本文发表于2015年6月4日的央行观察

业界期待已久的人民银行调查统计司盛松成司长与翟春老师合著的《中央银行与货币供给》终于发售了，这本书应该是盛司长继2014年4月发行的《社会融资规模理论与实践》之后的新作，也是继2014年12月人民银行货币政策司和货政二司的李斌与伍戈的《信用创造、货币供求与经济结构》之后人民银行的又一力作。《中央银行与货币供给》是对经典货币银行分析的总结。经典货币银行分析首先从货币的定义与度量出发，探讨货币的需求与供给，基础货币和货币乘数对于货币供给的影响，以及货币政策和传导机制对于基础货币、货币乘数、货币量等影响，而这也是本书的展开脉络。

在科技不发达的过去，中国的个人无法开设支票，也没有银行卡，无法通过转帐的方式进行交易，因此储蓄存款并不被看作M1的一部分，而被放在准货币中，是十分符合现实的。我们也由此看到，1999年底，流通中货币余额为1.3万亿，而单位活期存款为3.2万亿，纸币和硬币仍然是人们生活中主要的支付手段。但在今天，纸币和硬币的作用越发微弱，或许使用量最大的时候就是春节附近，而甚至这一功能目前也要被微信红包所取代。2015年3月底，流通中货币余额为6.2万亿元，但单位活期存款则达到27万亿元，而此时的54万亿个人存款必然也包含了一部分可以用来买卖商品和劳务的活期存款，例如工商银行2014年底个人存款余额为7.2万亿元，其中3万亿为活期存款。以此推算，54万亿个人存款可以约有20万亿的活期存款，那么6.2万亿的纸币与硬币和超过45万亿的活期存款相比，比重较1999年底大为降低。更不要说和近130万亿的M2货币量相比了。

上面讨论纸币的重要性，是想揭示在今天我们生活的社会中，纸币这种央行发行的负债（尽管现在纸币持有者不能持纸币找央行兑换任何东西，纸币仍然按照历史做法被放在央行的负债方）对货币总量的影响减小，这也和人们生活的体验相同，钱包里的钞票和在银行的存款相比越发微不足道。因此，传统上认为通货膨胀是开动印钞机的结果当然是早已过时了，而通过调节准备金存款余额以及法定准备金率的方式，间接调节主体为商业银行存款的货币量的增长速度，则成为中国现在数量框架的货币政策之核心，这也是盛司长的新作内容之核心，也就是基础货币、法定准备金率与货币量三者的联动机制。

《中央银行与货币供给》一书强调，准备金存款作为基础货币的一部分，是人民银行的负债，其规模完全由人民银行决定，其任何变化都反映了人民银行资产负债表的某一项目或几个项目的变化。当然，上面提到的“变化”可能出于人民银行的自主选择，也可能是人民银行由于其他原因而被迫选择，特别是近年来汇率政策导致的人民银行对商业银行的结汇，使负债侧准备金存款总额随资产侧的外汇占款不断增加。央行为了维持汇率的走势，只能将自身资产负债表当作一个“池子”，以维持合意的汇率目标，而不是自己站在一边，任由商业银行买卖来确定汇率。

除了外汇占款项目外，其他项目也会影响准备金存款余额，比如对金融机构发放贷款，此外还有购买国债，只不过央行持有的国债绝大多数为成立中投时的操作，该操作在购买的同时还包括了售汇以及财政存款的变化，过程相对复杂，不在本文展开讨论。而在总资产不变的情况下，央行也可以调整自身负债的结构，将准备金存款这一负债调整为央行票据，或通过回购使准备金存款转化为其他负债。此外，春节时发红包带来的纸币需求，财政收支通过财政存款带来的扰动，也会影响准备金存款的总量。固定收益市场人士对于上述机制都是十分清楚的。

如果说准备金余额使商业银行的资产与央行的负债联系起来，那么准备金率则将商业银行负债与央行负债联系起来。众所周知，银行在央行持有准备金存款作为资产，自身负债包括客户存款，准备金存款与客户存款的比是准备金率。央行规定一个法定准备金率，银行的准备金率要在特定日期达到法定准备金率以上。准备金存款和法定准备金存款要求之间是超额准备金存款，准备金率和法定准备金率之间是超额准备金率。在银行整体准备金存款规模不变的情况下，更高的法定准备金率会降低超额准备金率，而如果银行整体想要维持某个超额准备金率或超额准备金规模，银行整体惟有减少存款一途，传统手段是少发放和多催收贷款，这样客户用存款归还贷款的时候，银行的贷款资产和存款负债都会减少，准备金率中的分母小了，比例提高，更高的法定准备金率要求得以满足。

对单个银行而言，考虑法定准备金率的视角往往是“在给定存款基数的情况下需要多少准备金存款”，但对于央行而言，其考虑因素乃是整个银行业的存款，因而其考虑视角是“在给定存款准备金的情况下银行业如何调整自身的存款总额”。视角的不同产生两种央行看来“不乖”的情况，一是银行逼宫央行，即银行通过结汇或放贷创造了大量人民币存款，因此要求央行提供更多准备金存款或者降低法定准备金率，否则就通过银行间外汇买卖和人民币同业拆借冲击汇率和利率水平。二是央行对银行束手无策，央行试图宽松，但银行出于种种考虑惜贷导致信贷供给不足，也可能经济下行导致信贷需求不足，于是央行对准备金存款以及法定准备金率的调整无法体现在银行的存款量增长上，这当然就是目前欧美央行和银行间的关系，超额准备金已经非常多，但银行信贷扩张乏力。

在中国，准备金率十分复杂，首先，准备金率是一个比率，分子是在央行的准备金存款，分母是被人行称作“一般存款”或“各项存款”的一个交存范围。但对于不同银行而言，适用的会计科目并不同，例如在《中国人民银行关于调整交通银行和招商银行准备金存款交存范围的通知》（2013年4月2日，银发[2013]85号）中，招商银行的准备金存款交存范围有21个科目，而交通银行则有22项。不同银行的科目也都五花八门，中国银行居然有一个“8429长期不动产”的科目在交存范围里。最近的银发[2014]387号文把一些同业往来项目下的存款纳入存款准备金交存范围，存款类金融机构被要求报送全套会计科目及会计科目使用说明，人民银行也会按照上述情况调整交存存款准备金的会计科目范围。所以关于“一般存款”的信息，应该只有人民银行自己比较清楚。

即使所有需要交存准备金存款的机构科目全部一致，余额全部公开，我们也很难计算在央行处的法定准备金存款要求是多少，因为准备金率对于不同的机构并不相同，这一点在历次准备金率调整的公告中体现得很明显，定向降准当然更体现了这一点。此外2010年底人民银行引入了差别准备金动态调整机制，不但准备金率会根据银行运行状况变化，而且公式也会纳入新的参数，自身的参数数值也会调整，因此对于某一机构而言，人民银行公告的法定准备金率调整结果就只是一个参考。此外，银行的外币存款也需要有银行在央行处存放的外币准备金来对应，而一部分人民币存款准备金也是用外币缴存的，这样的复杂机制使得基于公开数据的市场分析基本上无计可施。

很多人认为，降准意味着银行有很多钱可以用于放贷，而提准意味着银行需要补交更多准备金，上述人行的公告里也是用“交存”这一词汇来表达。实践当中，这样的表达十分生动，但在理论上，人民银行是非常清楚法定准备金率变动对市场的影响机制的，这也反映在《中央银行与货币供给》以及其他央行其他人士的著作当中。银行并不能用自身存款在央行交存保证金存款，因为银行自身的存款是在银行的负债侧，而银行在央行的存款则是银行的资产，前者无法转化为后者。此外，对个别银行而言，在法定准备金存款的考核日前，余额不足的银行需要通过同业拆借、出售资产或拉存款等方式满足法定准备金要求，看上去是一个“借钱交存”的过程，但对整个银行体系而言，同业拆借、出售资产或拉存款都是不同银行间的资产转移，准备金存款从一家银行被转帐至另一家银行，但准备金存款总量由人民银行确定，所以“借钱交存”在人民银行看来只是银行之间在它这里来回转帐，银行作为整体并没有“交出”更多的准备金。

对于上面的机制有两个经典问题。一个是准备金率是否存在上限，这个问题在人行孙国峰老师的著作《第一排》的前言中被提出，有人认为这个上限是100%，即所有存款都转化为准备金。另一个问题是如果整个银行体系的准备金存款余额低于法定准备金要求，银行该如何做，有人认为可以将一些存款转化为准备金以解决这一问题。这两个答案当然都是错误的，错误的根源都是没有理解，在银行存款总额和准备金存款总额的关系中，两者不能互相转化，前者由银行整体确定，而后者由央行确定，这当然是盛松成老师在《中央银行与货币供给》一书中反复强调的。

孙老师对第一个问题的答案是没有上限，因为银行存款可以大于准备金存款，也可以小于准备金存款，这完全取决于银行整体的资产负债的结构，例如当大批定期存款转变为不计入准备金率分母的银行债券后，存款余额减少，但准备金存款总额完全不受影响，由此则准备金率可以非常大。或者存款余额不变，但央行大举买入银行贷款资产，央行的资产侧增加了对非银行部门的债权，负债侧增加了准备金存款；而银行总体则总资产不变，准备金存款增加，对非银行部门的债权减少，准备金率得以提升，当然这里的上限是银行总资产与存款之比，但这个上限必然在100%以上，因为净资产的存在使总资产大于总负债，而存款又只是一部分银行总负债。

而对第二个问题，由于准备金存款余额由央行确定，银行之间如何拆借买卖都不能创造出更多的准备金，而存款当然也不能转化成准备金，因此银行作为整体，能做的只有收缩分母减少存款，使准备金率达标，或者向央行出售资产或拆借，以获得更多准备金存款。注意，央行借款给商业银行是同时扩张资产和负债的方式，增加银行在央行持有的准备金存款，以及增记对银行的债权，央行并不是把某种自身资产借给银行。银行减少存款可能伴随着缩表（例如客户还贷），也可能不缩表但改变债务种类，但资产侧准备金存款余额都是不变的，存款减少也不是用来交存了准备金。

如果说非要使用“存款交存准备金”之类的表述，则银行需要同时变动资产和负债，央行的负债也需要变化。第一个例子是财政收入与支出。税款进入国库时，银行资产侧的准备金存款减少，负债侧的存款余额减少，而央行的资产负债总额不变，负债侧的准备金存款减少，财政存款增加。另一个例子是《信用创造、货币供求与经济结构》中描述的，境外银行在央行开户，境内客户将在境内银行的存款转帐至该境外银行，转帐使境内银行整体的资产侧准备金存款和负债侧客户存款同时减少，央行负债侧准备金存款减少但国外负债增加。上述两个例子的特点都在于，操作需要同时涉及银行存款以及央行在准备金存款之外的项目。除了准备金存款，央行还有货币发行、金融性公司存款、发行债券、国外负债、政府存款、自有资金以及其他负债等项目，银行的客户只要参与上面的任何一项，理论上都会造成银行在央行准备金存款和客户在银行存款的同时变动。银行与银行间或者银行与客户间无论如何买卖借贷，只要不影响央行资产或者负债侧的其他项目，准备金总量就不会变化。

上述的一些评论仍然没有超出央行-银行以及贷款-存款的讨论框架，《中央银行与货币供给》一书正文也同样讨论的是同样的架构。但一旦出现以银行存款作为准备金的新一层架构，即央行-银行-余额宝，准备金制度则变得更复杂，该书的附录2与3关于余额宝和准备金的内容则是一次很好的讨论，下一篇文章将会专门评价这两篇附录的货币意义。

Published on June 4, 2015, in the Central Bank Observer.

The long-awaited book co-authored by Sheng Songcheng, Director of the Statistics and Survey Department of the People's Bank of China (PBOC), and Mr. Zhai Chun, titled "Central Banks and Monetary Supply," has finally been released. This book should be Sheng's new work following the publication of "Scale Theory and Practice of Social Financing" in April 2014. It is also another masterpiece of the People's Bank of China after the work by Li Bin and Wu Ge from the Monetary Policy Department and the Monetary Policy Operations Department of the People's Bank of China in December 2014, titled "Credit Creation, Money Supply and Economic Structure." "Central Banks and Monetary Supply" is a summary of classic monetary and banking analysis. Classic monetary and banking analysis starts with the definition and measurement of money, explores the demand and supply of money, the impact of base money and money multiplier on money supply, and the effects of monetary policy and transmission mechanisms on base money, money multiplier, money supply, and more. This is also the context of this book's development.

In the past when technology was underdeveloped, individuals in China were unable to open checks or have bank cards. They couldn't engage in transactions through transfers. Therefore, savings and deposits were not considered part of M1 and were classified as quasi-money. This was quite realistic. From this, we can see that by the end of 1999, the balance of currency in circulation was 13 trillion yuan, while demand deposits were 32 trillion yuan. Paper currency and coins were still the main means of payment in people's lives. However, today, the role of paper currency and coins has become increasingly weak. Perhaps the time of greatest usage is around the Spring Festival, and even this function is now being replaced by WeChat red envelopes. As of the end of March 2015, the balance of currency in circulation was 62 trillion yuan, but demand deposits reached 270 trillion yuan. Among the 54 trillion yuan of personal deposits at that time, a part was certainly demand deposits that could be used for buying and selling goods and services. For example, by the end of 2014, Industrial and Commercial Bank of China's personal deposit balance was 7.2 trillion yuan, of which 3 trillion yuan was demand deposits. Based on this estimation, the 54 trillion yuan of personal deposits could have roughly 20 trillion yuan of demand deposits. So, when comparing the 6.2 trillion yuan of currency in circulation with the over 45 trillion yuan of demand deposits and more than 45 trillion yuan of demand deposits, the proportion has decreased significantly compared to the end of 1999. Not to mention comparing it with the nearly 130 trillion yuan of M2 money supply.

The discussion about the importance of paper currency is aimed at revealing that in today's society, the influence of paper currency, a liability issued by the central bank (even though paper currency holders cannot redeem paper currency with the central bank for anything, paper currency is still recorded as a liability on the central bank's balance sheet according to historical practices), on the total money supply has decreased. This is consistent with people's experiences; the cash in the wallet is becoming less significant compared to the deposits in the bank. Therefore, the traditional notion that inflation is a result of printing money is outdated, and the core of China's current quantity-based monetary policy lies in adjusting the reserve deposit balance and the statutory reserve ratio. This is also the core content of Sheng's new work, the interplay between base money, statutory reserve ratio, and money supply.

The book "Central Banks and Monetary Supply" emphasizes that reserve deposits, as part of the base money, are a liability of the People's Bank of China. The scale of reserve deposits is entirely determined by the People's Bank of China, and any changes reflect changes in specific items or several items on the People's Bank of China's balance sheet. Of course, the mentioned "changes" could be due to the People's Bank of China's autonomous choice, or it could be a result of being forced to choose for other reasons. Especially in recent years, the exchange settlement by the People's Bank of China with commercial banks due to exchange rate policies has led to an increasing total amount of reserve deposits on the liability side due to the increasing foreign exchange settlements on the asset side. In order to maintain the exchange rate trend, the central bank can only treat its own balance sheet as a "pool" to maintain the desired exchange rate target, rather than standing aside and allowing commercial banks' trading to determine the exchange rate.

In addition to the foreign exchange settlement item, other items can also affect the balance of reserve deposits, such as lending to financial institutions and purchasing government bonds. However, most of the government bonds held by the central bank were acquired when China Investment Corporation was established, and this operation included changes in foreign exchange sales and government deposits. This process is relatively complex and is not discussed in this article. In a situation where total assets are unchanged, the central bank can also adjust the structure of its liabilities, changing the reserve deposit liability into central bank notes, or transforming reserve deposits into other liabilities through repurchase agreements. In addition, the demand for paper currency brought by red envelopes during the Spring Festival and the disturbance caused by fiscal revenue and expenditure through fiscal deposits can also affect the total amount of reserve deposits. Professionals in the fixed income market are very clear about the above mechanisms.

If reserve deposits connect the assets of commercial banks with the liabilities of the central bank, then the reserve requirement ratio links the liabilities of commercial banks with the liabilities of the central bank. As we all know, banks hold reserve deposits at the central bank as assets, and their liabilities include customer deposits. The ratio of reserve deposits to customer deposits is the reserve requirement ratio. The central bank sets a statutory reserve ratio, and the reserve requirement ratio of banks must exceed the statutory reserve ratio on specific dates. The difference between reserve deposits and statutory reserve deposits is excess reserve deposits, and the difference between the reserve requirement ratio and the statutory reserve ratio is the excess reserve requirement ratio. With the overall reserve deposit scale of banks unchanged, a higher statutory reserve ratio will reduce the excess reserve requirement ratio. If the overall banking system wants to maintain a certain excess reserve requirement ratio or excess reserve scale, banks can only reduce deposits. The traditional approach is to reduce lending and increase loan collection. When customers use deposits to repay loans, both the bank's loan assets and deposit liabilities decrease. The denominator of the reserve requirement ratio becomes smaller, the ratio increases, and the higher statutory reserve requirement ratio can be met.

For an individual bank, the perspective of considering the statutory reserve requirement ratio is often "how much reserve deposit is needed given a certain deposit base." However, for the central bank, the consideration is about the total deposits of the entire banking industry. Therefore, the perspective is "how to adjust the total deposit amount of the banking industry given a certain deposit reserve." Different perspectives give rise to situations where the central bank doesn't seem to be "well-behaved" in the eyes of banks. One is when banks push the central bank into a corner. Banks create a large amount of RMB deposits through foreign exchange transactions or lending, demanding that the central bank provide more reserve deposits or lower the statutory reserve ratio. Otherwise, they will impact the exchange rate and interest rate levels through interbank foreign exchange trading and RMB interbank borrowing and lending. The second situation is when the central bank is powerless against banks. The central bank attempts to ease, but banks, due to various considerations, withhold loans, leading to insufficient credit supply. Additionally, economic downturns can lead to insufficient credit demand. Therefore, adjustments to reserve deposits and the statutory reserve requirement ratio by the central bank cannot be reflected in the growth of bank deposits. This is the current relationship between European and American central banks and interbank lending, where excess reserve deposits are already quite high, but bank credit expansion is weak.

In China, the statutory reserve requirement ratio is quite complex. Firstly, it's a ratio. The numerator is the reserve deposits held at the central bank, and the denominator is a deposit range referred to by the PBOC as "general deposits" or "various deposits." However, different banks use different accounting items. For example, in the PBOC's "Notice on Adjusting the Scope of Reserve Deposits for Bank of Communications and China Merchants Bank" (April 2, 2013, Banking and Insurance [2013] No. 85), China Merchants Bank's scope of reserve deposits has 21 items, while Bank of Communications has 22. Different banks have various accounting items. Remarkably, China Bank even has an accounting item called "8429 Long-term Real Estate" within the scope. Recent Banking and Insurance [2014] No. 387 included certain interbank transactions within the scope of deposit reserve requirements, requiring deposit-taking financial institutions to submit a complete set of accounting items and explanations of the use of accounting items. The PBOC will also adjust the scope of deposit reserve requirements based on the above circumstances. Therefore, only the PBOC itself is probably clear about the information related to "general deposits."

Even if the accounting items requiring deposit reserve requirements for all institutions are consistent and their balances are fully disclosed, it's difficult to calculate how much statutory reserve deposits are required at the central bank. This is because the reserve requirement ratio is not the same for different institutions. This is clearly reflected in the announcements of previous reserve requirement ratio adjustments, and especially in targeted reserve requirement ratio adjustments. Furthermore, the PBOC introduced a differentiated reserve requirement dynamic adjustment mechanism at the end of 2010. Not only does the reserve requirement ratio change based on bank operating conditions, but the formula also incorporates new parameters. The numerical values of these parameters also change. Therefore, for a specific institution, the results of statutory reserve requirement ratio adjustments announced by the PBOC are only for reference. In addition, banks' foreign currency deposits also need corresponding foreign exchange reserves deposited at the central bank. Moreover, a portion of RMB deposit reserve deposits is deposited in foreign currency. This complex mechanism makes market analysis based on public data almost impossible.

Many people think that lowering the reserve requirement ratio means banks have a lot of money for lending, while raising it means banks need to pay more reserve deposits. The use of phrases like "deposit reserve" is vivid in practice, but in theory, the PBOC clearly understands the mechanism of how changes in the statutory reserve requirement ratio affect the market. This is also reflected in "Central Banks and Monetary Supply," as well as other works by the PBOC and other authors. Banks cannot use their own deposits to deposit reserve deposits at the central bank, as banks' own deposits are on the liability side, while deposits at the central bank are assets. The former cannot be transformed into the latter. Furthermore, for individual banks, the perspective of considering the statutory reserve requirement ratio is often "how much reserve deposit is needed given a certain deposit base." However, for the central bank, the consideration is about the total deposits of the entire banking industry. Therefore, the perspective is "how to adjust the total deposit amount of the banking industry given a certain deposit reserve." Different perspectives give rise to situations where the central bank doesn't seem to be "well-behaved" in the eyes of banks.

The above analysis still falls within the framework of the central bank-bank and loan-deposit discussions. The main text of the book "Central Banks and Monetary Supply" also discusses the same framework. However, once a new layer of structure emerges, where bank deposits are treated as reserve deposits, the reserve requirement system becomes even more complex. The content of Appendices 2 and 3 in the book regarding Yu'ebao (a money market fund product) and reserve deposits is a good discussion in this regard. The next article will specifically evaluate the monetary significance of these two appendices.