Last week I attended a talk and panel discussion at Brookings, in which Roger Lowenstein discussed his new book on the Fed's origins. I have much to say about that book, and I eventually plan to say some of it here. But for the moment my concern is with another book, this one concerning, not the Fed's origins, but its recent conduct. I mean Ben Bernanke's The Courage to Act.
So why bring up the Brookings event? Because, in the course of that Federal Reserve love-fest, someone made a passing reference to those crazy people who actually want to limit the Fed's emergency lending powers. Having seen the Fed save the economy from oblivion, such people, one of the panelists observed (I believe it was former Fed Vice Chairman Donald Kohn), are determined to make sure it can never save it again! At this, the audience chuckled approvingly.
Well, mostly it did. My own reaction was more like a bad attack of acid reflux. Is it really possible, I asked myself (as I struggled to keep my gorge from rising), that nobody here takes the moral hazard problem seriously? Do they really suppose that Senators Warren and Vitter and others seeking to limit the Fed's bailout capacity are doing so because they like financial meltdowns and couldn't care less if the U.S. economy went to hell in a hand-basket?
To his credit, Ben Bernanke does understand the problem of moral hazard. Moreover, he claims, in his long but very readable memoir, to have struggled with it repeatedly over the course of the financial crises. "I knew," he writes at one point, "that financial disruptions" could
send the economy into a tailspin. At the same time, I was mindful of the dangers of moral hazard — the risk that rescuing investors and financial institutions from the consequences of their bad decisions could encourage more bad decisions in the future (p. 147).
Faced with this dilemma, what's a responsible central banker to do? The classic answer — and one that Bernanke has long endorsed — is what he calls "Bagehot's dictum," after Walter Bagehot, the Victorian polymath (and opponent of central banking) who set it forth in Lombard Street. According to Bernanke's own summary of that dictum, central bankers faced with a crisis should "lend freely at a high interest rate, against good collateral" (p. 45).
Did Bernanke's Fed follow Bagehot's advice? To answer, it helps to first consider Bagehot's own elaboration of his rules:
First. That these loans should only be made at a very high rate of interest. This will operate as a heavy fine on unreasonable timidity, and will prevent the greatest number of applications by persons who do not require it. The rate should be raised early in the panic, so that the fine may be paid early; that no one may borrow out of idle precaution without paying well for it; that the Banking reserve may be protected as far as possible.
Secondly. That at this rate these advances should be made on all good banking securities, and as largely as the public ask for them. The reason is plain. The object is to stay alarm, and nothing therefore should be done to cause alarm. But the way to cause alarm is to refuse some one who has good security to offer… No advances indeed need be made by which the Bank will ultimately lose. The amount of bad business in commercial countries is an infinitesimally small fraction of the whole business… The great majority, the majority to be protected, are the 'sound' people, the people who have good security to offer. If it is known that the Bank of England is freely advancing on what in ordinary times is reckoned a good security — on what is then commonly pledged and easily convertible — the alarm of the solvent merchants and bankers will be stayed. But if securities, really good and usually convertible, are refused by the Bank, the alarm will not abate, the other loans made will fail in obtaining their end, and the panic will become worse and worse.
Plainly, Bagehot's reasons for insisting on good collateral ("good banking securities") are, first, to protect the central bank itself against losses, and, second, to make sure that only "sound" institutions benefit from the central bank's protection.
The Fed's first, extraordinary use of its last-resort lending power during the subprime crisis consisted of its decision, on March 15, 2008, to assist JPMorgan's purchase of Bear Stearns by arranging for the purchase, through Maiden Lane, a limited liability company formed for the purpose, of $30 billion worth of Bear's mortgage-related securities. Although Bernanke claims that those securities were "judged by the rating agencies to be investment-grade" (that is, rated BBB- or higher) (p. 219), their value when the Fed acquired them was anything but certain, which is why JPMorgan was determined to limit its exposure to losses on them to $1 billion — its share of the Maiden Lane purchase.
Moreover, thanks to Bloomberg's having forced the Fed to disclose the contents of all three Maiden Lane portfolios, we now know that, by April 3, 2008, when Bernanke made the same "investment grade" claim in testifying before the Senate Banking Committee, some Maiden Lane securities had already been downgraded to below investment grade. Furthermore we know that Maiden Lane I's portfolio was chock-full of toxic securities. Reacting to these disclosures, Ohio Senator Sherrod Brown, a member of the Senate Banking Committee, opined that "Either the Fed did not understand the distressed state of some of the assets that it was purchasing from banks and is only now discovering their true value, or it understood that it was buying weak assets and attempted to obscure that fact."
That Bernanke should repeat the "investment grade" claim in his book, after the true nature of the Fed's purchases has been disclosed, seems pretty surprising. So, for that matter, does his admission — offered in defense of the Fed's subsequent decision to let Lehman go under — that the Fed "had no legal authority to overpay for bad assets." If the Fed really lacked such authority, then its purchase of Bear's assets wasn't legal. If it did have permission to overpay, then the reason Bernanke gives for the Fed's having let Lehman Brothers fail — a reason he only started referring to when questioned by the Financial Crisis Inquiry Commission (FCIC), almost two years after the rescue — is phony.
If saying that the the Fed's Bear bailout was secured by "investment grade" collateral is a stretch, calling the assets in question "sound banking securities" or "commonly pledged" ones requires an impossible leap: even the Fed itself commonly accepts only AAA-rated CDOs and MBSs as collateral for its discount-window loans.
Yet perhaps the biggest problem with the Bear loan was, oddly enough, the fact that its providers did not consistently maintain that Bear was being rescued only because it had plenty of good collateral. Instead, in explaining the Bear rescue to the JEC, Bernanke argued that Bear had to be saved because its sudden failure "could have severely shaken confidence." Tim Geithner made similar claims; and Hank Paulson, in justifying the rescue to the FCIC, actually scoffed at the suggestion that Bear might have been solvent at the time. "We were told Thursday night," Paulson testified, "that Bear was going to file for bankruptcy Friday morning if we didn't act. So how does a solvent company file for bankruptcy?" How indeed. In short, far from insisting that they were rescuing Bear because, though illiquid, it was fundamentally sound, those concerned made it clear that they were rescuing it because it was Too Big (or Too Systematically Important) to Fail.*
Peruse the pages of Lombard Street all you like. You will find no equivalent to the contemporary notion that some firms are Too Big (or Systemically Important) to Fail. Nor will you discover any other exception to the rule that emergency lending ought to be confined to "sound institutions." Suppose one recklessly-managed, gigantic firm to be in danger of going under, and of ruining 1000 sound firms in the process, unless the central bank intervenes. Lombard Street offers grounds for having the central bank lend generously to the sound 1000, but none at all for having it lend to the unsound one, however gigantic it may be.
Why not lend to unsound firms, or at least to gigantic (or Systematically Important) ones? Because, if you do, every gigantic firm will come to expect similar aid, and so will be inclined to take risks it would not take otherwise. (Notice how this isn't the case if lending is confined to "sound" firms.) Of course the moral hazard problem had been present before the Bear rescue. But until then it was mainly confined to commercial banks, which had so far been the only recipients of the Fed's largesse. Although the 13(3) loophole had been present since the 1930s, the Fed hadn't dared to make much use of it even then, and made none at all for decades afterwards.
The Bear rescue convinced surviving investment banks that they'd suddenly been moved from beyond the school-ground fence to the head of the Systematically-Important class. As Michael Lewis put it not long after Bear was saved:
Investment banks now have even less pressure on them than they did before to control their risks. There's a new feeling in the Wall Street air: The big firms are now too big to fail. Already we may have seen some of the pleasant effects of this financial order: the continued survival of Lehman. What happened to Bear Stearns might well already have happened to Lehman. Any firm that uses $1 of its capital to finance $31 of risky bets is at the mercy of public opinion… Throw its viability into doubt and the people who lent them the other $30 want their money back as soon as they can get it — unless they know that, if it comes to that, the Fed will make them whole. The viability of Lehman Brothers has been thrown into serious doubt, and yet Lehman Brothers lives, a tribute to the Fed's new policy.
Lewis wrote in June 2008. And he was far from being alone in his sentiments. (See also Joe Nocera's exit interview of Sheila Bair.) Lehman filed for bankruptcy in September 2008. These facts must be kept in mind in assessing Bernanke's own assessment of the Fed's action:
Some would say in hindsight that the moral hazard created by rescuing Bear reduced the urgency of firms like Lehman to raise capital or find buyers. … But in hindsight, I remain comfortable with our intervention. … Our intervention with Bear gave the financial system and the economy a nearly six-month respite, at a relatively modest cost (pp. 224-5; my emphasis).
What Bernanke calls "a six-month respite" is what some others might be inclined to call a six-month period during which failing firms, instead of either looking for more capital wherever they could get it, including from prospective purchasers, or planning for bankruptcy, could become more deeply insolvent.
Bernanke goes on to say that Lehman did, after all, raise some capital that summer, and that it ultimately suffered runs that proved that at last some of its creditors worried that it would not be rescued (ibid.). But these facts prove no more than that the market put the probability of a Fed rescue at something less than 100 percent. In fact they don't even prove that much, for as Bernanke observes elsewhere (p. 252), Lehman, besides refusing to consider selling itself, acquired more capital only after being heavily pressured by both the Fed and the Treasury to do so; and Lehman first confronted a broad-based run on September 12, when it finally became evident that the Fed might not rescue it after all (p. 258). Moreover it's clear from the Fed's internal email communications, as disclosed by the FCIC, that the decision to not rescue Lehman was a last-minute one, and one that came as a surprise even to employees at the New York Fed, who reported in favor of a bailout.
Besides allowing an insolvent firm to go on placing risky bets with other people's money, the expectation of Fed support makes both troubled firms themselves and their prospective buyers unwilling to clinch a deal until the pot has been sweetened. Had Bear been allowed to fail, or had Bernanke and company somehow been able to persuade larger investment banks that despite the Bear bailout their still greater Systematic Importance was no guarantee of Fed support, Lehman might have felt compelled to grab one of the lifelines thrown to it by CITIC securities and the Korean Development Bank, instead of waiting for the USS Fed to toss it a thicker one. Whether any of Lehman's prospective, later purchasers were also holding out for such a deal isn't clear, although Bernanke acknowledges that at one point both Bank of America and Barclay's, having found Bear's losses to be much bigger than had previously been assumed, "were looking for the government [i.e., the Fed] to put up $40-$50 billion in new capital" (p. 263), and that he worried at the time that the firms might be "overstating the numbers as a ploy to obtain a better deal."
In the case of AIG's rescue, it's even harder to avoid seeing a moral-hazard-inspired game of chicken being played out between the lines of Bernanke's account. "Every time we heard from the company and its potential private-sector rescuers," Bernanke writes, "the amount of cash it needed [from the Fed] seemed to grow" (p. 275). When two firms finally made offers, AIG's board "rejected them as inadequate," and then made sure its representatives let Fed Board members know that "it would need Fed assistance to survive" (p. 276). A day later AIG executives "were hoping for a Federal Reserve loan collateralized by a grab bag of assets ranging from its airplane-leasing division to ski resorts" (p. 127). Would those executives have entertained such hopes if Bear hadn't been rescued, or if the Fed had been prohibited by statute from rescuing potentially insolvent firms, or ones lacking "good banking securities" in the strict sense of the term?
The $85 billion loan that the Fed ended up making to AIG was in any case even less justifiable on Bagehotian grounds than its loan to Bear had been. As Bernanke acknowledges, the collateral for the AIG loan consisted, not of any sort of securities but of "the going concern value of specific businesses," the value of AIG's marketable securities having been "not nearly sufficient to collateralize…the loan it needed" (p. 281). Even granting Bernanke's claim that such collateral met the Fed's own legal requirements — a claim that is one of many reasons for entertaining serious doubts concerning Bernanke's insistence that the Fed could not legally have rescued Lehman Brothers — it certainly couldn't be said to consist of "good bank securities." On the contrary, it was so bad that when the Fed was forced to disclose its Maiden Lane holdings, those of Maiden Lane II and III, which held AIG's troubled assets, were worth 44 and 39 cents on the dollar, respectively.
Although the Fed's defenders, Bernanke among them, are quick to note that all three Maiden Lane portfolios eventually recovered, so that the Fed (or rather taxpayers) bore no losses, the fact that they did doesn't at all suffice to square the rescues in question with Bagehot's well-considered advice. That advice simply doesn't allow central banks to place risky bets on troubled firms. Bagehot never says that it's OK for a central bank to set his advice aside provided that its gambles end up paying off.** The Fed's apologists also fail to consider that, while the Fed itself may have come out of the deals it made smelling like roses, the same cannot be said for several of the private firms that took part in them.
And what about the moral hazard consequences of the AIG bailout? Time will tell, but at very least the bailout set the dangerous precedent of having the SIFI ("Systemically Important Financial Institution") stamp applied to non-financial firms. And although Bernanke assures his readers that the bailout's "tough" terms were such as would not "reward failure or…provide other companies with an incentive to take the types of risks that had brought AIG to the brink" (p. xiii), he fails to point out that the terms, though "tough" on AIG's shareholders, let its creditors, including Goldman Sachs, go Scot-free, instead of insisting that they accept haircuts. The trouble is that, unless creditors bear some part of the risk of failure, they will chase after high non-risk-adjusted returns, even if that means depriving safer firms of credit.
The plain truth is that, despite his professed devotion to Bagehot, Ben Bernanke was never able to heed the principles laid down by that great authority on last-resort lending.*** Nor is it hard to see why. When confronted by a failing SIFI, it generally takes more courage for a central banker to refuse aid than to grant it. After all, if the SIFI survives, the central banker can claim credit, whereas if it doesn't he can at least claim to have "acted." On the other hand, if the SIFI is left to fail, the costs are obvious and immediate, whereas the benefits are largely invisible and remote. Bad as it was, the drubbing Bernanke took for bailing out Bear and AIG was nothing compared to the horsewhipping he received, even from some people whose opinions he had reason to care about, after he let Lehman fold. The usual public choice logic applies. In any event, no one knows how to calculate the net present value of present and future financial losses. And who, in the midst of a crisis, would pay attention if someone managed to do it?
And that is why it makes little sense, after all, to blame Ben Bernanke for the Fed's irresponsible bailouts. Apart from allowing Lehman Brothers to fail, he only did what just about any central banker would have done under the same circumstances. For among that tribe, the courage to act is one thing; the courage to refuse to rescue large, potentially insolvent firms is quite another. And that is why we need laws that make such rescues impossible.
*Although throughout most of his memoir Bernanke insists that the Fed's emergency lending was "guided" by Bagehot's dictum calling for "unlimited credit to fundamentally solvent financial institutions and markets " (p. 268), at one point he writes that "In a few instances, we went beyond Bagehot by using our lending authority to rescue large institutions on the brink of collapse, including Bear Stearns and AIG" Note that "including." (Note added 11/6/2015.)
**Bernanke himself appears to confuse loans paid in full with loans made to solvent institutions when he observes that "Nearly all discount window loans [are] to sound institutions with good collateral. Since its founding a century ago, the Fed has never lost a penny on a discount window loan" (p.149). Apparently he is unaware of, or has forgotten about, the House Banking Committee's study of Fed discount window lending during the late 1980s and Anna Schwartz's St. Louis Fed article on the same subject.
***The conclusions appears to hold, not just for the Fed's more notorious rescues during the crisis, but also for its lending through the Term Securities Lending Facility (TSLF) and the Primary Dealer Credit Facility (PDCF), both of which lent on toxic and systematically overvalued collateral.