财新传媒
位置:博客 > 王韬 > 央妈二代

央妈二代

前言:介绍最近读的书《新朗博德街》,作者是Perry Mehrling,巴纳德学院的经济学教授,也在新经济学思维研究所INET任职。
 
了解央行,从Bagehot(白志浩)读起。这位《经济学人》的总编在1873年写的小册子《郎博德街》是此后一百多年央行家的行动手册。朗博德街是伦敦的银行街,当年的金融市场还比较简单,主要是商业银行和储蓄机构。流动性危机会传染,一家银行挤兑,另一家也可能会,撒手不管就是金融危机。所以作为熊孩子们的银行需要央妈,危机时刻她可以出手相救,央行因此成了最后借款人,英文里叫作Lender of Last Resort,直接翻译就是“最后可以求助的借款人”。与此同时,这种保护却不能变成娇惯纵容,鼓励投机,而是需要高高在上,在救与不救之间保留一丝暧昧,让市场摸不着猜不透。为获得短期支持,问题银行必须拿出一定担保,并且需要支付高于市场的利率,作为银行对风险管控不力的惩罚。这就是所谓的“白志浩法则”。
 
1913年美联储建立,大萧条之后有了FDIC。本以为央妈守住大门,外加存款保险护航,金融危机就不会再发生。但没想到2008年的金融海啸,传统商业银行之外的影子银行成了危机发源地,货币市场发生的挤兑原来与商业银行并无二致。时代已经大变,央行不仅要做商业银行的最后借款人,还要给整个货币市场担任借款人。因为救了贝尔斯通,美联储广受批评,这样的行为无疑鼓励了道德风险。为了给市场一个教训,也因为受限于民主程序无力进一步行动,政策制定者最后决定对雷曼兄弟撒手不管。可事实上,他们低估了货币市场崩溃之后的连锁反应,货币市场照样挤兑,缺钱卖资产,资产价格进一步下跌,局部危机迅速蔓延成一场金融风暴,甚至连装了安全阀的传统银行也未能幸免,就连历史上很少发生银行挤兑的英国银行也出现了挤兑的情况。
 
之所以发生这一切,在经济学家Perry Mehrling看来,是因为我们错误地认识了当前的金融体系和央行的角色。如果说过去,央行是一个高高在上的深不可测的严父,等待流动性紧缺的机构上门求情取款,二代的央妈发现这种主动上门的情况越来越少,市场流动性似乎十分充足,银行通过种种货币市场和回购保证了充足的流动性,信贷扩张之际,市场一片繁荣,流动性成了最廉价资源。相比之下,央妈日益门庭冷落。现在的美联储,不得不放低身段参与到市场当中。自此,央行从最后借款人,变成了最后的零售商,作者将其叫做Dealder of Last Resort,亲自到市场中买债券,卖债券,影响价格,进而干预利率。二战期间,美联储迫不得已成了联邦政府的券商,通过大举干预国债市场,稳定国债利率。联储的独立性此后虽有增加,但是因为美国巨大的国债市场,干预利率的最好方式就是在债券市场上亲自出马。另一个原因是,联邦基金利率并不能从金融市场传递给货币市场,通常情况下,回购利率都低于联邦基金利率,而联邦基金利率则低于欧洲美元的利率,但利差很小,只有几个基点。美联储只有直接介入货币市场和资产市场,才会对市场利率产生影响。
 
美联储深入地介入到市场当中,也成了一线的零售商,这是所谓央妈二代。央妈的转变,即是不得已为之,也是应该为之。不得已,是因为影子银行彻底改变了金融市场的游戏规则,货币市场基金、回购,利率不只是信贷价格,时间错配的价格,也还必须是流动性价格。必须为之,是因为宏观审慎的需要,信贷内在的不稳定性。明斯基说,信贷内在的不稳定性,抵押物值钱,信贷扩张,更容易借到钱,抵押物下跌,需要钱,却借不到钱。克服这种信贷的内在不稳定,要求央行扮演起这个角色,熨平信贷波动,早日避免泡沫,以及泡沫之后的破裂。
 
到此时就明白了,作者为啥把书名叫做《新朗博德街》。从最后借款人到最后零售商,两个词的交替对央妈角色的重新定义,带给我独特的启发。这本书值得一读,下面是kindle的部分摘录。
 

—————————————————————————————————————

  • In this regard, the decision of both the economics and finance disciplines to abstract from the monetary plumbing behind the walls, the better to advance scientific understanding on other dimensions, has had fateful consequences for our ability to sustain rational discourse in the face of a systemic plumbing failure. If we don't educate ourselves about how the system works when it is working, we will have no framework for understanding what is wrong when it fails.

  • These simple examples show how relaxation of the daily survival constraint depends on credit; some people can enjoy cash outflows greater than cash inflows only because other people are willing to enjoy cash inflows greater than cash outflows, and vice versa.

  • Payments elasticity in our decentralized payments system thus depends on interbank credit to relax the "reserve constraint" facing individual banks.

  • Typically, the repo rate is less than the federal funds rate, and the federal funds rate is less than the Eurodollar rate, but the spreads are very small, just a few basis points.

  • Knowing that the Fed will intervene to stabilize the federal funds rate, dealers rationally shift their riskreturn calculus in favor of taking larger positions in the term structure arbitrage, and such a shift can be expected to move the structure of asset prices closer to the EH theoretical ideal, but not all the way.

  • It is not that modern policymakers are unconcerned with liquidity but rather that they have convinced themselves, or rather have been convinced by economists, that matters of liquidity (the purview of an antiquated money view) can be conceptually as well as operationally separated from matters of economic stabilization (the purview of the modern economics view). The present crisis has posed a rather decisive challenge to this neat division of intellectual labor, but inevitably past habits of thought persist and it is these habits that must be confronted if we are to learn the lessons that the crisis has to teach us.

  • Time and risk are now explicitly modeled, but that is the only substantive change; abstraction from monetary plumbing remains of the essence, even more so today than in the past through the convenient analytical assumption of a so-called representative agent.

  • In a money view perspective, if the Fed fails to raise interest rates in the face of a credit-fueled asset price bubble, the bubble will feed on itself, growing ever larger and having ever greater distorting effects, until it bursts.

  • The money view emphasizes the inherent instability of a credit system driven by the private profit motive, but the problem is made worse when the Fed adopts a policy rule that denies any responsibility for preventing a bubble.

  • Abstracting from money may make our economic theory easier, but it does not make our economic policy better.

  • At its core, our monetary system is a dealer system that supports the liquidity of our securities markets, and the Fed serves as dealer-in-chief not only in wartime but also in peacetime, and especially in financial crisis time.

  • First currency swaps, then interest rate swaps, and then credit default swaps were introduced, and the eventual result was transformation of the rigid and highly regulated financial system that we had inherited from Depression-era reform.

  • A rehabilitation of the nineteenth-century money view is, I have suggested, the place to start, but it is not the place to end. The concept of liquidity that seemed appropriate for Bagehot is no longer appropriate for us. Long ago we switched over from Bagehot's emphasis on the "self-liquidating" character of certain short-term commercial debts to more appropriate emphasis on the "shiftability" of certain securities in liquid markets. But we have not yet switched over from Bagehot's conception of the central bank as "lender of last resort" to the more appropriate modern conception of it as "dealer of last resort."

  • The Fed intervened in the market for Treasury repo with the goal of stabilizing the federal funds rate at some target, and that was all.

  • Other repo rates and the Eurodollar rate got stabilized through money market arbitrage by private dealers, and the private money market then served as the source of funding liquidity for dealer operations in securities of all kinds, producing the two-way dealer markets that are the source of market liquidity.

  • That is how Martin thought the system should work, and how in fact it eventually did work, until it stopped working in August 2007.

  • From a long historical point of view, the central lesson of the crisis is that the American system requires the Fed's support as dealer of last resort, not just in the money market (as emphasized by Martin) but also in the capital market, and not just for Treasury securities (as emphasized by Martin) but also for private securities.

  • The practical intertwining of money markets and capital markets is the defining institutional feature of the American system, and that feature requires a similarly integrated backstop by the central bank.

  • Solvency risk was about the prospect of loan default, and it was handled by a buffer of bank capital, backstopped by deposit insurance at the FDIC. Liquidity risk was about the prospect of deposit withdrawals, and it was handled by a buffer of cash reserves, backstopped by the discount window at the Fed. This is the model of banking that was in the back of most of our minds as we looked at the new shadow banking system, and from this vantage point it seemed clear that the new system involved exposure to familiar solvency and liquidity risks, but those familiar exposures were handled differently. The important thing is that in the shadow banking system neither solvency risk nor liquidity risk was backstopped in any direct way by the government.

  • Lender of last resort to the traditional banking system. Like the parents of the shadow banks, the Fed professed not to be worried about the quality of the collateral, and made room for some of it at the discount window by relaxing collateral requirements and by expanding eligibility requirements.

  • Finally, in September 2008, with the collapse of Lehman Brothers and then AIG, even unsecured money market funding froze up.

  • The resulting scramble for funding drove LIBOR rates to unprecedented spreads over federal funds rates, and the Fed responded by extending lender of last resort even further, accepting a wider selection of collateral from a wider selection of counterparties.

  • From a Jimmy Stewart perspective, this final expansion of the Fed's role, dramatic though it was, seemed to be nothing more than an extension of traditional lender of last resort support. The only difference was the scale of the lending, which meant that the Fed could no longer fund its lending simply by liquidating its holding of Treasury bills.

  • The second set shows how the shadow bank parents stepped in when secured funding dried up because of concern about collateral values.

  • However, September 2008 was the moment when the Fed moved from lender of last resort to dealer of last resort, in effect taking the collapsing whole sale money market onto its own balance sheet.

  • When liquidity risk was thought to be the issue, it was the Fed's problem; when solvency risk was thought to be the issue, it became the Treasury's problem.

  • In both respects, the fact that the shadow banking system had collapsed onto the traditional banking system made it seem as though the problem was now just a traditional banking problem.

  • This financial crisis is not merely a subprime mortgage crisis or even a shadow banking crisis; it is a crisis of the entire market-based credit system that we have constructed since 1970, following Martin's 1952 report and Moulton (1918).

  • The big thing that happened in September 2008 was that the system of private dealer money market arbitrage, having been under stress for more than a year, finally froze up completely. And the big thing about the Fed's response was that it stepped in as the dealer of last resort to replace the private dealer system.

  • From the very beginning, the shadow banking system was completely dependent on a well-functioning dealer system in two senses.

  • Using this facility, any shadow bank parent that found itself holding an MBS that it could not repo, could swap that MBS for a Treasury bond that it could repo. (Initially, the facility was limited to MBSs rated AAA.)

  • In fact, however, by lending ninety cents on the dollar on a nonrecourse basis at a rate of 100 basis points over LIBOR, the Fed was doing essentially what Lehman and AIG used to do, but with less leverage and charging a higher price. (The credit risk involved in such lending was covered by funds allocated from the Treasury's Troubled Asset Relief Program under section 102, "Insurance of Troubled Assets.")

  • Operating as dealer of last resort, the Fed found itself inventing a new version of the Bagehot principle to guide its operations: insure freely but at a high premium. As dealer of last resort, what the Fed was insuring, it is important to emphasize, was not the payments that the debtor had promised to make but rather the market value of the promise itself; that is the difference between dealer of last resort and credit insurer of last resort.

  • As dealer of last resort, what the Fed was insuring, it is important to emphasize, was not the payments that the debtor had promised to make but rather the market value of the promise itself; that is the difference between dealer of last resort and credit insurer of last resort.

  • As in the original Bagehot principle, the idea is for the Fed to charge a price that provides incentive for the private market to undercut the Fed once it recovers.

  • These are bold and innovative experiments, but the basic pattern comes through clearly. The Fed now recognizes that, for our market-based credit system, it must remake itself as dealer of last resort.

  • More fundamentally, we can look forward to a remake of the framework for monetary policy, going beyond the precrisis fixation on tracking the "natural" rate of interest, and taking account for the first time of the key connection to asset prices that runs from funding liquidity to market liquidity.

  • To say that the essence of liquidity is shiftability is not to say that liquidity is or should be a free good, and it is not to say that we can safely abstract from liquidity when we consider questions of monetary policy and financial regulation. This is the central lesson of the crisis.

  • What are the implications of the Fed's new role as "dealer of last resort" for normal times? That is the question that we must confront looking forward, starting from the realization that our market-based credit system relies critically on two-way dealer markets that link funding liquidity in the money market with market liquidity in the capital market.

  • That is the question that we must confront looking forward, starting from the realization that our market-based credit system relies critically on two-way dealer markets that link funding liquidity in the money market with market liquidity in the capital market.

  • A key lesson of the crisis is that funding liquidity is not enough, since in a crisis funding liquidity does not get translated into market liquidity, no matter how hard the Fed works to push funds out the door.

  • The job of the Fed is not to eliminate the risk that dealers face but rather to put bounds on it, to establish an arena within which private calculation of expected profit and risk makes sense. For this purpose, it is helpful to think of the dealer of last resort function as a kind of tail risk insurance. Having set the bounds that establish the possibility of rational risk calculation, the Fed can then turn its attention to its more traditional function, setting the money rate of interest.

  • The classic money view urged central bankers to attend to the balance of discipline and elasticity in the money market, in order to manage the inherent instability of credit. The classic money view urged central bankers to attend to the balance of discipline and elasticity in the money market, in order to manage the inherent instability of credit. Our modern world is not Bagehot's world, by a long shot, but at the highest level of abstraction the classic money view holds as true in our world as in his.

推荐 2