The incredible shrinking TARP bailout

I seem to be the only blogger talking about this, which makes me think either I am ahead of the curve, or more likely making a bonehead error.  But as of yet no commenter has yet found the bonehead error I am making.

Last time I wrote on this subject the eventual cost to the government from bailing out the big banks was estimated at a negative $7 billion–in other words a profit to Uncle Sam of $7 billion.  There was an expected loss on the AIG bailout of $36 billion, and I acknowledged that could be viewed as a backdoor bailout of the big banks.  Indeed the entire TARP is a giant favor to the banking industry even if every dollar is repaid.  And now it appears it all will be repaid, even the AIG bailout.  (I am excluding the loans to automakers, which I regard as a separate issue.)

When the crisis first broke we were told the main problem was the big banks, and the underlying regulatory problem was “Too Big to Fail,” which encouraged the big banks to take excessive risks.  I now think my original view was wrong.  It appears that at the end of the day the biggest banking fiasco in the history of the universe will not result in any long run net taxpayer transfer to big banks.  And yet the owners and managers of those banks incurred mind-boggling losses.  So how plausible is it that TBTF was the primary cause of excessive risk taking in 2004-07?  If even a crisis this big didn’t result in  the long-run transfer of one cent of taxpayer money to big banks, does it seem likely that expectations of those transfers were a powerful motivating factor in the the MBSs they bought? Were they expecting an even bigger banking fiasco?  I suppose it’s possible, but I just don’t see it.

In contrast, massive quantities of taxpayer funds will be transferred to depositors at smaller banks, who (in collusion with the banks themselves) gambled recklessly by lending taxpayer-insured funds out to risky construction projects.  There I really do see a moral hazard problem, indeed what happened was essentially a repeat of the 1980s S&L crisis.  Once might be a fluke; twice is a systemic problem.  I’m increasingly likely to view the big bank crash as a fluke and the smaller bank crash as a chronic policy problem.  Indeed didn’t the same dichotomy occur during the Great Depression, with most failures being smaller banks?

So here are the eventual taxpayer losses we are looking at:

Fannie and Freddie  — $165 billion and rising

FDIC — Over $100 billion

FHA — Who knows, even today they’re still encouraging new sub-prime loans.

AIG — $0

The big banks — negative $7 billion

Would it be fair to say that the initial reporting of the crash of 2008 was a bit misleading?  The reporting that led most people to form indelible opinions that they will probably never re-visit or re-evaluate?

Some will argue that the Fed policy of buying MBSs indirectly helped the big banks.  Maybe so, but if we are talking about indirect effects from government programs, then what about the indirect effects of the Fed letting NGDP fall 8% below trend in 2008-09?  That hurt banks far more than any Fed MBS purchases helped them.

So tell me, why is my hypothesis wrong?

Tags: , , , ,


48 Responses to “The incredible shrinking TARP bailout”

  1. Gravatar of Morgan Warstler Morgan Warstler
    14. September 2010 at 14:14


    1. The MASSIVE unrealized gains sustained by the winners who got out early and were not able to acquire major parts of the banks at deep discounted prices AND DESERVE to be far richer.

    2. The loss of economic growth as the banks sat on reserves waiting patiently so they can acquire the assets of the failing banks through FDIC “loss sharing” agreements.

    3. The growth from the same effects of paying off interest on reserves.

    4. The major losses to Pension Funds, etc. that cannot get a decent return on their money.

    That’s whats wrong with your hypothesis…

  2. Gravatar of Morgan Warstler Morgan Warstler
    14. September 2010 at 14:17

    Scott, I think you generally fail to believe / understand – there would still be banks, the question is who should be owning them. That you get this answer so wrong and shrug speaks volumes.

  3. Gravatar of 123 – TheMoneyDemand Blog 123 - TheMoneyDemand Blog
    14. September 2010 at 14:24

    Andy Harless has a grain of truth here:

  4. Gravatar of scott sumner scott sumner
    14. September 2010 at 14:29

    Morgan, The direct subsidy from the IOR is small. The indirect effect of the program probably hurts banks, as it reduces AD and that depresses asset values and increases defaults.

    123, Bank profits plummeted when we went into recession, and the estimated losses to banks got much smaller after March 2009, when the asset markets turned around. So I’d say the banks would prefer prosperity, if they could have their way.

  5. Gravatar of Liberal Roman Liberal Roman
    14. September 2010 at 14:37

    “4. The major losses to Pension Funds, etc. that cannot get a decent return on their money.”

    As usual I disagree with everything you wrote, but even you have to realize how ridiculous your claim #4 is.

    These pensions funds would be dead flat broke as all of their assets would crash in nominal value. I am not sure Scott is defending TARP here, but he has agreed that Federal Reserve QE did pump up asset prices which had to be good for pension funds.

    Unless you are again talking about the small nominal interest rates that these Pension Funds now get in the Treasury Market. For that all I have to say is that you are again being fooled by the Money Illusion AND rates would be low anyway without any Federal Reserve actions because it’s not just the Fed that has bought into Treasuries, it’s all the financial players who were diving in at the height of the crisis.

  6. Gravatar of Owen Owen
    14. September 2010 at 14:54

    Three things strike me about the dichotomy between the realized losses at small and large banks post-TARP.

    1) Walter Bagehot famously advised that during a liquidity crisis Central Banks, “lend freely but at penal rates of interest.” Since we lent freely at less than penal rates we shouldn’t be surprised that this worked well for banks.

    2) Immediately after the crisis many commentators opined that securitization of real estate loans caused underwriting standards to slip. If the small banks are losing their shirts on loans (more likely to be held on their books that loans from large banks) to developers or commercial real estate owners that undermines this very widely circulated opinion. To the extent that “reform” relied on hasty conclusions like this, it may seem sensible now but fail miserably down the line.

    3) Maybe the big bank boosters are right and diversification across asset classes and geographic regions really did make them more stable. They weren’t resistant to liquidity shocks, but perhaps the portfolios they held really were more resilient to economic shocks (8% below trend NGDP). This doesn’ mean that TBTF doesn’t create moral hazard, but it does imply that an abundance of small regional banks is an equally vulnerable financial architecture. To the extent that we’re now trying to attract new investors to the struggling banks maybe our policies should favor consolidating batches of struggling/ insolvent regional banks rather than just looking for new money to take over their existing portfolios/ assets/ liabilities.


  7. Gravatar of MikeSandifer MikeSandifer
    14. September 2010 at 15:03

    Perhaps it’s due to ignorance on my part, but I see the S&L crisis as a very different animal. For one thing, didn’t it primarily involve S&Ls, which were in trouble going back to at least the 70s due to dysfunctional regulatory structures specific to them?

    Since the FDIC was founded, how long did we go without a banking crisis in this country? Certainly there were no existential ones.

    But, at least the deregulation story is more temporally consistent. Some deregulation in the earlier 80s which made the S&L problems worse helped move up the date and increase the severity of what otherwise would be less extreme consequences. Additional deregulation preceded the LTCM threat, and the recent’08 crisis.

    Also, someone like Roubini or Krugman might point to imprudent deregulation leading to an increased frequency of financial crises internationally.

    The truth is likely complex and certainly beyond my knowledge and other resources to investigate properly and I’m not saying all deregulation is bad. But, deregulation is often done poorly, as is regulation.

  8. Gravatar of MikeSandifer MikeSandifer
    14. September 2010 at 15:06

    That is, we went from the 30s to the late 80s without a crisis.

  9. Gravatar of Carl Lumma Carl Lumma
    14. September 2010 at 15:12

    Scott, where are you getting the $100B number for the FDIC?

  10. Gravatar of Ted Ted
    14. September 2010 at 15:12

    Scott, I never thought the “too big to fail” story made any sense – at least if you believe in reasonably efficient markets (you’ll see in a moment). Also, it always seemed really risky to rely on some supposedly implicit guarantee.

    How exactly would we detect whether “too big to fail” was a contributor? Some people like to look at borrowing costs of large banks versus small banks, but this could merely reflect the relative credit risk of big bank vs. small banks portfolios rather than expectations of a future bailout. A natural place to look is instead the credit default swap market on bank debt. If expectations of a future bailout were responsible for the excessive risk taking, investors should rationally internalize this and this expectation should be incorporated in CDS prices. If rational investors were anticipating bailouts galore we shouldn’t have expected the CDS price to move too substantially when the markets started to get in trouble since you shouldn’t be afraid of losing your money if the government is going to bail out the institutions. Instead we saw behavior totally at odds with this. We saw the prices skyrocket – behavior we would not expect if creditors expected bailouts. We saw Citigroups CDS price rise from around 15 in late 2007 to almost 250 bp in early 2008 and and eventually reaching above 450 in late 2008. We saw AIG CDS price rise from the low 30s to above 250 bp in early 2008 and then above 800 bp in late 2008. What about Bank of America? We saw that go around 10 bp up to above 150 in early 2008, eventually rising to above 200 bp in late 2008. What about Lehman Brothers? It went from a trivial amount in late 2007 to above 500 bp in early 2008.

    To me such substantial rises seem to reflect that the market expected these banks to default – to fail – not an established belief that we were going to see a bailout. Now perhaps you can say that creditors thought this before and then something led them to revise their expectations, but I see no event that would suggest that. If anything if these bailout expectations existed the handling of Bear Sterns, a relatively small institution, should have given greater confidence in the market that their expectations were correct.

    I agree that deposit insurance probably did create moral hazard though. I’ve seen some interesting theoretical proposals that can prevent bank runs, but I’d have to look into more.

  11. Gravatar of Dave Dave
    14. September 2010 at 15:12


    I think the major question here is one of solvency. The overriding fear which TBTF removed (implicitly) was not loss of value, but rather the question of continued existence.

    As bad as big losses may be for banks, there is probably a certain awareness that there’s always next year to make your money back. Bankruptcy (or being acquired for pennies on the dollar and folded into the competition) are the only things that permanently remove the chance to recoup your losses in the future.

    Implicitly, I think TBTF provided security not against ugly losses, but against plain old insolvency.

    Possibly an unsophisticated viewpoint, I’m interested to hear your thoughts…

  12. Gravatar of Ted Ted
    14. September 2010 at 15:16

    As a side note, our regulators also need to be watching the CDS prices. They can easily tell you what institutions are in trouble long before you are staring into the abyss.

  13. Gravatar of TravisA TravisA
    14. September 2010 at 15:34

    A possible response would be that the most heavily regulated entities (the big banks) did the best while the more lightly regulated entities did the worst.

    Massive losses at Fannie and Freddie were inevitable since their entire loan base was RE, even though it seems they had better underwriting standards than the big banks.

  14. Gravatar of Owen Owen
    14. September 2010 at 16:45

    Mr. Sandifer,

    the S& L crisis of the late 80s was caused not by problems going back to the 70s but by the way the industry responded to the S& L crisis of the late seventies (now less remembered, but at the time potentially huge). Essentially S &Ls were only allowed to make loans, not hold equities or other assets. so after the inflation shocks the entire industry was insolvent. Rather than explicitly bail them out, the govt. changed the rules to let them hold real estate directly.the govt. intended for them to use this as an inflation hedge, but to still primarily be small lenders. Instead they piled into real estate because returns were better and carrying costs lower than portfolios of small loans. Then came the real-estate crash and the industry was insolvent again.

    For those of you counting bank crisis that’s three in thirty years- just in the US and excluding LTCM.

  15. Gravatar of Morgan Warstler Morgan Warstler
    14. September 2010 at 16:51

    Scott, I’m glad you came to realize the size of IOR, but the real weight there is in my #1 and the #2… the smaller regional banks are all sitting on assets trying to look pretty hoping the Fed comes calling on Fridays.

    More importantly, we’re talking about likely $1T in hard assets that without TARP would be sitting on other guys, smarter guys books… that you don’t want to own and eat this isn’t my fault, eventually you are tacitly supporting Fed sponsored theft.

    The losses are giant, any attempt to forget this makes you suspect.

  16. Gravatar of Lorenzo from Oz Lorenzo from Oz
    14. September 2010 at 16:59

    Speaking as an citizen of Downunder, financial de-regulation has been a great success here, but it was very specific type of de-regulation where government stopped doing much to structure or control financial transactions per se (apart from the normal rules about fraud, etc) and concentrated on prudential regulation. This seems to be what Canada did too, from my (extremely limited) knowledge.

    So I find arguments that put things in terms of “regulation/de-regulation” unhelpful. The devil is all in the details. There is a need to separate regulation of financial products and transactions in terms of price, quantity etc from regulation of backing for institutions and activities. I realise there is some fuzziness in such separation, but that does not, on Australian experience, invalidate the distinction.

  17. Gravatar of q q
    14. September 2010 at 18:25

    i don’t understand your math here. say the net loss to the taxpayer for TARP is negligible; this is and has always been plausible imho. round it off to $0. within that $0 there are lots of plus signs and minus signs. to take a couple examples, the government made a lot of money on its goldman sachs notes: by doing this it taxed goldman sachs equity holders (tarp coupon came out of their dividend, stock dilution through warrants). this tax paid for the losses at other banks. so, i think your statement that the taxpayer didn’t pay for this in the end isn’t at all true, unless you for some reason don’t think of goldman sachs stockholders as taxpayers.

    btw, the reason the banks made it out of 2008/9 was that they made huge profits on the steep yield curve; this provided them with huge interest margins. the fed had something to do with this but also general risk aversion did too.

    not to beat you up re: the EMH as you are much more of an expert than i am, but you really shouldn’t say that fannie and freddie’s expected losses “are rising”. that’s a meaningless statement. you can say they have risen over the past X amount of time, but as they are an estimate of losses yet to come, “rising” is a prediction and unknowable in our filtration.

    as far as fannie and freddie go, the owe the government around $150B but are sitting on $100B in cash (on which they are allowed no interest) and have been cash flow positive not counting the 10% interest they pay the government for more than the past year. they have had strong revenues because of a steep yield curve and deep credit losses because of the housing price collapse and high unemployment. if the steep yield curve persists and the housing market does not collapse again they will work through the credit problems in their book and be able to pay back the government. this should especially be possible if you, as you probably do, see them as governmental entities which would mean that you’d account for their funding cost at probably 2% rather than 10%.

    fannie published a very good summary of the kind of loans they made leading up to the crisis as part of their last quarterly filing. it’s a quick read and you might find it informative; you’ll be able to see exactly how much and of what kind of subprime loans they supported. i’ll find a link tomorrow and post it for you. to my knowlege freddie mac has not published anything so accessible but their portfolio is similar.

  18. Gravatar of Joe Calhoun Joe Calhoun
    14. September 2010 at 19:10

    The deal being discussed by AIG will have the Treasury converting preferred stock to common. I don’t know the exact breakeven price but it is somewhere north of here so I don’t think you can call AIG a full payback or at least not yet.

    As for TBTF, I’m not much for so called systemic risk. Frankly, we don’t know if large banks failing really does result in systemic risk since the theory hasn’t been tested. I don’t find the 21st century domino theory of finance any more convincing than the 1960s geopolitical version. Having said that, it does seem likely that at a minimum, without TARP, some bank bondholders would now be holding equity. And those far down the capital structure would be holding nothing but a worthless certificate. The rental of some taxpayer capital for a while must have cost someone something since those bondholder losses were avoided.

    As you’ve pointed out many times, current interest rates are a reflection of past tight monetary policy. The steep yield curve then is a result of Fed policy, and is a transfer from savers (taxpayers) to the banks’ bottom line. And obviously some of the banks were only able to take advantage of the steep yield curve because of the capital provided by TARP. So even if TARP is fully repaid the shareholders and bondholders of the banks have enjoyed the use of and profits from the capital provided. Since I don’t believe in free lunches, I have to believe a transfer from taxpayers has taken place.

    One last transfer results from the change in accounting rules that allows banks to carry assets at prices that amount to a wink and a nod from regulators. Allowing banks to mark assets to fantasy frees up capital for other, more profitable uses.

    I suppose if you believe in the domino theory, the price was worth it to avoid the fallout from mass bank failures. But it most assuredly wasn’t free.

  19. Gravatar of Larry Larry
    14. September 2010 at 20:48


    Scott’s hypothesis is about whether that taxpayers got screwed, not about whether TARP changed the list of winners and losers among the playas. It’s worth noting that every government action reorders that list to some extent. See the auto bailout-shafting that GM and Chrysler creditors got as the unions jumped to the head of the line. If savvy investors (Paulson) made less money and less-savvy investors (pension funds, me) made more, that’s not a sign per se that a bad thing was done.

    I do count the AIG bailout as mostly a bank bailout, because of where the money went. But if the AIG number is down from what 170B to 36B, I’d count that as “not so bad in the end”. It’s probably small enough that we could hit up the big banks for it…It also shows the extent to which this was a panic (liquidity) rather than a crash (solvency), although we may still be kidding ourselves given the number of zombie mortgages still haunting the landscape.

    Is it ironic that the securitizing (big) banks ended up in worse shape than the paper-holding (small, construction lending) banks?

  20. Gravatar of Joe Joe
    14. September 2010 at 21:16

    Professor Sumner,

    Wait… just three months ago you were throwing praise on Russ Roberts’ paper on too big to fail,

    Are you now completely repudiating everything in that post? Are you now denying the moral hazard of FDIC?

    Also, you might be interested in Arnold Kling paper on the crisis,

    I’d love to hear what you think.

  21. Gravatar of MikeSandifer MikeSandifer
    14. September 2010 at 22:19


    Some of those details ring a bell. It’s been a while since I read about it, but it seems you know more about it than I ever did. I appreciate the clarification.

    My general point remains, however. Why would the FDIC and GSEs necessarily provide too much moral hazard, when each existed for several decades before any financial crisis?

  22. Gravatar of Nick Rowe Nick Rowe
    14. September 2010 at 22:46

    It’s good you write about this stuff Scott, especially if others ignore it.

    The question we were asking 2 years ago: “Is this a solvency or liquidity crisis?” Everyone was genuinely worried there would be missive govt losses. Ex post, the answer then is “liquidity”.

    And that answer is even more surprising when you consider that the Fed has failed to reflate the economy, so real GDP and the price level are lower than they should be, and those assets are worth less than they should be, if the Fed had done its job properly.

    This really makes me wonder if, historically, this is (nearly) always the case?

    I remember asking a friend once, who worked at the IMF, whether the IMF made a profit on its bailouts. He said it did. OK. Case for shutting the IMF is dismissed.

  23. Gravatar of Ashwin Ashwin
    14. September 2010 at 23:12

    Scott – you’re ignoring all the backstop “liquidity” programs instituted by the Fed during the crisis, most of which are still in place and have not been unwound. And you’re ignoring the benefit to the banks of the virtual socialisation of the housing market via Fannie/Freddie.

    For example, the biggest bailout in the JPM/Bear Stearns affair was the Maiden Lane vehicle which took on all of Bear’s lousy assets. And Citi got almost $300 bn of its assets backstopped by the Fed.

  24. Gravatar of Charles R. Williams Charles R. Williams
    15. September 2010 at 02:19

    Suppose I co-sign for my alcoholic brother-in-law who has bad credit and lost his job. I lend him more money and let him live in my house while he looks for work. Suppose after 3 years he pays me back with 3% interest – about what I’ve been earning on my T-bond portfolio. In a couple of years he says he’ll pay off the loan I co-signed.

    All is well. This didn’t cost me anything. It doesn’t even count as charity. Why was I worried? I should just put my brother-in-law on my credit card and relax.

  25. Gravatar of 123 – TheMoneyDemand Blog 123 - TheMoneyDemand Blog
    15. September 2010 at 02:22

    Here is Buiter in November 2007:
    “The Bank of England’s failed as a provider of liquidity. It failed in two distinct tasks; as a lender of last resort (LoLR) and as a market maker of last resort (MMLR). It is key to distinguish between these two functions. (see my earlier contributions on the subject, many joint writings with Anne Sibert at (1), (2), (3), (4), (5).) The lender of last resort provides liquidity to individual banks and other individual financial institutions in distress. Following Bagehot it does so on demand, at a penalty rate and against collateral that would be good in normal times, when markets are orderly, but that may have become illiquid in abnormal times because of disorderly markets. The market maker of last resort provides liquidity to markets for financial instruments that are normally liquid but have become disorderly and illiquid. It does so either by purchasing the potentially illiquid assets outright or by accepting them as collateral in ‘sale and repurchase agreements’ or repos. Since it corrects a market failure – illiquidity of a market is a public bad – there is no need to impose a penalty rate on repos used to mitigate such illiquidity. No adverse incentives (‘moral hazard’) are created as long as the securities that are offered as collateral are priced fairly and subject to appropriate liquidity ‘haircuts’ or discounts on these fair valuations. I discuss how such fair valuations can be established when the securities offered as collateral do not have a relevant market reference price or benchmark because they are or have become illiquid. The Bank of England mismanaged both the lender of last resort function and the market maker of last resort function.”

    The Fed has also failed in both of these two functions. It has underprovided liquidity to illiquid markets – this has caused the very steep NGDP expectation crash in September-October 2008, and was very harmful for large financial institutions. It together with the Treasury has provided liquidity to individual banks without proper penalty rate, this was a huge benefit for some large financial institutions.
    And there is Andy’s point that artificially steep yield curve (that would disappear if there was no output gap) is a huge subsidy to banks.
    So while you are right that MMLR problems have created the huge losses for banks, it is wrong to completely dismiss the moral hazard issue.

  26. Gravatar of Steve Steve
    15. September 2010 at 03:37


    the TARP was designed to help banks that are most likely to survive, but have trouble with, well, “troubled assets”.

    FDIC is for other banks that are less likely to survive. It is not surprising that FDIC loses money, whereas TARP does not. Thus, I am not sure the divide big/small banks is correct, rather relatively healthy/unhealthy.

    But you are right, the problem is probably moral hazard of implicit or explicit insurance (of which TARP is a part, even though this time we did get lucky). However, I don’t see the “mind-boggling losses” to owners and managers. After all, the rescue worked and now they are repaying the loans and are back to earn money. Only those banks that failed inflicted losses on their owners (managers, we don’t know), but also on taxpayers via FDIC.

  27. Gravatar of Steve Steve
    15. September 2010 at 03:50

    As an added question: isn’t the rescue of Fannie and Freddie a backdoor bailout of the big banks?


  28. Gravatar of Robert Waldmann Robert Waldmann
    15. September 2010 at 04:22

    YOu may be ahead of the curve, but you are well behind me. I predicted the news which surprises you on July 1.

    “Now there was also the bailout of Fannie Freddie and AIG so the over all rescue might come out as costing more than zero (I very much doubt it but it is possible).”

    I admit that my prediction has not been confirmed (yet) as I add in the GSEs. I think you can’t doubt that I got the sign of the error in the official forecasts right. It isn’t as if there has been surprisingly good general economic news.

  29. Gravatar of Morgan Warstler Morgan Warstler
    15. September 2010 at 06:52

    “Scott’s hypothesis is about whether that taxpayers got screwed, not about whether TARP changed the list of winners and losers among the playas.”

    This is EXACTLY the same thing. It is un-interesting to pretend otherwise. Someone who pretends otherwise, has an agenda not being fully explored, explained.

    As I’ve said before, the REALITY is a Fed that gets this… thus the effort to hew towards 1% inflation. The commitment of low inflation is a promise made to all the “playas” that the government will not get an unfair house cut, and that everyone can bank on the chips being worth the same amount.

    In capitalism, even in China, it is about making sure the losers lose, and the winners win, or the whole system = FAIL.

    This is a profit driven enterprise, not a socially driven one.

  30. Gravatar of Owen Owen
    15. September 2010 at 07:30

    Mr. Sandifer,

    I don’t believe that FDIC insurance does create much moral hazard because I don’t think most people (myself included) could evaluate a bank’s stability from publicly disclosed (mis)information. Lots of countries have deposit insurance and many of them have much more stable banking sectors than the US.

    I’m more skeptical about the GSEs. They were started in the 60s and we’ve had two real-estate crashes and two S&L crises since then. GSEs have certainly not made our real estate market stable or affordable and they haven’t been free.

    Also yesterday I said we’d had three banking crisis in thirty years. I should have pointed out that Nixon was forced to break the Bretton Woods agreements in the early seventies to stabilize the financial system which might otherwise have been our fourth bank crisis. And in the eighties a rather desperate Fed extricated Solomon Brothers from a crisis to prevent it from spreading to other banks. Throw in LTCM and we seem to average about 2 crisis or near crisis a decade. But don’t worry, now that we’ve passed a financial reform bill things will really be different from now on.


  31. Gravatar of thruth thruth
    15. September 2010 at 08:14

    Scott/Robert Waldman: The way official TARP estimates were computed for the budget was on a fair value basis. That is, the recorded cost of the preferred stock purchases (at time of purchase) was the difference between the price paid and estimated market value of the securities purchased. Because those assets earn a risk premium, one should expect that on average the government will “earn” the risk premium from holding them. Ex ante there was a subsidy. Ex post the government earned the risk premium (give or take) and the subsidy was re-estimated down. That’s the source of the so-called shrinking TARP bailout. (One area where the fair value estimates may have been lacking was in anticipating early repayment of the TARP injections.)

  32. Gravatar of thruth thruth
    15. September 2010 at 08:28

    Scott, you said

    “When the crisis first broke we were told the main problem was the big banks, and the underlying regulatory problem was “Too Big to Fail,” which encouraged the big banks to take excessive risks. I now think my original view was wrong. It appears that at the end of the day the biggest banking fiasco in the history of the universe will not result in any long run net taxpayer transfer to big banks. And yet the owners and managers of those banks incurred mind-boggling losses. So how plausible is it that TBTF was the primary cause of excessive risk taking in 2004-07? If even a crisis this big didn’t result in the long-run transfer of one cent of taxpayer money to big banks, does it seem likely that expectations of those transfers were a powerful motivating factor in the the MBSs they bought? Were they expecting an even bigger banking fiasco? I suppose it’s possible, but I just don’t see it. ”

    Protecting creditors is enough to encourage incremental risk taking, especially for highly leveraged organizations. Prior to the crisis, a lot of this extra risk showed up as off balance sheet activity of the TBTF banks and insurers (contingent lines of credit or backstops to SIVs, Hedge funds etc that allowed those entities to take on the leverage that they did)

  33. Gravatar of JKH JKH
    15. September 2010 at 10:41

    You’re more right than anybody else I’ve read on the subject.

    The big bank crisis was fundamentally a crisis in marked to market accounting.

    The big question is whether the endemic accounting lie was on the part of banks who didn’t want to put a fictitious value on non-marketable securities in the middle of a financial crisis, or on the part of an accounting profession that wanted to force fictitious values through bank capital positions regardless of the non-marketable nature of the asset or the surrounding financial environment.

    Either way, the effect of TARP was to stabilize market perceptions of bank capital, at least on a relative basis, and to allow the banks time to start to earn their way back from the loss effects of highly deviant “marked to market” positions and to allow the market time to start to reverse volatility discounts from all types of assets.

    Relative stabilization of market perceptions, and time, was all that was required to do this. It is not surprising that the cost relating to the big banks ended up being minimal.

    Of course, the blogosphere is still bristling with a multitude of geniuses who assure us that the US banking system is fundamentally insolvent.

  34. Gravatar of Keith Keith
    15. September 2010 at 11:04

    Simple accounting cannot tell us whether TARP had net costs to the taxpayer or not because we do not know the alternative, nor can we.

    TARP was financed by borrowing the savings of Americans and others. This savings would have flowed elsewhere but for TARP. Are the American Taxpayers better off by having this savings flow to the banks through TARP or would the American Taxpayer have been better off by having their savings finance other endeavors? Who knows? More appropriately, no one knows.

    Our income and savings are limited, and there are opportunity costs with all choices. I think a reasonable human being could make a good argument that the opportunity costs of TARP were greater than the benefits, but this is something we cannot truly know.

  35. Gravatar of Freestate Freestate
    15. September 2010 at 11:28

    Scott, the FDIC bailouts are not directly taxpayer funded. The FDIC is funded by insurance payments from the banks themselves. So you can’t count FDIC bailouts as taxpayer funded costs. All in all, the FDIC is a very good regulatory scheme.

    What I didn’t understand is the logic of this statement: “If even a crisis this big didn’t result in the long-run transfer of one cent of taxpayer money to big banks, does it seem likely that expectations of those transfers were a powerful motivating factor in the the MBSs they bought?”

    One of the biggest risks any company faces is temporary illiquidity that puts it out of business (into Chapter 11). For example, I was in Finance in a Fortune 50 company that was using overnight funding. And when that market seized up we barely survived – simply because we had a major liquidity problem due to too much short term financing risk.

    The illiquidity risk to banks is far greater than it is to other companies. So if a major bank knows that the government will step in and:
    1) Drive a steep yield curve to allow it to earn healthly profits
    2) Ease capital requirements to allow those profits to build up and increase its capital
    3) Allow it to hold under performing assets until their prices recover

    So the FDIC banks gambled – and lost. The banks were taken away and shareholders went bust and bondholders took big haircuts Other banks subsidized the losses. In the big bank situation the banks lost and even though they were insolvent and normally would go chapter 11 the government allowed them to survive. So anyone who runs a big bank figures out quickly that he has option and bonus upside without any bankruptcy downside. That is about as big as it gets in terms of moral hazard. No senior executive has incentive to “stop dancing” when things get risky.

  36. Gravatar of Accounting for Scott Sumner | economy eyes Accounting for Scott Sumner | economy eyes
    15. September 2010 at 14:29

    […] Scott Sumner thinks he is the first to note that the cost to the US government of bailing out the bi… Clearly he doesn’t read angry bear much, as I have been predicting that for months. […]

  37. Gravatar of scott sumner scott sumner
    15. September 2010 at 16:04

    Owen, Those are good points. Just to be clear, I agree that:

    1. TARP creates moral hazard problem, and I also think monetary stimulus would have been preferable to TARP. So I am not trying to defend TARP.

    2. TARP was a big favor to banks.

    But I keep coming back to the massive losses incurred by banks. Since there will be no net flow of taxpayer funds to banks, you really have to wonder whether their owners and managers were thinking “Yes, these loans are risky, and we may lose trillions, but we’ll get a little back in bailouts.” What sort of crisis would have been necessary to lead to a net flow of taxpayer funds to the big banks? Obviously even bigger than the once in a century crisis we did have. So how likely was it that the owners and managers made decisions under the assumption “well, if it’s a once in 500 year crisis, our losses will be a bit smaller due to government bailouts.” I do think TBTF is a bad policy, and probably had a slight impact on the crisis. But I no longer think it was the main problem.

    Mike Sandifer, We had a banking crisis in the early 1980s even before deregulation. Indeed the deregulation was a reaction to the banking crisis. I have not heard people suggest the 1980s deregulation be reversed. There were some changes made after the crisis of the 1980s, and I predicted they would not work, because they didn’t address the core issue of FDIC. And the re-regulation did not work.

    I don’t see any important differences between the current FDIC crisis and the earlier FSLIC crisis. It was banks and S&Ls making lots of bad construction loans in both cases. Why do you think they were different? And did the new regulation bill address this problem? Not that I can see.

    Carl, The $100 billion figure is the latest estimate I have seen. I have a reference in an earlier post.

    Ted, Those are good points.

    Dave, You said;

    “As bad as big losses may be for banks, there is probably a certain awareness that there’s always next year to make your money back. Bankruptcy (or being acquired for pennies on the dollar and folded into the competition) are the only things that permanently remove the chance to recoup your losses in the future.”

    There’s some truth to this, but in an efficient market any stock price decline is pretty much expected to be permanent. So if the stock price falls from $45 to $3, that not much different from bankruptcy. I’m not saying your wrong, I just wonder if it is a strong enough effect to explain such recklessness.

    TravisA, You said;

    “Massive losses at Fannie and Freddie were inevitable since their entire loan base was RE, even though it seems they had better underwriting standards than the big banks.”

    Yes, but I’m not willing to cut F&F any slack. It’s exactly because they were focused on a single industry that they needed to be especially careful, and obviously they took advantage of the implicit taxpayer guarantee.

    Owen, Good point about the S&Ls. Also recall that 1934-80 was a period of almost steadily rising inflation, so you didn’t have many periods where property prices fell sharply.

    Morgan, I am not supporting the TARP program.

    Lorenzo, Those are good points.

    q, That argument makes sense, but before I can comment I’d need to know which large banks aren’t able to repay their TARP loans. Does anyone know?

    You said;

    “btw, the reason the banks made it out of 2008/9 was that they made huge profits on the steep yield curve; this provided them with huge interest margins. the fed had something to do with this but also general risk aversion did too.”

    I completely reject this argument, as any gains from a steepening yield curve were more than offset by the harm done by the Fed’s tight money policy.

    The most recent estimates of F&F bailout costs was $165 billion, a few months ago. The bailout of the big banks is now expected to earn Uncle Sam a profit of $7 billion.

    Joe Calhoun, It’s the Fed’s job to prevent a systemic collapse of the banking system, that’s not really a “bailout” if it doesn’t cost the public anything. Say it was done though the discount window, as after 9/11 if I recall correctly. And I keep coming back to the point that the problem was caused (mostly) by tight money. So I simply don’t buy the argument that the Fed has done banks any huge favors. They may have tried to but they haven’t.

    I can certainly see how others would look at the situation differently, but I now doubt that TBTF was leading to much risk-taking.

    Larry, The article I linked to said the expected cost has now fallen from $36 billion to zero.

    Joe, I haven’t changed any of my policy views–I still oppose all policies that create moral hazard in banking. I still think FDIC was a major cause of the crisis. I still agree with 95% of the RR paper, and will have my students read it. But I no longer think TBTF was a major factor in the excessive risk-taking.

    Nick, Good point about the IMF. It should not be eliminated, it should be privatized! I’ll buy shares.

    Seriously, those are good points. Think about the Great Depression when the Fed let NGDP fall in half. Even in that case most big American banks survived, and all the Canadian banks survived or merged. And yet people (like Krugman) say it shows how reckless banks were in the 1920s. How many people today could repay their loans if NGDP fell in half? How many banks would survive? I think the role of the business cycle is hugely underrated, at least in the US. But I’m not claiming it explains every financial crisis in every country.

    More to come . . .

  38. Gravatar of scott sumner scott sumner
    15. September 2010 at 16:36

    Ashwin, Read my early replies, I don’t think Fed policy helped banks in net terms, although some individual actions did. F&F were a huge mistake, but I see that as a separate issue. I think they should be abolished.

    I agree that there are so many separate programs it is hard to disentangle all the effects–but did anyone expect the TARP money to be repaid in an economy that never really recovered from the recession?

    Charles, Read my post, I said the Federal government did the big banks a huge favor. That’s not the issue I’m discussing.

    123, I agree with everything you say, and in my view if you net it all out the Fed hurt banks more than they helped.

    Steve, Other than Bear Stearns, aren’t the bank failures mostly the smaller banks? I seem to recall reading that the stock prices of Citibank, Bank of America, etc, plummeted in value in 2008-09. Is that wrong? Someone had to take big losses. The IMF estimates nearly a trillion in losses to US financial institutions.

    I’ve heard arguments both ways on Fannie and Freddie. If it was a backdoor bailout, it is another reason they should be abolished. But I’ve never heard any convincing arguments. The Fed bought some MBSs, right? But has the Fed lost money on those purchases?

    Robert Waldmann, I tip my hat to you. I knew things were getting better, but until yesterday I never thought AIG might be able to pay it all back.

    thruth, I agree a risk premium should be accounted for. Are you saying that they are getting all their money back, including the risk premium, or just all the money back in nominal terms? If I read you correctly you seem to be saying that they are earning the sort of profit that a government would deserve for taking on that sort of risk. Is that right?

    thruth#2, Yes, I think the bank creditor angle is the weakest part of my argument. The owners themselves didn’t really absorb all the risk, because they were so leveraged. I guess the question is what would have happened if there was no TARP, but the Fed prevented NGDP from falling more than it did. In that case would the bank creditors have come out OK, as the conversion of debt into equity led to a short term loss, but then eventually they were made whole, as the government is now being made whole? I’m still thinking about that angle. I’m not trying to deny there was some moral hazard here, I just wonder about the severity of it, given the dimension of the crisis and lack of actual taxpayer transfers.

    JKH, That’s a very interesting perspective, and I’ll defer to your expertise on mark to market. It certainly sounds plausible to me. Of course if I’m “right” now I was “wrong” last year. :)

    Keith, I agree, and am not defending TARP nor denying it was a huge favor. But did it really create much moral hazard? That’s what I’m not sure about. I certainly created some.

    Freestate, That’s a common misconception. FDIC is basically a government organization and the fees are essentially taxes. That means they are costs that are passed on to bank customers in the form of higher loan rates or lower deposit rates. Taxpayers most definitely are picking up the cost, even if FDIC doesn’t have to be bailed out.

    I disagree with your perspective on implicit Fed subsidies, for reasons discussed above.

    Banks that did “stop dancing” like Goldman Sachs did well, so banks certainly had an incentive to be smart, unfortunately most acted very stupidly, even from the perspective of their own self-interest.

  39. Gravatar of thruth thruth
    15. September 2010 at 17:18

    “thruth, I agree a risk premium should be accounted for. Are you saying that they are getting all their money back, including the risk premium, or just all the money back in nominal terms? If I read you correctly you seem to be saying that they are earning the sort of profit that a government would deserve for taking on that sort of risk. Is that right?”

    If they had charged the then exorbitant market rates, then they would have given away no subsidy ex ante, and given the actual outcome, made out very well ex post. But they undercharged significantly (as one would expect of a govt intervention) and hence gave away a big subsidy ex ante (as measured by the initial price tag), but obviously did ok after the fact. So I say there was a subsidy involved, but the benefits from providing that subsidy may well have been worth it (to the extent we believe it stabilized the financial sector and the economy). Of course, in this sort of situation it’s a bit difficult to tell what constitutes a fair market rate, since distress pricing was rampant.

    Will get back to you on #2.

  40. Gravatar of scott sumner scott sumner
    16. September 2010 at 05:28

    thruth, Thanks that helps. And it is sort of what I assumed in the post. I agree that TARP was a big favor to banks, despite all the money being repaid.

  41. Gravatar of 123 – TheMoneyDemand Blog 123 - TheMoneyDemand Blog
    16. September 2010 at 06:11

    Scott, you said:
    “123, I agree with everything you say, and in my view if you net it all out the Fed hurt banks more than they helped.”
    For some specific banks the opposite might well be the case.

  42. Gravatar of Carl Lumma Carl Lumma
    16. September 2010 at 10:32

    Thanks Scott! If I’m reading that right, it looks like the $100B may still come from banking customers (member premiums) rather than taxpayers.

    Regarding the UAW… not that I supported that bailout, but it looks like the cost may be down to $12B ish


  43. Gravatar of ssumner ssumner
    16. September 2010 at 18:26

    123, Maybe, but it wouldn’t change my argument.

    Carl, Yes, but even if the cost was zero the UAW bailout was a bad idea.

    Taxpayers will ultimately pick up the cost of FDIC expenses, as the bank fees that support FDIC are an implicit tax that is passed on to the public by the banks.

  44. Gravatar of Scott Wimer Scott Wimer
    17. September 2010 at 06:54

    You touched on the solvency issue, but moved on before considering its implications.

    The little banks that have failed, taking big bites of the FDIC’s Deposit Insurance Fund (DIF) with them, are victims of insolvency. However, there are a couple of things to keep in mind here. The first, is that by law, the FDIC is required to take over banks BEFORE they are insolvent. It’s covered by section 38 of the Federal Deposit Insurance Act, titled “Prompt Corrective Action”. Well before a bank’s liabilities exceed its assets, the FDIC is required to step in and ensure that the bank either increases its assets or reduces its liabilities. If the bank can’t/won’t do that, but still well before the liabilities exceed the assets, the FDIC is required to take the bank into receivership and sell off the bank’s assets to cover the liabilities and make the depositors whole.

    It is Prompt Corrective Action, and only Prompt Corrective Action that allows the FDIC with its minuscule DIF to insure the deposits of the banks in the U.S.

    Unfortunately, there has been a problem here. The banks have effectively been lying about the value of their assets. Further, the big banks bullied Congress into standing on FASB’s neck to ensure that they would be allowed to lie about the value of their assets.

    “Lie,” is the wrong term, but that’s the effective result. The principal reason that an FDIC insured institution needs to report its assets and liabilities is because the U.S. taxpayers (through the FDIC and its US Treasury backing) are on the hook for the difference. Investors might care about these numbers as well, and that’s nice. But, the big deal is that when a bank makes enough bad loans, they are effectively writing checks on the backs of you, me, and our neighbors. Even better, so long as those will-eventually-go-bad loans get paid, the bank makes money.

    You can see the impact of this each Friday as the FDIC takes over yet another small to mid-sized bank and books a hit to the DIF either directly or through some staggered loss-sharing agreement (which makes the true hit to the DIF harder to quantify). So far, the over-valuation of assets in these banks has ranged from 5 to 30%! As a result, the FDIC has a negative effective balance in the DIF. And this is just from closing down little banks.

    The FDIC simply doesn’t have the DIF funds to close the big banks. To think otherwise is to think that the big banks who pushed like mad for mark-to-model (aka mark-to-fantasy) rather than mark-to-market for more of their capital base — aren’t making use of this profit enhancing mechanism to the greatest extent possible. I call mark-to-model a profit enhancing mechanism for the simple reason that it reduces the amount of capital that must be set aside for loss provisions. Loss provisions are a direct hit to the profit of a bank… therefore… each dollar NOT set aside for loss provisions becomes a dollar of profit.

    Making things better is that about 10% of the capital at the big U.S. banks is composed of CDOs issued by other big U.S. banks. They chop of the crap that they couldn’t unload on moronic MBS investors, and then pass it back and forth. That creates a market price for it, and that particular game works so long as all the participants are willing to play. It’s a bit like chicken though, because the first guy to bail wins, but at the cost of having access to that dirt cheap capital. By better, I meant something akin to “their capital position is even worse than you thought.”

    So, you haven’t gotten to pay for the big banks failures yet. That they will fail seems pretty certain — they’ve been insolvent for a handful of years now, and only allowed to continue existing through political corruption and the fact that they’re so more broke than the FDIC can make whole — without tapping a huge hunk of their $500 billion line of credit with the Treasury dept (which would incidentally require an act of Congress for them to do since we’re getting closer to the debt ceiling again).

    Gack. Too many words. back to work for me.

  45. Gravatar of scott sumner scott sumner
    18. September 2010 at 06:06

    Scott, Those are good points, but ultimate failure of the big banks wouldn’t change my argument, as long as the TARP funds had already been repaid.

  46. Gravatar of TheMoneyIllusion » The incredible shrinking TARP bailout | Adult Society TheMoneyIllusion » The incredible shrinking TARP bailout | Adult Society
    24. September 2010 at 02:47

    […] this link: TheMoneyIllusion » The incredible shrinking TARP bailout This entry was posted in Business, Finance and tagged even-if-every, even-the-aig, every-dollar, […]

  47. Gravatar of Jon Jon
    26. September 2010 at 10:52

    Scott, you’re right.

  48. Gravatar of TARPing Small Banks into (Too) Big (To Fail) Banks | The Barstool Economists TARPing Small Banks into (Too) Big (To Fail) Banks | The Barstool Economists
    30. September 2010 at 07:14

    […] happened? Scott Sumner has expressed the following view: In contrast [to the big banks], massive quantities of taxpayer funds will be transferred to […]

Leave a Reply