SIFIs.
On the one hand, it is a bit surprising that a federal judge yesterday rejected the Financial Stability Oversight Council's attempt to classify MetLife as a "systemically important financial institution" subject to enhanced prudential regulation by the Federal Reserve.
Classifying and monitoring too-big-to-fail firms was in many ways the centerpiece of post-crisis financial reforms, and judicial review of FSOC's designations was supposed to be limited and deferential. Certainly MetLife is big and owns lots of financial assets, and, while it came through the crisis more or less fine, its fellow giant insurer, American International Group … did not.
On the other hand, it isn't that surprising. A "systemically important financial institution" can't just be, you know, a big company with a lot of money. Apple has a lot of money. Vanguard, good lord, has so much money. (More than $3 trillion under management, versus $878 billion of assets at MetLife.)
Any sensible SIFI designation can't just be about bigness and money; it has to be about whether a company's financial distress could damage the broader financial system. (Or, as section 113 of the Dodd-Frank Act puts it, whether "material financial distress at the U.S. nonbank financial company, or the nature, scope, size, scale, concentration, interconnectedness, or mix of the activities of the U.S. nonbank financial company, could pose a threat to the financial stability of the United States.")
In some vague sense that is true of any big company — if Apple disappeared tomorrow, things would get weird — but it seems to me that the real damage comes from companies that could be subject to runs. (Here again I follow Morgan Ricks, Gary Gorton and others in thinking about financial panics as primarily a problem of short-term debt, broadly defined.)
That's why the Fed regulates banks: because banks are preeminently in the business of funding long-term assets with short-term debt (deposits), a model that we have long known creates the risk of instability. But the last crisis featured runs on companies — investment banks, asset-backed commercial paper conduits, AIG — that weren't banks, but used similar funding models without the prudential regulation and capital buffers of banks.
On that theory, FSOC's job is really to subject runnable non-banks to prudential regulation by the Fed. That job is in some ways less important than it sounds, just because a lot of the runnable non-banks from the last crisis have disappeared on their own: The big investment banks have become, or been bought by, Fed-regulated banks; asset-backed commercial paper has vanished.
But AIG's still around! So it is natural that the FSOC would start with AIG, and would, reasoning by analogy, figure that other big insurers are also systemically important.
But the argument that MetLife is particularly run-prone seems pretty weak. MetLife is an insurance company. As it said in its briefing in this case:
Like other conventional life insurers, MetLife writes long-term policies and invests premiums in long-term assets to satisfy its obligations when they come due. Unlike banks, which take liquid, short-term deposits and wholesale funding and invest in long-term assets such as commercial loans, MetLife matches its long-term liabilities with long-term assets, and its investment portfolio is linked to, and driven by, the profile of its insurance liabilities.
You don't see a lot of runs on life insurance policies; though I suppose it is possible. (MetLife argues that the FSOC relies on the "irrational and ahistorical supposition that distress at MetLife would lead its customers to surrender their policies en masse–despite evidence that customers do not hold insurance contracts for liquidity purposes and would face adverse tax consequences and contractual penalties associated with surrender.")
But who can say in advance what will be subject to runs? AIG is an insurance company too, but its policies were not the problem in 2008. The problems came from its securities lending business, which does have characteristics of short-term financing, and from its derivatives business, which kind of doesn't, but which did feature run-like demands for collateral as AIG's credit deterioriated.
(The FSOC's ruling designating MetLife as a SIFI points to its derivatives liabilities and securities lending business as potential sources of exposure, though they don't seem to be as significant as AIG's were.) And while the FSOC seems to have given up on regulating the big asset managers as SIFIs, there has been a lot of talk recently about the risk of a run on mutual funds — which was not really something people thought about a few years ago.
In any case, it is hard to know what yesterday's decision means, because the judge issued it under seal; a possibly redacted public version will be issued after April 6. Perhaps the judge found some minor flaw in the FSOC's work, and it can revise and resubmit and make MetLife a SIFI again.
Or perhaps the FSOC will win on appeal. But MetLife's win might make FSOC regulation of mutual fund activities more difficult. It also makes it a little weird that MetLife already announced that it will break itself up, partly to avoid the SIFI designation. That now seems a bit hasty.
In other too-big-to-fail news, here is Euromoney on Neel Kashkari's "radical and very public campaign to end too-big-to-fail," which it calls "the monetary equivalent of trolling."
Asset management.
Elsewhere in just-barely-non-SIFI news, BlackRock "plans to cut about 400 jobs in what may be the biggest round of layoffs to date at the world's largest money manager." But BlackRock is fine: It "attracted $54 billion in net new money in the fourth quarter," and "expects to end the year with a higher headcount" despite the layoffs. The hedge fund industry has perhaps bigger worries:
Pension funds, insurers and university endowments helped pump up hedge funds to a record $3 trillion in assets over the last decade. But with results falling behind a more traditional mix of stocks and bonds for six straight years and the high-fee structure now politically sensitive in some states due to uneven results, many of them are pulling back.
From New Mexico to New York, big investors are dramatically reducing their commitments and opting for cheaper imitations. Investors globally asked for more money back from hedge funds than they contributed in the fourth quarter of 2015, according to HFR Inc.—the first net quarterly withdrawal in four years. They pulled an additional $15.3 billion in this year's first two months, according to eVestment.
And at least some of them are using "liquid alternatives" instead, to get sort-of-hedge-fund-like exposures at much lower costs. It seems pretty reasonable that a lot of the work that hedge fund managers do could be replicated by computers, though I guess it would be more flattering if those computers were, like, Go-playing neural nets rather than systematic investing strategies.
Speaking of hedge fund managers and neural nets:
Two hedge fund "quants" have come up with an algorithm that diagnoses heart disease from MRI images, beating nearly 1,000 other teams in one of the most ambitious competitions in artificial intelligence.
They both actually seem to be former hedge fund employees (at Crabel Capital and Two Sigma). You sometimes hear complaints that in modern financial capitalism people with the talents to cure cancer end up programming computers to trade stocks instead. But on the other hand, it's possible that the lure of financial-industry riches creates an environment in which a lot of smart ambitious young people are trained to solve problems using big data and artificial intelligence. And then they can go cure cancer in their spare time.
