A while back, Seth Ackerman — a writer for the radical left magazine Jacobin — published a long, detailed, and bracing article calling for the socialization of the financial industry. Then, over the weekend, he got a surprising (albeit inadvertent) endorsement from Berkshire Hathaway CEO Warren Buffett.

In his annual letter to shareholders, the octogenarian investor extraordinaire gave a big conceptual boost to Ackerman's proposal.

To explain: Berkshire Hathaway is a conglomerate, an umbrella corporation for a whole bunch of different companies and stock holdings. It's into insurance, railroads, electric utilities, underwear, private jets, the works. Academic theory says that's a really inefficient way to do things. The interior of a firm, after all, is not a market. It's a centralized bureaucracy. It may respond to external market pressure, but all internal decisions are made by bureaucratic management. And in the case of a conglomerate, those decisions include allocating capital between all the various companies that make it up.

Conversely, with a normal company, it just does one thing, and the financial markets allocate capital between it and all the other companies doing one thing. Markets are supposed to be more efficient at this.

Yet Berkshire Hathaway is one of the most successful American companies of the last half-century, and its stock has performed admirably for decades. So what gives?

In his letter, Buffett put forward a few ideas about the disadvantages of the market approach, which Berkshire Hathaway's conglomerate structure vastly reduces or eliminates.

First off, there are overhead costs: lots of individual companies are all going to have individual boards of directors, individual regulatory costs, administrative costs, etc.

Next, there are transactions costs: all the lawyers and bankers and consultants and more who are needed to move capital in the financial markets. "Money-shufflers don't come cheap," as Buffett noted.

Finally, there are a lot of political and social frictions in financial markets: "We are free of historical biases created by lifelong association with a given industry and are not subject to pressures from colleagues having a vested interest in maintaining the status quo," Buffett wrote. "That's important: If horses had controlled investment decisions, there would have been no auto industry."

Another aspect of Buffett's management style is ruthlessness about one key decision: entry and exit. As long as a company is doing well, Buffett will keep supporting it. Otherwise, not. And he stays out of managing the individual companies under him. Buffett argues this approach is attractive to families or owners who are emotionally invested in their business, want to sell it, but don't want to see it dismantled by shareholders. It's also arguably attractive for management, which gets to run its company free of meddling.

What's remarkable is how closely all this mirrors Ackerman's case for just having the government buy up every last bit of stock on the financial markets and run the whole thing as a public trust.

One long-held argument for capitalism is that markets and price signals allocate resources more efficiently than centrally planned economies. But a range of studies during the Cold War, comparing western capitalist countries to Soviet economies using standard neoclassical measures of efficiency, found minimal advantages for the former.

These findings then forced economists to explain why capitalism produces so much more prosperity than socialist or communist systems. The answer they came up with was that, under capitalism, firms are autonomous and they can experiment. Any company or business can form anytime it wants, and if its solution to some human problem is effective, it will make a profit and capital will be allocated to it. If its idea is ineffective, it won't receive capital and it will be allowed to fail.

So by both Buffett and Ackerman's lights, it appears that, at least under some circumstances, centralized bureaucracy can move resources around as intelligently as markets. How would it look in practice?

Well, a number of states already have sovereign wealth funds (SWF), which are just government-run investment portfolios that generate revenue, including the deeply red Texas, Alabama, North Dakota, Louisiana, and Alaska. That last one actually takes a good portion of its investment income and just hands it back out to Alaskan residents as a per capita check. The state enjoys a poverty rate that's 33 percent lower than America as a whole, and one of the lowest levels of income inequality.

So we could give every state an SWF, and then expand them until they've covered the entire American financial market. (Or half of it if you're feeling cautious.) That diversity would probably be preferable to just having the federal government run one big pot for the whole country.

Management of the funds could be appointed by the executive (state or federal), then confirmed by the legislature, both to make sure the process is accountable to democratic will and to put a firewall between daily political pressure and the daily running of the funds. The funds' mandates and bylaws could require both Buffett-style hands-off management and his ruthlessness with regards to when you provide a firm more capital and when you don't. You would want to pay the managers of these funds very well to attract talent, but you would also want to link their pay to fund performance. Hell, you could even provide politicians in the legislature a regular bonus indexed to the funds' performance, just to make sure everyone has an in-built incentive to manage the fund appropriately.

The ultimate goal is a humane capital structure that broadly benefits society as a whole, but also maintains economic realism about when to let a failing firm die, even if it will cause job dislocation and even if local communities are emotionally invested in its survival. The idea is just that the tension between these two approaches would be diminished, through a kind of hybrid market-socialist system. Meanwhile, you could still handle social insurance, basic aid, unemployment benefits, job retraining, monetary policy, labor law, and the rest through the same avenues our government does now.

Is this idea a little out there? Well, sure. But it's not like private financial markets have been performing admirably or even usefully as of late.

Buffett obviously didn't intend this conclusion from his argument, and he would probably consider this all radical crazy talk. But maybe he and Ackerman should at least get together for a beer.