原创 2016-02-06 季天鹤央行观察央行观察

结售汇不是外汇交易

一个常见误解，是认为居民把人民币存款变为美元存款是在购入外汇。的确，把人民币存款变为美元存款后，储户不再享有人民币利率回报，也无法进行人民币汇款，而是享有美元利率回报，可以进行美元汇款，并且人民币贬值美元升值时，美元存款的人民币估值会因人民币贬值而增加。由此，人们广泛认为在境内的美元存款就是美元，结售汇是一种外汇交易。

但一笔交易肯定包括买方和卖方，如果结售汇是一笔外汇交易，居民是美元买方，那么卖方是谁呢？或许上述人士会说卖方是银行，但在居民把人民币存款变成美元存款的过程中，银行的资产负债规模是不变的，并且银行的资产侧是完全不发生变化的。这和股票交易不同：在股票交易中，股票买方和卖方的资产都要变化，一个出钱收券，一个出券收钱。

结售汇不是外汇交易，人们往往关注银行客户的感受。从银行客户的角度讲，结售汇前后他们的资产有变化，感觉上是交易。但是从银行的角度讲，结售汇其实并不是一种交易，而是一种特殊的金融服务。同样是负债，我们没有听说过银行人民币债券的持有者某天突然找银行要求变成美元债，这也说明了结售汇的特殊性。

银行的对冲操作

有观点认为，储户不停地购汇，银行发现自己负债侧美元存款越来越多，为了减少汇率变动对自己的影响，会购入等量的美元。对一家银行而言，买入美元意味着另一家银行要卖出美元。而当所有银行都要买入美元的时候，总要有一个最终的美元卖者。当然，这个最终的美元卖者在现实中就是外汇储备。

上述观点的逆过程在2008年以前人民币升值的进程中比较明显。当外汇不断进入中国，最初始的状况是银行资产侧持有外币，银行负债侧是储户的外币存款。但当储户将外币存款通过结汇变成人民币存款后，银行便出现了币种的错配。如果银行不找人行结汇，而是自己持有上述外汇，那么银行资产的10%以上就都是外汇资产。

考虑到银行当时在进行改革并准备上市，让银行承担10%的资产币种错配，以及20%水平的外币贬值损失，恐怕报表不够好看，银行自己当然更不愿意。所以我们看到，2008年以前的情况是银行找人行结汇，自己在人行负债侧持有人民币准备金资产，人行在资产侧持有外币资产。这样，人行代替了银行承担外币贬值的汇率损失。

不过，2008年前人民币升值和2015年起人民币贬值，尽管在价格上是对称的一升一贬，但在数量上则并不对称。人民币升值过程中，银行创造人民币的能力是无限的。储户如果汇来10万亿美元，银行也能轻松创造出来65万亿人民币存款出来。这是由于汇来外汇这一操作，对银行来说是资产负债双增，而资产负债的增长并没有天花板。因此，银行最为担忧的，乃是币种错配造成的估值损失。

但人民币贬值则不然。首先，银行负债侧的人民币存款不都是客户结汇的结果，很多乃是银行放贷而创造出的人民币存款，而这部分人民币存款并不像客户结汇的人民币存款那样，有外币在银行资产侧作为对应。其次，当储户把人民币存款变成外币存款汇走，这个过程会减少银行的外汇资产，而这个减少幅度则是有上限的，一旦耗尽便再无外汇可以汇走。

因此在贬值时期，银行最担心的并不是价格变动，而是数量，即自己资产侧的外汇流动头寸不足以应对储户购汇并汇出的操作。2016年起人民币贬值对银行的压力，和2008年前人民币升值对银行的压力，并不相同。银行从外汇储备购买外汇的操作，并不是为了应对资产负债币种错配的损失，而是应对储户汇出外币的数量压力。

目前的情况

2015年8月和9月，外汇局口径的外币每月净流出均高达4600亿人民币等值，10月份为内流126亿人民币等值，11月和12月又变成外流1600亿人民币等值和2000亿人民币等值，共计12000亿。而观察银行业的国外资产，2015年7月底为42000亿等值人民币，12月底为41600亿等值人民币，减少仅为400亿人民币。

可见，上述超过12000亿人民币等值外汇汇出，主要是由央行提供支持，其中包括5000亿元人民币等值的商业银行外汇缴纳人民币存款准备金。我们知道，央行国外资产从7月底的273500亿元人民币等值下降至253800亿元人民币等值，减少20000亿元，主要应对了上述12000亿人民币等值的外汇汇出。

剩下8000亿元人民币等值差额的奥秘，在于银行的国外负债从7月底的20000亿元人民币等值，下降至12月底的13000亿人民币等值，减少7000亿人民币等值。这部分反映了央行在离岸的主动外汇干预，以及境外银行把在境内的人民币变为外币取出，而境外银行的这个操作恐怕反映的是离岸客户的操作。

同时我们看到，境内银行的境内外汇贷款，从2015年7月底的5712亿美元下降至2015年12月底的4656亿美元，降幅超过1000亿美元。这意味着，有约7000亿人民币的人民币存款或者等值外币存款被用于还贷。按理说这样多的还贷和出逃，应该会使得外币存款规模急剧下降。但事实上从2015年7月底到12月底，境内外币存款仅从5342亿美元下降至5262亿美元，即600亿等值人民币左右。

上述数据说明，真实发生的事情是三种情况：一些人在汇出外币，这里的外币可能是购汇的结果，也可能是原有的外币存款，其中包括央行的干预；一些人在归还境内外币贷款，这里的外币可能是购汇的结果，也可能是原有的外币存款；一些人则在增持外币存款，使得前两个途径减少的外币存款，在这里又被补齐。外汇储备和储户在银行外汇存款的关系，并不是简单对应的。只有当储户整体产生对外转账要求，且银行整体不能满足这一要求时，银行整体才会找外汇储备解决问题。

如果不考虑转账带来的数量压力，而只考虑汇率变动的价格压力，央行完全可以缓解银行的损失而不消耗外汇储备。如果银行要的只是避免人民币贬值造成的人民币利润下降，那么央行完全可以通过和银行进行人民币结算损益的不可交割远期甚至即期交易来帮助银行解决这个问题，银行亏钱了央行补给人民币，反正人行创造人民币的能力无限。

并且，央行资产侧有很多外汇，而负债侧都是人民币，人民币贬值对央行来说是带来利润的，此外，这一操作不影响外汇储备，因为是人民币交割的。当然上述策略只是一个假想情况，但本身其实说明，价格显然不是银行在目前人民币贬值下面临的最大挑战，因为这个挑战比跨境汇款带来的挑战容易应对得多。

下一阶段的形势

尽管从价格上，居民通过在银行购汇来期待人民币贬值中获取利润，但居民得到的并不是国外资产，而是一种国外资产的兑现权，并且这种兑现权在资本项目管制以及居民缺乏手段的现实情况下，基本上没有真正的兑现功能，而居民或许本来也没有兑现的意图，只是想挣点钱，即行权能力和行权意愿两者都很有限。

无论是直接持有境外外币还是在国内持有外币，从以搏收益为目标的资产配置方面，两者带来的损益是一样的，都由人民币汇率变动决定。但在国内持有外币和在境外持有外币，在外汇储备和汇率稳定上的意义上则完全不同，前者不涉及对外汇储备的提取，而后者则在目前的情况下涉及。

如果中国居民通过在境内增持美元存款的方式实现资产的多币种配置，而不是通过在境外持有的方式，那么这一配置方式对于人民币汇率改革将是非常好的，可以为实体经济的转型和升级争取更多时间，最终实现由实体经济情况决定人民币汇率水平，由市场主体自身的操作应对人民币汇率波动。而从准备金率的角度，这也是我在之前多篇文章中指出的，让外汇准备金率提供弹性，来应对购汇需求。

人行目前的牌出得还是不错的，从行政和市场两个角度打击以金融稳定为代价的跨境投机类汇款，同时通过商业银行各种拖拉来延缓美元现钞取现。人们配置美元的热情最终被引导到增持境内美元存款上，反正大家的目的也不过是博收益。上述操作反映在外汇信贷收支表上，则是外汇存款余额不因还贷和对外汇出而减少，反而因居民购汇而增加。期待2016年的外汇信贷收支表提供答案。

A common misconception is that when residents convert their Chinese yuan deposits into US dollar deposits, they are purchasing foreign exchange. Indeed, after converting Chinese yuan deposits into US dollar deposits, depositors no longer enjoy Chinese yuan interest rate returns, can't make Chinese yuan remittances, and instead enjoy US dollar interest rate returns, can make US dollar remittances, and when the Chinese yuan depreciates and the US dollar appreciates, the Chinese yuan valuation of US dollar deposits will increase due to the Chinese yuan's depreciation. As a result, many people widely believe that US dollar deposits within China are indeed US dollars, and foreign exchange transactions involve buying and selling foreign exchange.

However, a transaction inherently involves a buyer and a seller. If foreign exchange transactions involve buying and selling foreign exchange, and residents are the buyers of US dollars, then who is the seller? Some might argue that the seller is the bank, but during the process of residents converting Chinese yuan deposits into US dollar deposits, the bank's asset and liability scale remains unchanged, and the bank's asset side remains entirely unchanged. This differs from stock trading: in stock trading, the assets of both the stock buyer and seller change – one pays money and receives stocks, while the other pays stocks and receives money.

Foreign exchange transactions are not equivalent to foreign exchange trading; people often focus on the experience of bank customers. From the perspective of bank customers, their assets change before and after foreign exchange transactions, creating a sense of trade. However, from the bank's perspective, foreign exchange transactions are not actual trades; instead, they are a form of specialized financial service. Similarly, as liabilities, we've never heard of holders of bank Chinese yuan bonds suddenly asking the bank to convert them into US dollar bonds one day, indicating the uniqueness of foreign exchange transactions.

Bank's Hedging Operations

Some viewpoints suggest that if depositors continuously purchase foreign exchange, banks find themselves with increasing liabilities in the form of US dollar deposits. To reduce the impact of exchange rate fluctuations on themselves, banks will purchase an equivalent amount of US dollars. For a bank, buying US dollars means that another bank must sell US dollars. And when all banks need to buy US dollars, there must be a final US dollar seller, which in reality, is the foreign exchange reserves.

The reverse process of the above was more evident during the period of the Chinese yuan's appreciation before 2008. As foreign exchange continuously entered China, the initial situation was that banks held foreign currency assets on the asset side, and bank liabilities were in the form of customers' foreign currency deposits. However, when customers converted foreign currency deposits into Chinese yuan deposits through foreign exchange transactions, banks experienced a mismatch in currency. If banks did not seek settlement with the central bank but instead held the above foreign exchange themselves, then over 10% of bank assets would be foreign exchange assets.

Considering that banks were undergoing reforms and preparing for listing at the time, it would have been unattractive to let banks bear a 10% currency mismatch on their assets and a 20% foreign currency depreciation loss. Therefore, the situation before 2008 was that banks sought settlement with the central bank, holding Chinese yuan reserve assets on the central bank's liability side while the central bank held foreign currency assets on its asset side. Thus, the central bank replaced banks in bearing the exchange rate loss due to currency depreciation.

However, during the period of Chinese yuan depreciation starting in 2015, the situation is different from that before 2008, even though they appear symmetric in terms of price changes (one being appreciation and the other depreciation). During the Chinese yuan's appreciation, banks had the unlimited ability to create Chinese yuan. If customers brought in $10 trillion US dollars, banks could easily create around ¥65 trillion Chinese yuan deposits. This is because the foreign exchange operation for bringing in foreign exchange is a double increase in assets and liabilities for banks, with no ceiling on the growth of assets and liabilities. Thus, banks' primary concern was the valuation loss caused by currency mismatch.

However, Chinese yuan depreciation does not work the same way. Firstly, not all Chinese yuan deposits on the liability side of banks are a result of customer foreign exchange transactions; many are Chinese yuan deposits created by banks lending, and these Chinese yuan deposits do not have corresponding foreign exchange on the asset side like customer foreign exchange deposits. Secondly, when customers convert Chinese yuan deposits into foreign currency deposits and remit them out, this process reduces the bank's foreign exchange assets, and this reduction has an upper limit – once exhausted, there is no more foreign exchange that can be remitted.

Thus, during periods of depreciation, what banks are most concerned about is not price changes but quantity – that is, the shortage of foreign exchange flow positions on the asset side to accommodate customers' foreign exchange purchases and remittances. The pressure on banks from the Chinese yuan's depreciation since 2016, and the pressure on banks from the Chinese yuan's appreciation before 2008, are not the same. Banks buy foreign exchange from foreign exchange reserves to counter the pressure of the number of remittances from customers.

Current Situation

In August and September of 2015, the net outflow of foreign currency according to the State Administration of Foreign Exchange reached an equivalent of ¥460 billion per month, and in October, it turned into an inflow of ¥12.6 billion, followed by net outflows of ¥160 billion and ¥200 billion in November and December, totaling ¥1.2 trillion. When observing the foreign assets of the banking industry, it was ¥4.2 trillion at the end of July 2015 and ¥4.16 trillion at the end of December 2015, a decrease of only ¥40 billion.

This implies that the significant outflow of over ¥1.2 trillion is primarily supported by the central bank, including ¥500 billion of commercial banks' foreign exchange deposit reserve requirements. It's known that the central bank's foreign assets decreased from ¥27.35 trillion at the end of July to ¥25.38 trillion at the end of December, a decrease of ¥2 trillion, mainly to counter the above ¥1.2 trillion outflow.

The remaining mystery of the ¥800 billion difference lies in the fact that the foreign liabilities of banks decreased from ¥2 trillion at the end of July to ¥1.3 trillion at the end of December, a decrease of ¥700 billion. This reflects the central bank's active foreign exchange intervention offshore and foreign banks converting their Chinese yuan holdings into foreign currency and withdrawing them. This operation by foreign banks likely reflects the actions of offshore customers.

At the same time, foreign exchange loans of banks, which were at $571.2 billion at the end of July 2015, decreased to $465.6 billion at the end of December 2015, a decrease of over $100 billion. This means that approximately ¥700 billion of Chinese yuan deposits or equivalent foreign currency deposits have been used to repay loans. In theory, such a significant repayment and outflow should result in a sharp decrease in the foreign currency deposit balance. However, from the end of July 2015 to the end of December 2015, foreign currency deposits decreased only from $53.42 billion to $52.62 billion, a decrease of only around ¥60 billion.

The above data illustrates that three scenarios have occurred: some individuals remitting foreign exchange, where the foreign exchange might be a result of foreign exchange purchases or existing foreign currency deposits, including central bank interventions; some individuals repaying domestic and foreign currency loans, where the foreign exchange might be a result of foreign exchange purchases or existing foreign currency deposits; and some individuals increasing foreign currency deposits, compensating for the decrease in foreign currency deposits through the first two methods. The relationship between foreign exchange reserves and bank foreign exchange deposits is not a simple correspondence. Only when customers as a whole generate demands for external transfers and banks as a whole cannot meet this demand will banks as a whole seek foreign exchange reserves to solve the problem.

If we don't consider the quantity pressure brought by transfers, only the price pressure brought by exchange rate changes, the central bank can completely alleviate the bank's losses without consuming foreign exchange reserves. If all banks need is to avoid a decrease in Chinese yuan profits due to Chinese yuan depreciation, then the central bank can fully address this issue by engaging in non-deliverable forward or even spot transactions with banks to settle Chinese yuan profit and loss. If banks are losing money, the central bank provides Chinese yuan; after all, the central bank's ability to create Chinese yuan is unlimited.

Furthermore, the central bank has many foreign exchange assets on its asset side and liabilities mostly in Chinese yuan on its liability side. Chinese yuan depreciation brings profits to the central bank. Moreover, this operation does not affect foreign exchange reserves because it involves Chinese yuan settlement. Of course, the above strategy is only hypothetical. However, it demonstrates that price pressure is clearly not the primary challenge that banks are facing under the current situation of Chinese yuan depreciation. This challenge is much easier to deal with than the challenges brought by cross-border remittances.

Next Phase Situation

Although in terms of price, residents anticipate making profits through purchasing foreign exchange from banks in the hope of benefiting from Chinese yuan depreciation. However, residents do not actually obtain foreign assets; they gain a realization right to foreign assets. Given the reality of capital controls and the lack of means for residents, this realization right practically has no true realization function. Residents may not even have the intention to realize it; they just want to earn some money, meaning their ability and willingness to exercise these rights are both limited.

Regardless of whether directly holding foreign currency abroad or domestically, from the perspective of asset allocation aimed at profit, the gains and losses are the same, determined by Chinese yuan exchange rate fluctuations. However, in terms of the significance of foreign exchange reserves and exchange rate stability, holding foreign currency domestically and abroad are entirely different. The former does not involve the extraction of foreign exchange reserves, while the latter does under current circumstances.

If Chinese residents achieve multi-currency asset allocation by increasing domestic US dollar deposits rather than holding foreign currency abroad, this method would be beneficial for the Chinese yuan exchange rate reform. It would buy more time for the transformation and upgrading of the real economy and ultimately lead to the Chinese yuan exchange rate being determined by the situation of the real economy, with market participants responding to Chinese yuan exchange rate fluctuations. From the perspective of reserve requirements, this is what I've mentioned in previous articles: allowing foreign exchange reserve requirements to be flexible in order to meet foreign exchange demand.

The central bank's current strategy is working well. By combating cross-border speculative remittances at the cost of financial stability, and by delaying customers' withdrawal of US dollar cash through various tactics of commercial banks. People's enthusiasm for holding US dollars is ultimately guided towards increasing domestic US dollar deposits, since everyone's goal is just to make a profit. The above operation is reflected in the foreign exchange credit and debit statement, with foreign exchange deposits not decreasing due to loan repayments and foreign exchange outflows but increasing due to residents' foreign exchange purchases. We await the foreign exchange credit and debit statement for 2016 for more answers.