Bailout bleg

I’d be very interested in seeing an estimate of how much taxpayer money will ultimately be lost in the following areas:

1.  AIG

2.  Big banks (TARP)

3.  Auto union company bailouts

4.  The GSEs

5.  Losses to FDIC (which are ultimately paid for by taxpayers, contrary to what the press often reports.)

6.  Other taxpayer losses associated with the recent bailouts.

I find it very hard to get reliable information.  Free Exchange recently linked to this AIG press release:

“With AIG repurchasing all outstanding warrants issued to the U.S. Treasury, we are turning the final page on America’s assistance to AIG,” said Robert H. Benmosche, AIG President and Chief Executive Officer. “We appreciate the opportunities this support allowed and are proud to have returned to America every cent plus a profit of $22.7 billion.”

That surprised me.  Is it true?  I’m not interested in the market value of guarantees, which are a hidden subsidy, but actual cost to taxpayers.  Where does one go to get that information?



27 Responses to “Bailout bleg”

  1. Gravatar of OneEyedMan OneEyedMan
    4. March 2013 at 07:08

    Check out Paulson’s Gift by Veronesi and Zingales (NBER 2009)

  2. Gravatar of babar babar
    4. March 2013 at 07:13

    in many of these cases nobody will know the full amount until the books are closed. (GSEs are a case in point – you can get estimates, all of which come through an ideology-tainted lens of one sort or another.)

  3. Gravatar of Reader Reader
    4. March 2013 at 07:39

    Brief rundown of my understanding for all of these.

    1 AIG – the US govt has now completely exited its loans and ownership and made an accounting profit, but there are two big caveats commonly raised. (a) the terms were too favorable and were even improved several times, representing losses vs what the government should have received, and (b) the exit price was boosted by preserving AIG’s NOL asset value, which will lower future tax revenues from AIG, and some contend the government control should have zeroed the NOLs instead (causing a lower exit price but no future tax deductions)

    2 TARP – this is largely wound up, but maybe not fully.  There were some losses of TARP funds with some smaller banks, but gains from others mean there will be a net accounting profit.

    3 Auto bailouts – these were a loss of many billions, but the consensus that letting them fail unassisted and be liquidated would have been worse for the economy is probably right.

    4 GSEs – many billions lost backstopping these, with funding from the treasury needed to keep them running.  The estimates for losses do keep improving though.

    5 FDIC – the cost of the guaranty is collected from the industry through fees.  The fees were changed to be weighted by bank assets rather than just by deposit base, putting more burden on larger banks.  FDIC did run low on cash to meet the cost of failures during the crisis but made banks prepay their fees for several years in a special collection to replenish.  I don’t think FDIC has ever taken taxpayer funds to meet its insurance fund costs.

    Search around a bit for quantification of 3 and 4.

  4. Gravatar of maynardGkeynes maynardGkeynes
    4. March 2013 at 07:39

    While an interesting question in itself, I hope that you don’t believe that taxpayer’s losses are the correct metric for gauging the true cost of the bailouts.

  5. Gravatar of J Mann J Mann
    4. March 2013 at 07:49

    Ritholz collects some of the complaints about the AIG accounting.

    1) It generally combines the profit from shares acquired by the NY Fed and shares purchased through TARP, although if you’re interested in total cost to the taxpayer, that may be appropriately. (Especially if, absent TARP, the NY Fed would have taken a loss).

    2) It ignores the cost of granting AIG a substantial tax concession. (Which might also be appropriate, if you assume that AIG wouldn’t have been able to pay the tax if there had been no bailout).

    3) It assumes no discount rate. (Which again, might be appropriate given the government’s costs of borrowing during the relevant period).

    I haven’t done the math, but it looks to me that while you can argue about the amount, and while there are all kinds of other stories about risk and moral hazard, the taxpayers seem to have come out ahead in the specific case of AIG.

  6. Gravatar of ssumner ssumner
    4. March 2013 at 08:08

    OneEyedMan, Does something written in 2009 have the ultimate cost to the taxpayer? That’s hard to believe.

    Thanks babar.

    Maynard, No, but a point many have missed is that the taxpayer costs are very important in establishing the the non-taxpayer costs.

    For instance, an additional cost is that moral hazard distorts the financial system. So bailouts encourage risk taking. But how much? My tentative hypothesis is that for big banks the answer is “only a little,” and for small banks the answer is “a lot.” But I can’t be sure because I have almost no reliable data to work with.

    J Mann, Thanks, that’s helpful.

  7. Gravatar of Errorr Errorr
    4. March 2013 at 08:10

    The problem is that the FED programs are folded into the accounting and those have all been profitable. Overall though, the GSEs are certainly losers, as were the auto bailouts in the end, TARP will come out positive although not as positive as it could because they let a bunch of the healthier banks exit early. It also includes profits from FED programs like Maiden Lane II & III and other dividends from their operations before and during TARP. Alot of those losses come from the MBSs from specific companies, although most of those seem to be profitable minus the 5B (16B estimated for the entire program by the end) for mortgage restructuring.

    The best estimate I’ve seen is from the non-partisan ProPublica at

    Note: “Altogether, accounting for both the TARP and the Fannie and Freddie bailout, $606B has gone out the door””invested, loaned, or paid out””while $360B has been returned.

    The Treasury has been earning a return on most of the money invested or loaned. So far, it has earned $101B. When those revenues are taken into account, $144B is the net still outstanding as of Mar. 1, 2013”

  8. Gravatar of ssumner ssumner
    4. March 2013 at 08:34

    Thanks Errorr, So it looks like TARP will help us to pay down the national debt.

    I still think TARP was a bad idea, but that’s interesting.

    Thanks reader, I’d quibble over a few points. The auto bailouts were a mistake–as long as people buy cars, firms will supply them. It was a union bailout. FDIC losses are taxpayer losses, I don’t understand why so many people miss that points. The so-called “fees” are simply a tax on banking, and like other taxes are mostly picked up by consumers (although even if they weren’t they’d still be a tax.

  9. Gravatar of TallDave TallDave
    4. March 2013 at 09:10

    It’s not unusual for the lender to see a profit, the same thing happened to Northern Rock in Britain. It’s the advantage of being lender of last of resort.

    The owners, otoh, generally lose about everything, and existing creditors seem to often get pennies on the dollar (especially in cases like GM where the government rewards its clients ahead of creditors, in what would be a violation of bankruptcy law if anyone else did it).

    Unfortunately, in the case the the GSEs the government was a creditor, in the sense that it guaranteed their loans, even though the charter explicitly says it does not. Bleah.

  10. Gravatar of Doug M Doug M
    4. March 2013 at 09:12

    I don’t think that realized gains and losses are all that relevant. We should be looking at what was put at risk. If the government was lucky or unlucky after the fact, doesn’t shed much light on whether this was good policy or bad policy.

  11. Gravatar of TallDave TallDave
    4. March 2013 at 09:13

    2) It ignores the cost of granting AIG a substantial tax concession.

    That happened with GM, too — typically you don’t get to carry over losses through a bankrupcty for tax purposes. There’s a $17B gift to GM, courtesy of taxpayers.

  12. Gravatar of TallDave TallDave
    4. March 2013 at 09:16

    the consensus that letting [the automakers] fail unassisted and be liquidated

    Does anyone seriously think they would have been liquidated? Courts don’t liquidate companies with billions in revenue! The reason the government stepped in was to deliver a big fat kiss to the UAW.

  13. Gravatar of Doug M Doug M
    4. March 2013 at 09:17


    Regading the auto bailout. I know what of the larger creditors to GM prior to the bailout. He and a group of investors had a large package at the ready but were front-run by the government. The private plan would not have been union friendly, and would have gutted the pension plan. But, had the government not stepped in, GM would still be in business.

  14. Gravatar of Steven Kopits Steven Kopits
    4. March 2013 at 10:15

    That the government made a profit on financial bailouts should not come as a surprise.

    It’s pure Gorton.

  15. Gravatar of Ron M Ron M
    4. March 2013 at 10:57

    “Where does one go to get that information?”

    For bailouts by the Treasury start here.

    Summaries are in the daily reports.

    Details are in the transaction reports. AIG starts on page 31 of the 3/1/13 investments pdf. Auto industry is on page 28.

    The Fed’s view of the AIG bailout is here.

    More on the Maiden Lane programs is here.

    Other Fed lender-of-last-resort programs are here.

    AIG’s view of their bailout is here.

  16. Gravatar of dbeach dbeach
    4. March 2013 at 11:12

    Yes, TallDave, GM would have been liquidated. In order for a restructuring under Chapter 11 to occur, the company would need DIP (Debtor-in-Possession) financing in order to keep operating during the restructuring. Given the state of the economy and the financial system in early 2009, no such financing was going to be forthcoming from private sources at that time, meaning the company would have been liquidated. Note that the private sector’s reluctance to provide said financing has been somewhat justified by the fact that the government has taken losses on those loans (though presumably private banks might have made loans on more favorable terms than the government received).

    I don’t have a strong view as to whether the auto bailout was good policy, and yes the UAW was a huge beneficiary of it, but I have little doubt that absent the bailout GM and Chrysler would have ceased to exist.

  17. Gravatar of OneEyedMan OneEyedMan
    4. March 2013 at 11:55

    I guess if you want to know the ex-post cost it would have to be more recent. In general, you get a rough sense of the ex-ante costs at the time of a bailout decision but not the ex-post until it is much too late. The ex-ante cost seems like the economically important quantity and was known in 2009.

  18. Gravatar of ssumner ssumner
    4. March 2013 at 12:16

    TallDave, Good point.

    Doug, You are in no position to decide whether the information is useful until you know what I plan to use it for. The interesting amount is not the funds put at risk, but rather the expected net flow of taxpayer money to bank creditors.

    I happen to think it was a bad public policy, but I need all the facts to know exactly why it was a bad public policy.

    dbeach, Regardless, there’s no doubt that with a US auto market averaging 15 million units per year most of those factories would have been bought up by other companies, and would have been producing cars.

    Ron M. Very helpful.

    Everyone. So the big banks repaid the bailouts, and the AIG profit will be considerably larger than 22 billion. The auto industry will probably not fully repay the loans. We know that FDIC will cost the taxpayers lots of money, (unless one sees the high cost of cigarettes as being due to a “fee on tobacco companies,” and not a tax.)

  19. Gravatar of Doug M Doug M
    4. March 2013 at 12:23

    I will consider myself repremanded.

  20. Gravatar of Steve Steve
    4. March 2013 at 13:34

    1. The AIG bailout was hugely profitable, but the political consensus is that it was immoral because it helped bankers abd speculators. AIG’s greatest sin was selling CDS on the rest of the financial system, effectively a suicide pact with the rest of wall street.

    2. TARP was hugely profitable, but the political consensus is that it was immoral because it helped big bankers.

    3. The auto bailouts were big money losers, but the political consensus is that it was morally good because they rescued industrial companies and their unions.

    4. The GSE bailouts were colossal spectacular money losers, but the political consensus is that it was morally good because it helped creditors like Russia and China, as well as homebuyers. Besides, the GSEs were screwed over by the big banks and should have to right to cherry pick every non-performing loan and put back those loans to the big bankers. Snark, snark.

    5. The FDIC lost money to small banks, but made money off large banks. The political consensus is that this is morally good, because community bankers are upstanding pillars of society who lend to small business, but big bankers are greedy wall streeters who sit on excess reserves.

  21. Gravatar of dtoh dtoh
    4. March 2013 at 15:57

    No surprise on AIG. AIG was claiming from day one that there would be no default on any of its assets and that the problem was not one of solvency…. only liquidity (or more specifically a mark to market problem).

  22. Gravatar of Reader Reader
    4. March 2013 at 17:26

    I’m lost on how FDIC is funded by taxes, unless the point is just to elide all distinction between fees connected to specific services and taxes. Where do park visitor fees fall?

    The cigarette tax doesn’t help clarify at all, because I think those are paid into general revenues and never adjusted to match any designated costs.

    I do understand about incidence of costs and that bank clients ultimately bear a good chunk of the FDIC fees. A taxpayer can avoid that by using cash or credit unions, of course. How is it related?

  23. Gravatar of Reader Reader
    4. March 2013 at 18:04

    OK, bit of devil’s advocate on the auto *union* bailout, which I understand but think is overplayed here.

    The main driver behind capital losing out a lot (87.5% per senior CDS auction) while unions “lost” out much less, was that labor would play a future role in any sensible plan to keep things running, while capital’s contribution was over and done. The unions still had to come in to work, and pension assets are held in trust for the employees and not on the table.

    There could absolutely have been a private bid to compete with the government’s in the bankruptcy 363 sale, but no one could pull it off. Bondholders agreed to the plan because their alternative was drawing the bankruptcy out and taking possession of factories without functioning operations or supply chain — this would have caused a lot of deadweight loss and change in global structure of the industry.

    FWIW, I think the government losses ($25B in one article) can be attributed to the pre-bankruptcy loans (under the Bush administration and in contrast to post-BK DIP financing) to keep the industry on life support waiting for the next administration. Perhaps a step too far in devil’s advocacy, but you could call those taxpayer losses a partial bailout of the bondholders, who should have been cut off earlier.

  24. Gravatar of Steve Steve
    4. March 2013 at 19:25

    “I’m lost on how FDIC is funded by taxes”

    The FDIC is a tax on bank depositors, just as the ciggie tax is a tax on smokers.

    The FDIC tax gets redistributed to loser banks and a handful of bureaucrats.

  25. Gravatar of Suvy Suvy
    5. March 2013 at 07:21

    I think that TARP was conducted poorly. I don’t think the banks should’ve been “bailed out”; either they should’ve been allowed to fail outright, be nationalized, or their debt could have been swapped for equity. If I’m injecting capital into banks, I should get what those that inject capital are entitled to: a share in equity. The price of the shares of those banks should have gone to 0 while all of those banks were wound down. I think it’s clear that having a financial sector of that magnitude is not healthy for an economy and counterproductive. The only reason the financial sector got so large is because of the implicit/explicit bailouts from 1980-2008 that even exist today. In order to have capitalism, you have to have poorly managed firms go bankrupt.

    As for the auto-bailout, I have my doubts about saving that industry. I certainly do understand the situation they were in–they ran into difficulties when the credit crunch was happening so it was impossible for them to get private sector funding. I’m a lot more sympathetic towards saving the auto companies rather than saving the banks, but I still have major problems with doing so.

    As for the GSEs, I think they should be would down. I don’t see a reason why they need to exist anymore. We don’t need to keep pushing people into houses under the assumption that everyone needs to own a house. I’d like to see a mandated law requiring a 20% mandatory down payment required on every house. I also have a major problem with subsidizing people to take on leverage(this means eliminating the mortgage interest rate deduction).

  26. Gravatar of ssumner ssumner
    5. March 2013 at 09:10

    Doug, Sorry if I was too grouchy.

    Steve, Great comment.

    Reader, Yes, and the taxpayer can avoid the cigarette tax by not smoking.

    Suvy, I’ve also proposed the 20% downpayment.

  27. Gravatar of dbeach dbeach
    5. March 2013 at 10:55

    While it’s undoubtedly true that in the event of the liquidation of GM and Chrysler most of their factories and machinery (and employees) would eventually have been purchased and put to use by other auto manufacturers, consider the situation in early 2009. US auto sales were 9 million, not 15 million. Unemployment was on its way to 9 percent. Those factories and their workers would not have been put to use in anything like the short term. Many of GM’s and Chrysler’s parts suppliers would have failed, resulting in many thousands more unemployed. (Again, much of that capital and labor would eventually be put to use by other manufacturers, but not immediately.) Probably there would have been many municipal if not state bankruptcies in the Rust Belt.

    Of course much of this damage could have been prevented or at least mitigated with appropriate monetary policy, but as you have pointed out over and over, we didn’t have appropriate monetary policy back then, so I see no reason to assume it.

Leave a Reply