By Simon Johnson

Goldman Sachs is investing $450 million of its own money in Facebook, at a valuation that implies the social networking company is now worth $50 billion.  Goldman is also apparently launching a fund that will bring its own high net worth clients in as investors for Facebook.

On the face of it, this might just seem like the financial sector doing what it is supposed to – channeling funds into productive enterprise.  The SEC is apparently looking at the way private investors will be involved, but there are some more deeply unsettling factors at work here.

Remember that Goldman Sachs is now a bank holding company – a status it received in September 2008, at the height of the financial crisis, in order to avoid collapse (for the details, see Andrew Ross Sorkin’s blow-by-blow account in Too Big To Fail.)  This means that it has essentially unfettered access to the Federal Reserve’s discount window, i.e., it can borrow against all kinds of assets in its portfolio, effective ensuring it has government-provided liquidity at any time.

Any financial institution with such access to such government support is likely to take on excessive risk – this is the heart of what is commonly referred to as the problem of “moral hazard.”  If you are fully insured against adverse events, you will be less careful.

Goldman Sachs is undoubtedly too big to fail – in the sense that if it were on the brink of failure now or in the near future, it would receive extraordinary government support and its creditors (at the very least) would be fully protected.  In all likelihood, under the current administration and its foreseeable successors, shareholders, executives, and traders would also receive generous help at the moment of duress.  No one wants to experience another “Lehman moment.”

This means that cost of funding to Goldman Sachs is cheaper than it would be otherwise – because creditors feel that they have substantial “downside protection” from the government.  How much cheaper is a matter of some controversy, but estimates made by my co-author James Kwak (in a paper presented at a Fordham Law School conference last February) put this at around 50 basis points (0.5 percentage points), for banks with over $100 billion in total assets.

In private, I have suggested to leading people in the Obama administration and in Congress that the “too big to fail” subsidy be studied and measured more officially and in a transparent manner that is open to public scrutiny, for example as a key parameter to be monitored by the newly established Financial Stability Oversight Council.  Unfortunately, so far they have declined to take up this approach.

However, there is consensus that the implicit government backing afforded to Fannie Mae and Freddie Mac in recent decades allowed them to borrow at least 25 basis points (0.25 percent) below what they would otherwise have had to pay; this is a significant difference in modern financial markets.  In 13 Bankers we refuted the view that these Government Sponsored Enterprises were the primary drivers of subprime lending and the 2007-08 financial crisis – that debacle was much more about extreme deregulation and private sector financial institutions seeking to take on crazy risks.  But it is still the case that Fannie and Freddie were badly mismanaged – and followed the market in 2005-07 with bad bets based on excessive leverage – in large part because they had an implicit government subsidy.  Those institutions should be euthanized as soon as possible.

Goldman Sachs now enjoys exactly the same kind of unfair, nontransparent, and dangerous subsidy; it has effectively become a new form of Government Sponsored Enterprise.  Goldman is not a venture capital fund or primarily an equity-financed investment fund.  It is a highly leveraged bank, meaning that it borrows through the capital markets most of the money that it puts to work.

As Anat Admati (of Stanford University) and her colleagues tirelessly point out, the central vulnerability in our modern financial system is excessive reliance on borrowed money, particularly by the biggest players.

Goldman Sachs is a perfect example.  Most of this firm’s operations could be funded with equity – after all, it is not in the retail deposit business.  But issuing debt is attractive to shareholders because of the subsidies associated with debt funding for banks, and compelling to bank executives whose compensation is based on return on equity – as measured, this increases with leverage.  If they have more debt relative to equity, that increases the potential upside for investors.  It also increases the probability that the firm could fail – unless you believe, as the market does, that Goldman is too big to fail.

Social networking firms should be able to attract risk capital and compete intensely.  They do not need subsidies in the form of cheaper funding (seen today as a more favorable valuation for Facebook) or in any other form.

Social networking is a bubble in the sense that email was a bubble.  The technology will without doubt change forever how we communicate with each other, and this may have profound effects on the nature of our society.  But investors will get carried away, valuations will become too high, and some people will lose a lot of money.

If those losses are entirely equity-financed, there may be negative effects but they will likely be small – in the revised data after the 2001 dotcom crash, there isn’t even a recession (i.e., there were not two consecutive negative quarters for GDP).

But if the losses follow the broader Goldman Sachs structure and are largely debt-financed, then the US taxpayer will have helped create another major financial crisis.

And if you think that sophisticated investors at the heart of our financial system can’t get carried away and lose money on Internet-related investments, read up on Webvan:

“During the dot-com bubble, Goldman invested about $100 million in Webvan, the online grocer that never got off the ground and eventually collapsed in bankruptcy.”

An edited version of this post appears today on the NYT’s Economix blog; it is used here with permission.  If you would like to reproduce the entire post, please contact the New York Times.

By James Kwak

Ezra Klein points out a new tax expenditure database from The Pew Charitable Trusts. More attention to tax expenditures — exceptions in the tax code that reduce tax revenue or, put another way, subsidies channeled through the tax system* — is always a good thing. But Klein also says something interesting that I don’t agree with:

“they’re basically the welfare state for the middle class, cleverly arranged such that they don’t look like the welfare state for the middle class. If every year, the government sent every American — from the richest CEO to the greenest public-school teacher — a check covering 30 percent of their health-care costs, we’d think that a bit weird. We’d think it much weirder if we only sent the checks to the workers who happened to be at firms that offered benefits. . . .

“Yet that’s pretty much exactly what we do. We just hide it in the tax code rather than write it on a check.”

I agree with all of that, except the bit about the middle class. Tax expenditures primarily benefit the rich, for a few reasons.

• First, most of them are deductions from taxable income, which means their value to you is proportional to your marginal tax rate. Not only do rich people have higher marginal rates than middle-class people, but remember that many middle-class people pay exactly zero in income taxes. (The average tax rate for people in the middle income quintile is 2.3 percent.) This is true, for example, of the deduction for employer-provided health insurance.
• Second, because most of them are itemized deductions, you only get them if you itemize your deductions–which usually means either that you have a big enough house to have a big mortgage or you have enough income to pay a lot of state and local taxes.
• Third, the size of the deduction is highly correlated with income. To take the most obvious example, rich people have bigger houses, and so they have bigger mortgages. (In another universe, they might buy the same size houses and pay for them with cash — but that’s not our universe.)
• Fourth, there are the tax expenditures that you get on investments, which are disproportionately held by the rich. There’s the tax exemption for life insurance investments. There’s the big one: tax-advantaged investment accounts, including 401(k)s, IRAs, Roth IRAs, 529s, etc.  This is even leaving aside the question of whether lower tax rates for capital gains and dividends count as tax expenditures.
• Fifth, there are the tax expenditures for businesses, since in theory those flow to their shareholders, who are disproportionately the rich. (I say “disproportionately” because someone with a net worth of $1 million has much more than ten times the capital market investments of someone with a net worth of $100,000.)

Tax expenditures are basically exhibit A for my pet theory of the tax code. And there should be broad support for getting rid of them. But there isn’t, because the median household does get some benefit from the mortgage interest tax deduction. They just don’t realize how much more benefit rich households are getting from it.

* You can have a debate about what is or is not a tax expenditure. You can even have a debate about whether the whole concept makes sense, since the identification of something as a tax expenditure presumes some baseline in which that particular loophole doesn’t appear. But the latter debate is, in my opinion, philosophical at best. It’s pretty clear that the tax code includes some basic principles (like taxing income) and some exceptions (like not taxing a portion of your income equivalent to the amount of mortgage you pay on your house and your vacation house); the latter are tax expenditures.

By James Kwak

Or perhaps a leading candidate for the Republican presidential nomination.

When it comes to deficits and government spending, the strategy of Republicans in Congress is to assert things that are simply not true or that defy economic logic. Ezra Kleinnails House Majority Leader Eric Cantor misstating the CBO’s ten-year projection for health care reform so he can make a false claim about its longer-term effects (ignoring the fact that the CBO explicitly said health care reform would be deficit-reducing in the second decade). The same Republican leadership that rails against deficits is introducing rules that will make it easier to increase the deficit, since tax cuts will no longer have to be paired with offsetting spending cuts.

Apparently, the ability to say things that are not true is something that is learned quite early.

My four-year-old daughter is currently enamored of a series of fairy books. (If they have not entered your house yet, do not let them in; bar the doors, do not accept packages, do whatever you need to do.) In these books, the fairies are good, and the goblins are bad. We had read about seven of these books, and one feature of them was that although the heroines (two girls named Kirsty and Rachel who are completely interchangeable because they have no personality) and the fairies are afraid of the goblins, the goblins had shown themselves completely incapable of doing any harm to anyone. As I said to my daughter, “The goblins have never caught a fairy.”

Alas, in the next book, the goblins did catch the fairy (although she subsequently escaped). So I said, “We can no longer say that the goblins never caught a fairy.”

The next day, my daughter came to me and said: “We can still say the goblins never caught a fairy. Just watch. ‘The goblins never caught a fairy.’ See? You can still say it!”

She has a bright political future ahead.