Fed profitability datapoint of the day

By Felix Salmon
January 10, 2011
end of 2010, the Federal Reserve system had $2.423 trillion in assets and $2.367 trillion in liabilities, which means that the simplest measure of its total equity -- assets minus liabilities -- comes to $56.6 billion.

" data-share-img="" data-share="twitter,facebook,linkedin,reddit,google" data-share-count="true">

At the end of 2010, the Federal Reserve system had $2.423 trillion in assets and $2.367 trillion in liabilities, which means that the simplest measure of its total equity — assets minus liabilities — comes to $56.6 billion. The Fed also managed to earn net income of $80.9 billion in 2010. Which means that its return on assets was incredibly high at 3.3%, while its return on equity was an astonishing 143%.

I think it’s fair to say that no bank in the history of the world has ever had income of anywhere near $80 billion in one year: that’s over $700 per US household. Somehow, the Fed is making roughly $60 per household per month, and remitting that money straight to the Treasury. Of course, its cost of funds is ridiculously low, and in any event the Fed can simply print new money any time it wants. But still, $80 billion is enormous — more than four times the Fed’s profit in 2004, for instance. And it’s a useful reminder of just how massive the Fed’s balance sheet has become — and also of how monetary policy can make a serious dent in the funding-need side of fiscal policy.


We welcome comments that advance the story through relevant opinion, anecdotes, links and data. If you see a comment that you believe is irrelevant or inappropriate, you can flag it to our editors by using the report abuse links. Views expressed in the comments do not represent those of Reuters. For more information on our comment policy, see http://blogs.reuters.com/fulldisclosure/2010/09/27/toward-a-more-thoughtful-conversation-on-stories/

That’s what comes of insider dealing

Posted by ottorock | Report as abusive

What is their cost of capital? Zero?

Easy to make large amounts of money if you own the printing presses! Of course they don’t exactly count the new money as income, they merely rent it out…

Posted by TFF | Report as abusive

a) As you and Ben have assured us, the Fed is not printing money (at least not $100 bills)

b) Score one for ottorock

Posted by ARJTurgot2 | Report as abusive

It’s also worth remembering, though, that if quantitative easing works better this time, the inflation expectations will cause the values of those long-duration assets to drop, and some of them will probably be sold at significant losses when the expansion of the fed’s balance sheet is unwound. Keep in mind to mentally offset this against the first $80 billion in losses that the fed recognizes. (On the other hand, if this ability to fund fiscal policy is countercyclical, perhaps even in the absence of its monetary-side effect the simple business cycle PnL understates its value.)

Posted by dWj | Report as abusive

To expand on @dWj point, I wonder how these numbers will play out politically if conservative economist fears come true and the Fed’s independence is threatened because of balance sheet losses (forcing it to go to Congress). Will the Fed bring up the point that it has paid money to the treasury every year from profits?


Posted by Chris_Gaun | Report as abusive

“if quantitative easing works better this time, the inflation expectations will cause the values of those long-duration assets to drop, and some of them will probably be sold at significant losses when the expansion of the fed’s balance sheet is unwound”

So will the Fed hit Treasury with an enormous losses at the time of unwinding? Will the Fed ever unwind its balance sheet? Does it have a legal obligation to unwind? What are bounds of what the Fed is allowed to do?

Folks at Zero Hedge: “Oh you sweet, dear, fool did you think there was actually going to be an unwind?”

Posted by DanHess | Report as abusive

For some reason I can’t load comments #3-#5…

To answer DanHess, I expect the “unwind” to be a very slow process over many years (if it happens at all). They can likely unwind through the steady maturity of their assets.

Has anybody seriously suggested a more rapid “unwind”?

Posted by TFF | Report as abusive

If the Fed holds bonds to maturity and then gets a bunch of cash at that time, shouldn’t it ‘destroy that cash’ if it is being honest? (In the same way that it was wished into existence)? If it remits that cash to treasury, then we have pure monetization.

For the Fed to remit its earnings to the treasury is already pure monetization.

Posted by DanHess | Report as abusive

All of the Fed’s earnings of course are stolen funds, returns that would have accrued to savers if capital had been scarcer during the crash as economics would have dictated, if new cost-free Fed capital hadn’t been created. The ends are supposed to justify the means, and we don’t know what the ends will be yet.

I do know that if you are stealing capital as the Fed is, it darn well better be an emergency. QEII was definitely not an emergency.

Posted by DanHess | Report as abusive

DanHess, the cash that the Fed remits to the Treasury represents the *interest* on the bonds they purchase, not the bond principal. When the bonds are redeemed, the outstanding liabilities are reduced and the cash is effectively destroyed (at least until it is spent again to purchase a new bond).

It is a little strange that we don’t have clearly stated goals for QE (other than juicing the economy and increasing employment). Some goals I have variously guessed:

* Push China to allow a more appropriate relative valuation between the yuan and the dollar. With massive QE, they will face high inflation as long as they keep the link tight. (Essentially the Cold War strategy in which the US spent the Soviet Union into the ground.)

* Devalue the dollar to make American labor more competitive in a global market. Dropping wages by 30% to 50% would be devastating for anybody with a loan and would not share the pain between workers and retirees. Better to inflate the currency while wages stagnate.

* Punish investors who have fled to “safe” investments. You can still put half of your money in Treasuries, but it won’t be worth anything by the time they mature. Maybe the stock market is looking tempting again?

* Encourage Congress to enact fiscal stimulus by directly monetizing the new debt.

Posted by TFF | Report as abusive

A system capable of generating $80B in profits is also capable of generating $80b in losses. The response to such an outcome may however by slightly different.

If the fed were to show losses of $80b, the general public and international response may reach a more appropriate level of inquiry about the processes generating such returns.

The asymmetry of response reflects the common problem with risk and psychological bias. No one questions variance and its causal source when positive. Soul searching only starts after negative variance shows up, at which point many act surprised that any such thing could happen even after the same mechanism and process showed such extremes before.

Posted by Nick_Gogerty | Report as abusive

Nick_Gogerty, so lets say my system is to take 8trillion stick it in the bank at 1% and 80bn interest then how can that generate 80bn in losses?

Would be good to know the duration on these holdings. I seem to vaguely recollect the fed was going to hold the GSE MBSes to maturity. Would love to know how Nick thinks that can generate a loss.

Posted by Danny_Black | Report as abusive

My service business generates varying profits but can never generate a loss — I have essentially zero fixed expenses.

The Fed’s book is a little like that. Their ‘cost of funds’ is literally zero, so ANYTHING they receive is pure profit. The only way they can ever lose money is if the bonds they purchase default, and don’t they restrict their holdings to Treasury securities and others with a government guarantee?

Okay, so they could theoretically lose money if forced to sell into a down market. But they aren’t ever forced to sell and they don’t practice “mark to market”. So I don’t think this is terribly likely.

Posted by TFF | Report as abusive

TFF, you are of course assuming they hold to maturity.

Posted by Danny_Black | Report as abusive

True, Danny_Black. But how are their maturities structured? They have typically stayed short in the past — and while they are definitely buying up longer-term assets now, I believe the intent remains to hold them to maturity. Do you expect circumstances will force them to sell ahead of that?

Posted by TFF | Report as abusive

Janet Yellen would certainly have you believe that they will be sold – over roughly a 5 year period starting as soon as mid 2012. I still remain convinced that the intention IS to sell the bonds down, just as I remain convinced that the Fed has no idea quite how hard this will be in practice.

Posted by drewiepe | Report as abusive

Absolutely agree on that, drew.

It will be very hard to sell longer-term bonds in an environment with rising interest rates. Even harder in an environment with persistently high unemployment.

This is the report you refer to:
http://www.bloomberg.com/news/2011-01-10  /yellen-speech-may-offer-proxy-for-plan ned-unwinding-of-fed-qe.html

Any idea where we can look for a peek at the Fed’s book? Most interested in the maturity dates.

Posted by TFF | Report as abusive

Never mind, I found this one:
http://www.federalreserve.gov/releases/h 41/current/h41.htm#h41tab2

The MBS holdings are all longer-term securities. But of the $1T in US Treasury obligations, about 60% have a maturity of five years or less.

Posted by TFF | Report as abusive

When you control the money is it any wonder that you make a profit?

Of course the truth is that they win even if all those assets on their book vaporize because the American people would be the one’s on the hook then.

Posted by goldtracker | Report as abusive