The cost of bailing out Frannie

By Felix Salmon
December 16, 2010

Let’s say I buy a $3,000 pair of handmade shoes but don’t have that kind of cash to hand, so I put them on my credit card. I then rack up another $2,000 in penalties and interest before I’ve paid them off. Then the total cost associated with my poor investment in footwear is $5,000. The credit card company bailed me out but charged through a lot of money for doing so, and the money is absolutely part of the total sum I end up paying for those shoes.

Weirdly, Jeffrey Goldstein, the under secretary for domestic finance at Treasury, doesn’t seem to think that way. Fannie and Freddie have already borrowed $151 billion from Treasury, and they’re set to borrow another $90 billion by the end of 2013. That’s hardly chump change. Yet Goldstein says, with a straight face, that “the GSEs have already absorbed the vast majority of costs associated with the poor investments they made during the housing boom”. His argument:

Under the baseline scenario, FHFA projects that $90 billion in additional draws will be necessary through 2013. But this is why accounting for dividends is important – $71 billion of those additional draws will be used to pay dividends back to Treasury. This means that nearly 90 percent of total GSE losses have already been absorbed, since these future draws will primarily be returned through dividend payments.

This just doesn’t make any sense to me. Every business has money coming in, from various sources, and money going out, to various sources. If at the end of the day you end up having to borrow $90 billion from Treasury, then that means the money going out exceeds the money coming in by $90 billion. Most of us would consider that a “loss”. But not Goldstein. The way he sees it, if $71 billion of the money-going-out is going to Treasury, then it doesn’t count towards the GSEs’ total loss. But why is Treasury special in that way? If the interest payments were going to anybody else, then they would count towards Frannie’s losses.

Goldstein is right that from Treasury’s point of view, the net amount of money being pumped in to the GSEs is decelerating, even if the total cost of the GSE bailout is still going up rather than down. And it’s probably nice to be getting $71 billion in dividend payments from the agencies, even if that money has to be turned around and sent straight back again. What’s more, to be fair to Treasury, if it just reduced the interest rate on its bailout funds, it could reduce the headline cost substantially, even if the actual net cost to taxpayers went up.

But in the context of a world where all the other bailouts — even AIG, amazingly — look set to be paid back in full, the fact that we’re still pumping billions of new dollars into Frannie only serves to underline how massive and disastrous the agencies’ failure was. I’m glad they’re still around to make their interest payments. But the only reason they’re still around is because Treasury has promised them unlimited funds — a promise which means they can never go bust, no matter how much money they end up losing.


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

But Frannie is the only pair of bailed-out companies that had all their eggs in the housing basket. AIG was diversified, GM was diversified. Perhaps a better question to ask (and a more fair comparison) would be to look at the assets of Countrywide that BofA bought in 2008 and ask if they are still worth $4 billion (or if they even have positive net value).

All in all, Fannie and Freddie didn’t do too bad, they appear to have lost 150B and they expect to pay 90B of interest to the Treasury, out of 7 trillion dollars worth of loans. House prices fell in half in many places, and they only lost 2% to 3% (depends on whether you want to count interest) on top of the capital they had when times were good.

Posted by Nameless | Report as abusive

I don’t see why Fannie and Freddie should pay 10% interest when TARP charged the big banks much less. As a first approximation, if one assumes 5% interest, these 10% payments would pay off a loan in around 14 years. On the other hand, Treasury seems to be implying that the 10% “dividends” will cancel out the loan in 10 years.

Posted by bob33 | Report as abusive

bob33, from recollection GS paid around 11% for the roughly 6 months it had TARP money. Which means it was paying a coupon of around 20% a year and last time I checked 20 > 10.

Nameless, not sure you are reading it right. Through to 2013 they are expected to need 150+90bn and pay 71+16.4 in dividends and interest. Thats around 150bn losses right there.

PS Mr Salmon thx for the focus on Frannie…

Posted by Danny_Black | Report as abusive

Nameless, if you look at the document from the FHFA – ctions_102110.pdf – Frannie from 2009 to 2013 are expect to make credit losses from 131bn[low-end] to 254bn[high-end]. And thats not including what was pumped in during 2008.

Posted by Danny_Black | Report as abusive

“But the only reason they’re still around is because Treasury has promised them unlimited funds”

While true, the reason they are still around is because they serve a critical role in the housing market. It is for that reason that they were established initially and for that reason that they were bailed out despite heavy losses.

The interest payments on the “borrowing” help to ensure that the shareholders don’t profit at the expense of the Treasury. If the Treasury ends up losing money on this, then the shareholders will necessarily be wiped out.

Posted by TFF | Report as abusive

TFF, not sure there is that much lower to go than 30 cents… The bailout was for the creditors, a large percentage of the paper was owned by China and Russia – the stood to lose hundreds of billions had Frannie gone under.

Posted by Danny_Black | Report as abusive

Thanks for the correction, Danny_Black.

Where do the Treasury loans stand with respect to those creditors?

Posted by TFF | Report as abusive

Bit after my time so not au fait with the exact technicalities but basically the understanding is that any shortfall that Frannie have in servicing their debt will be made up by the taxpayer. It was a pretty major transfer of wealth from the US taxpayer to the Chinese and Russian SWFs – who were threatening very very very loudly to pull out of the US debt markets if the GSE’s paper was not honoured – as the paper was trading pretty rich, reflecting concerns about their health. From recollection foreign investors held just under a trillion dollars of paper before the conservatorship and again from recollection China was holding around half of that. Put another way virtually all the money the Chinese used for their stimulus over the last two years more or less would have gone up in smoke!!

Maybe someone more up to date can give precise details..

Posted by Danny_Black | Report as abusive

254bn of losses on 7 trillion of loans are still less than 4%.

In contrast, Countrywide had 200bn of assets and a mortgage book of 1.3 trillion or so, as of the end of 2006. By the time the dust settled, its assets evaporated, and people holding bonds and MBS of Countrywide took substantial haircuts.

Posted by Nameless | Report as abusive

Nameless that 254bn is on top of 150bn. No matter how you try and spin this Frannie has lost more than was **spent** – and mostly repaid with profit – by TARP. Thats why the relative lack of focus on them is simply astonishing. I have read people get worked up about 3bn **loan** to Israel but somehow there is no anger about the US taxpayer saving 20% of China’s FX reserves.

Posted by Danny_Black | Report as abusive

Upon closer look at the article, I’m not sure where the 254bn figure comes from at all, but figure 6 (scenario 3) shows projected treasury draw of 363bn _including_ all the money already given to Frannie, less dividends of 104bn projected to be paid back to the treasury by the end of 2013, with the net loss of 259bn (and falling) as of 12/31/2013. The median forecast is the loss of 154bn. Net loss will eventually get back to zero, and at some later point Frannie will be cut loose again.

I’m not arguing about the relative focus, my point is mostly that I dislike talking about agencies’ “massive disastrous failure” when they lost significantly less money per dollar loaned than virtually any private lender in the market during the bubble, and the big part of the reason why they lost what they lost is because the government essentially promised to make good on their obligations, and therefore all losses came out of their own pockets rather than out of pockets of people who held their bonds and MBS.

I’m sure that sovereign funds hold a lot of Frannie bonds, but I’m not sure that they have the majority. Domestic investors probably hold the majority. There was a massive rush into bonds long before Frannie looked to be in any danger.

And I’m not sure that it’s all about saving China’s FX reserves, because, in all likelihood, if we’d let Frannie crash rather than loan them a few hundred billion, mortgage interest rates would’ve been above the 7% mark now, house prices would’ve been correspondingly lower, the recession would’ve been quite a bit deeper, tax receipts would’ve been lower, and the net effect on the state of the treasury balance sheet would’ve been negative.

Posted by Nameless | Report as abusive

Nameless, the best case scenario is that 2009-2013 Frannie will have credit losses of 247bn. All companies eventually have zero net losses in that year. If have a portfolio of 100USD of stocks and it goes to zero 2010 then in 2011 I will have a zero net loss.

As for focus, look at the coverage given to GS – net cost to the taxpayer negative – vs Frannie – net cost to the taxpayer more than all the other TARP recipients put together, let alone being responsible for virtually ALL the losses.

As for making the obligations good, that was always the issue with Frannie. When they got into trouble the investors – and the loudest were the chinese – called the US Treasury’s bluff.

Posted by Danny_Black | Report as abusive

The big difference between GS and Frannie is that GS is still a public company, so we can and should be worried about using taxpayer money to benefit shareholders of GS. Frannie is essentially owned by the treasury.

Posted by Nameless | Report as abusive

Actually that is not the difference. The difference is that without the government taking it over it is simply indisputable that that Frannie would have gone bankrupt, they were insolvent rather than just illiquid. That is most certainly not the case with GS and arguably even without the liquidity operation that the Fed ran, GS would have made it.

Posted by Danny_Black | Report as abusive

Excellent point by Nameless. If BAC and C are “too big to fail”, then Frannie is “too critical to do without”.

Posted by TFF | Report as abusive

TFF, Frannie simply has more friends in government, not sure exactly what purpose a private company with open access to a Treasury credit line serves.

Posted by Danny_Black | Report as abusive

Frannie were NOT insolvent, not in the sense that they had to run out of money and be unable to meet their obligations. Even with all the losses they are taking due to the government forcing them to do mortgage modifications, Fannie Mae is reporting positive net interest income every quarter and sitting on 260bn of cash and securities. They are having trouble with market valuation of their assets, which is the only reason they need to draw from the Treasury (they like to have positive net worth).

It’s really hard to tell whether GS was insolvent by any definition of the word, it’s far less transparent than FNM/FRE.

Posted by Nameless | Report as abusive

Danny_Black, their guarantee has kept the mortgage market from shutting down. Worth a few hundred billion for that alone.

Posted by TFF | Report as abusive

Nameless, again simply untrue. They were having difficulty rolling over their paper and the only reason they were able to is that they were given a stronger guarantee by the US gov. They had 332billion in short-term paper outstanding they had to roll with an average maturity of 91 days. Again in case it is hard to work out 332 > 260, even assuming that 260bn was unencumbered cash. Even assuming they COULD have rolled the debt, they burnt through nearly half that amount in two years and under even the most optimistic of views credit losses will have burnt through most of that 260bn.

As for Goldmans, it is not and was not hard to tell. They were able to roll their debt. Frannie wasn’t, just like AIG wasn’t and BSC wasn’t and LEH wasn’t.

TFF, not saying it was wrong thing to do. Am saying that for all the fuss about IBs, the cost to the taxpayer is a rounding error compared to what Frannie and the car companies are going to cost. Should also be obvious that these GSEs should never ever be allowed to be even close to what they were.

Posted by Danny_Black | Report as abusive

Even that “net interest” claim is reliant on the cheaper funding that the explicit government guarantee gives them along with a couple of accounting tricks:

“Net interest income increased driven by lower debt funding costs and the purchase from MBS
trusts of the substantial majority of the single-family loans that are four or more monthly payments delinquent,
as the cost of purchasing these delinquent loans and holding them in our portfolio is less than the cost of
advancing delinquent payments to security holders.”

From their last 10Q –  /2010/q32010.pdf

PS LEH had about 50bn in cash and marketable securities the day before they went bankrupt.

Posted by Danny_Black | Report as abusive

1. were the biggest holders of fnm -banks or China
2. didn’t fnm securities count as tier 1 capital until it didn’t–(mark-to-market fiasco)
3. what is the average mortgage yield on the portfolio-certainly above 5%
4.what is the market value of their holdings
5. is there really any loss at all if there are massive unrealized gains in the portfolio
6. why pay the govt 10%-when lending at 4.5%–reduce the dividend to 2%
7.was the goal simply to nationalize the mortgage industry and penalize banks?

Posted by dominickspez | Report as abusive

What exactly do you mean by “having difficulty rolling over their paper”? You can always roll over your paper, the question is about the interest rate your market can bear. At the time conservatorship was announced, Fannie Mae debt was trading at something like 80 bp over treasuries. Very low levels of concern, by any standard.

The cost to taxpayer is negative, seeing how all we do is lend some money at 10% to a company that brings in more money than spends every quarter (hence positive net interest income) and is at no real risk of bankruptcy (not now, anyway) so we can be 100% sure that we get back every penny, with interest.

Posted by Nameless | Report as abusive

Nameless, really? Honestly? Seriously? Are you really that stupid? BSC manage to roll their paper? Did LEH? How exactly is the cost to the taxpayer negative? You do realise that if the 90bn they are expected to draw is drawn and they pay back 71bn in dividends that is a cost to the taxpayer? If you are having difficulty understanding 90 > 71 then I suggest you learn to count. Under the the most benign conditions, they are not expected to pay that money back, despite being able to borrow at US treasury rates and having an unlimited tap to draw from the US taxpayer. 80bips is what UNSECURED CP was trading at when they were taken in conservatorship. This was ABCP and with an implicit guarantee from the US government. That “very low levels of concern”? What would constitute “high levels” to you then?

Seeing as you appear to be utterly utterly clueless, I think I will leave you to wallow in your belief that Frannie are fantastic never had any issues and are a superb investment unlike GS or MS who merely paid back their debts in a few months at a rate twice what Frannie are paying.

Posted by Danny_Black | Report as abusive