go up.

This can’t be literally true, in part because people don’t all ask ChatGPT the same questions, and it doesn’t give them all the same answers. But here’s a fun little website called Agentic Equities that tries to track ChatGPT recommendations:

Here are the stocks ChatGPT is telling people to buy. We query GPT-4o-search using multiple analyst personas for the S&P 500 and select stocks. 

The analyst personas include “value investor,” “risk averse,” “quant/technical,” “meme investor,” “macro strategist,” “growth focused” and “balanced analyst,” and you can read the little memos ChatGPT writes for each of them. If ChatGPT is telling all of those personas to buy a stock, does that mean the stock will go up? Man, I have no idea, and this is not investment advice. But it does feel a little bit like the future.

For a while it looked like President Donald Trump’s tax bill might include a “revenge tax.” This tax, in Section 899 of the draft bill, would have imposed withholding tax on certain US income earned by foreigners from countries that have “unfair foreign tax” rules. So if a citizen of those countries — which apparently included Canada, the UK, most of the European Union and many other countries — owned US stocks or bonds, the US government would take as much as 20% of their dividend or interest income. This ended up not happening, but it was a worrying possibility for a while.

During that time, one reader emailed me to ask: Would this be a default on US debt? The logic is:

1. The US issues Treasury bonds, which pay interest.
2. Many of those bonds are owned by foreigners.
3. If the US seized the interest on those bonds for itself, rather than sending it to the (foreign) bondholders, that does seem like not quite the deal the bondholders signed up for.
4. Isn’t that a default?

I think the rough answer is “no,” in part because the Section 899 tax probably wouldn’t have applied to Treasuries and in part because the US, as a bond issuer, can kind of do whatever it wants; if it’s a “default,” it is not a default in any actionable way. (There is no bond contract setting out events of default, and the bondholders are not going to accelerate their bonds, sue to get them paid, seize US assets abroad, etc.)

But it is a more generally applicable question in sovereign debt. A country issues bonds. The bonds pay interest. [6] The country decides it no longer wants to pay the interest. If it simply says “we are no longer paying interest,” that’s a default. The bondholders might have some legal remedies. Emerging-market bonds, in particular, are often issued under New York or English law and waive sovereign immunity, so the bondholders really can go to court in New York or London and demand payment and get some sort of legal order that makes life unpleasant for the issuer. 

But what if the country does something else with a similar effect? What if it passes a new tax, saying “the tax rate on interest paid on our bonds to foreign holders is now 100%”? Then it could “pay” the interest to the bondholders (not a default), but take all of it back in tax. But people have thought of that; most sovereign bond contracts have a tax gross-up clause saying that, if the issuing country taxes any payments on the bond, it has to pay those taxes rather than imposing them on the bondholders. [7] So that doesn’t work, for most countries, though it might work for the US.

But you could be more creative. What about a law like “any payments on our bonds will be deemed made if we deliver them to our central bank in an account with the holder’s name on it”? That way, the company has “made” the payment, but no cash has gone out the door to investors. Or “any payments on our bonds can be made in this new cryptocurrency we just made up”? There are ways to change your laws to change how the payments are made.

Does this work? Well, it shouldn’t. The whole point of issuing sovereign bonds under New York or English law is to prevent this sort of thing: The issuer’s obligations are governed by some outside law, so it can’t just change its own laws to avoid paying.

But here is a fun paper by sovereign-debt legal experts Lee Buccheit and Mitu Gulati about a glitch. Most English-law sovereign bond contracts say something like this:

All payments are subject in all cases to any applicable fiscal or other laws, regulations and directives in the place of payment, but without prejudice to the provisions of [the tax gross-up clause].

Buchheit and Gulati explain:

The clause is a simple acknowledgment that a jurisdiction other than the jurisdictions in which the issuer and the investors are located may enact laws or regulations affecting payments under the instrument. The most likely measure of this kind would be some form of transaction tax on payments flowing through that jurisdiction or perhaps a politically motivated sanction on the issuing country. The Russian Federation, for example, was effectively prevented by EU sanctions from making payments on its Eurobonds following Russia’s invasion of Ukraine in 2022.

Under this clause, it is not a default on the bond if the issuing country makes a payment to investors, but the country where the payment is received does something to prevent or tax the payment. The investors take the risk of a change in law in other countries, though not the risk of a change in law in the issuer country.

The glitch is that some bond contracts (roughly a quarter of English-law bonds) don’t have the normal wording:

In our perusal of Cypriot bonds in 2013, however, we stumbled on the fact that the Fiscal Laws clause appearing in Cypriot bonds omitted the codicil “in the place of payment.” The formulation of the clause used by Cyprus said that all payments are subject in all cases to any applicable fiscal or other laws, regulations and directives. The important geographical qualification “in the place of payment” was absent. Does this mean, we asked, that Cyprus itself could enact laws, regulations and directives affecting is own payments under the bonds? …

Read literally, a Fiscal Laws clause that omits the “in the place of payment” codicil evidences the consent of the investors to the sovereign borrower enacting laws or regulations that will affect how the issuer can discharge its payment obligations under the instrument. This could arguably include the timing of payments, the place of payments, even perhaps the currency of payments. The one measure that would not be authorized under this provision is any form of withholding tax or similar levy imposed by the sovereign issuer’s own jurisdiction. The Fiscal Laws clauses we have examined (all of them) could not be used to override the obligation of the issuer to “gross up” (that is, to make the investors whole) for any tax of that kind. …

What might such a measure look like? Consider, for example, the situation of a country facing a severe — perhaps existential — need to conserve its foreign exchange. A situation brought on, perhaps, by the need to repel a brutal invasion by its neighbor. [8]  Confronted by that necessity, a plausible fiscal measure might require payments under external debt obligations to be discharged by crediting the amount due to an account in the name of the foreign creditor at the local central bank.

A number of countries that restructured their external debt in the 1980s employed programs of this kind covering foreign currency debts of both public and private sector obligors. The Philippines and Mexico, for example, each instituted a program that allowed private sector borrowers in those countries to discharge their foreign currency denominated debt obligations by paying the local currency equivalent to the country’s central bank (which then assumed the obligation vis-à-vis the foreign creditor on restructured terms). In other countries, public sector borrowers settled their debts to foreign creditors by paying the amount due to the local central bank which then restructured the debt on agreed terms. 

That is, if your sovereign bond contains a clause saying “all payments are subject in all cases to any applicable fiscal or other laws,” you can just make some new applicable laws to change the payments.

Okay:

A former Point72 intern says he asked the firm for a quieter desk — and got fired instead. Now he's suing Steve Cohen's $39.9 billion hedge fund.

Andrew Pardo, who was in the firm's 2023 summer intern class after completing his junior year at the University of Michigan, claims in his suit that he was fired after he asked to move to a desk in a less-trafficked area because he suffers from PTSD, or post-traumatic stress disorder, a mental health condition he said he was diagnosed with after suffering physical abuse. …

"This complaint, for which the plaintiff is demanding $20 million in damages in connection with his summer internship, is ridiculous and without merit. We intend to address these matters in the appropriate forum," the hedge fund wrote in a statement.

There almost certainly is some hedge-fund employee somewhere for whom the correct desk location really is worth $20 million, but probably not an intern.

Stagflation Concerns Ripple Through Wall Street as Tariffs Hit. In Land of 25% Inflation, Crypto Is Starting to Replace Money. Waller Emerges as Favorite for Fed Chair Among Trump Team. Microsoft Raids Google’s DeepMind AI Unit With Promise of Less Bureaucracy. Developer of Crypto Mixer Tornado Cash Found Guilty on One Criminal Charge. Cryptocurrency group Ripple buys stablecoin platform in $200mn deal. UK crypto investors hail