By Henry C K Liu
Oct 27, 2005
Source(
Part I)
(
Part II)
In order to understand the systemic risk implications of the astronomical growth of the repo (repurchase agreement) market, it is necessary to have some basic knowledge of the dollar money market, of which the repo market has become a major component.
The telling information is that repos now chalk up an average daily trading volume of about US$5 trillion in the United States, accounting for half of the money supply ($9.97 trillion as of September). Conventional perception notwithstanding, the repo market is no longer as risk-free as presumed because the money-creating proceeds from repos are mostly channeled toward speculation that contributes to systemic risks. The risk of contagion, a term given to the process of distressed contracts dragging down healthy contracts in the same market as speculators throw good money after bad to try to stem the sudden tide of temporary losses, can be magnified by the widespread use of repos.
The money deposited by commercial banks at the central bank is the real money in the banking system; other versions of what is commonly thought of as money are merely promises to pay real money. These promises to pay are circulatory multiples of real money. For general purposes, people perceive money as the amount shown in financial transactions or amount shown in their bank accounts. But bank accounts record both credit and debits that cancel each other. Only the remaining central-bank money after aggregate settlement is real. Real money, more properly called final money, can take only one of two forms: 1) physical cash, which is rarely used in wholesale financial markets, and 2) central-bank money. The currency component of the money supply is far smaller than the deposit component.
Currency and bank reserves together make up the monetary base, sometimes known as high-powered money. The US Federal Reserve has the power to control the issuance of both of these components of the monetary base. By adjusting the levels of banks' reserve balances, the Fed can achieve a desired rate of growth of deposits and of the money supply over time. When market participants and banks change the ratio of their currency and reserves to deposits, the Fed can offset the effect on the money supply by changing reserves and/or currency. The Fed and the US Treasury supply banks with the currency their customers demand, and when their demand falls, accept a return flow from the banks. The Fed debits banks' reserves when it provides currency, and credits their reserves when they return currency. In a fractional reserve banking system, drains of currency from banks reduce their reserves, and unless the Fed provides adequate additional amounts of currency and reserves, a multiple contraction of deposits results, reducing the quantity of money in the economy.
Banks can create loan money out of central-bank high-power money from a partial reserve regime. For example, $1,000 of high-power money with a bank reserve requirement of 10% can create $7,922 of new loans, each time holding back 10% reserve from the resultant deposit from the borrower before lending it out again. As the bank reserve requirement falls, the amount of bank-created loan money rises. Money deposited by customers with a commercial bank is not final money, as it is merely a promise to pay by one party to another. When a bank customer writes a bank check to pay another party, the money in his account shifts to the payee's account within the banking system, even if the shift is between banks. No money is added or removed from the money supply. Money is merely shifted from one account to another within the same bank or from one bank to another within the banking system.
The US central bank, the Federal Reserve System, is composed of 12 regional reserve banks, and money at any of these regional reserve banks is final money. Most banks use an account at the Federal Reserve Bank of New York to settle US dollar activities in the wholesale financial markets. The European Central Bank (ECB) is part of the European system of national central banks, which includes the Bundesbank (Germany's national central bank), the Banque de France, the Banca d'Italia, and others in the eurozone. Money denominated in euros at any one of the eurozone's national central banks (NCBs) is final money. In sum, the monetary definition of "money" is money on account at the central bank. Any other form of money is really just a promise to pay central-bank money before final clearing.
The main purpose of an interbank deposit market, a money market, is to equalize the payment system among banks. Banks attract deposits by offering interest payments. They make a profit out of deposits from customers by lending the deposited money at a higher rate to other customers or to other banks. Banks lend money to one another in all major currencies of market relevance in globalized financial markets at agreed interest rates and agreed maturities. Typical maturities for interbank money range from one day for overnight money to six months, and even out to one year, though with much less active trading in the longer maturities. Funds at each bank are swept electronically at the end of each business day for interbank lending.
The money market is so important that many banks maintain active screens showing the latest prices at which they are willing to borrow and lend. At any time on a trading day, a major money-center bank would post a particular rate at which it is willing to accept three-month deposits, say 3.4%, and to lend to other high-quality banks for the same period at a higher rate, say 3.45%. The bank is making a market in these deposits, bidding for three-month funds at one rate and offering three-month funds at a higher rate. The intent of such market-making is to profit from the bid-offer spread. Banks lend to customers at prime rate plus, usually a few percentage points higher than they can borrow at interbank rates. A bank with excess funds that cannot be lent at prime rates will lend to other banks with needs to lend more than funds at their disposal.
An institutional/corporate borrower that wants to borrow funds for six months has several options. One would be to borrow funds for six months from the commercial bank at the screen price of, say, 3.53%. But there are alternatives. It can borrow the money for only three months (at the rate of 3.45%), and after three months re-borrow the money, or roll over the loan at the higher or lower prevailing rate. Why do this? To have the same cost as a six-month loan, the re-borrowing would have to be at 3.58%, as the six-month rate is the average of 3.45% and 3.58%, the rates for the first and second three-month periods respectively. If the borrower thinks that in three months time, the cost of three-month money will be less than 3.58%, then it would be cheaper overall for it to borrow now for three months at 3.45%, and then in three months to re-borrow at the rate then prevailing. If it thinks that in three months time the cost of borrowing for three months is likely to be higher than 3.58%, it should borrow for the entire six-month period now.
So the break-even cost of three-month money in three months time is 3.58%. This is the forward price of money. The current price, also known as the spot price, of three-month money is 3.45%; the three-month forward price of three-month money is 0.13% over spot. Market prices are implying that short-term interest rates are rising. The borrowing institution or corporation has still more choices. If it believes that rates are unlikely to rise, then it might be cheapest to borrow for one day, and re-borrow the money each subsequent day. Or if it thinks that rates are about to rise to very high levels, the best course would be to borrow money for one year (at 3.65%), and in six months to lend at the then-prevailing rate, which it hopes will be much higher then.
No matter which view it takes, by choosing to borrow money at one maturity rather than another, the borrower is implicitly expressing an opinion on the future path of short-term rates. That opinion is measured, and can only be measured, against the current forward prices. A long-term borrower cannot avoid some form of implicit speculation; by choosing to borrow at one maturity rather than another, it is taking a view on the future path of interest rates, and that view should be measured against the market's forward prices. Thus risk is inherent in any financial commitment over time, even if both the rate and duration are fixed, in which case the risk is in opportunity cost. The risk for each market participant is related to the opinions of other market participants. Contrarians to market sentiment risk profit and loss.
This is the basic logic of technical analysis, which is concerned with market behavior resulting from market sentiment, as opposed to fundamental analysis, which is concerned with economic fundamentals of supply and demand. As John Maynard Keynes famously said, markets can stay irrational longer than market participants can stay liquid. The only exception is the central bank operating in a fiat currency regime. Only the Fed has the power to defy market sentiments, which explains why market participants try desperately to second-guess pending Fed decisions on interest rates by informing on the ideology and decision-making rationale of the Fed.
The same applies to exchange rates of currencies. By selecting which currency to borrow or lend and the rates and maturities at which loans are structured, borrowers and lenders express their views on the future path of exchange rates of particular currencies. The aggregate effect of market-participant opinions determines the market price of money and the exchange value of all freely convertible and floating currencies. This is how currencies are at times attacked by speculators who manage to create market sentiments against particular currencies regardless of fundamentals.
Central banks intervene from time to time to defy market sentiments by making market speculation unprofitable. In recent decades, the spectacular increase of the size of the foreign-exchange market has made central-bank intervention less effective. Though not the biggest market participants, central banks have unlimited funds denominated in their national currencies to effectuate short-term rates within a policy framework. For US dollars, the Fed calls all the shots, as it alone can print dollars at will. And since the US dollar is the world's major reserve currency for international trade and since all key commodities are denominated in dollars, the Fed commands a disproportional influence on the global money markets.
Insolvency risk, also known as credit or default risk, is faced by all market participants except the Fed, which alone can print dollars at will. Foreign central banks can print national currencies but they cannot print US dollars. Thus all foreign central banks face insolvency risks on dollar obligations. When that happens, foreign central banks have to turn to the International Monetary Fund (IMF) as a lender of last resort to avoid default on their dollar obligations. Banks deal not only with one another and not only with top-quality financial institutions from countries with clear and competent financial supervision, but also with riskier entities (individuals, trusts, partnerships, companies, or even governments). With some of these entities, the risk of insolvency is significant and sometimes unpredictable.
A solution to credit risk can be found in the repo market. The bank lends money to a borrower collateralized by government bonds of the same market value. Under normal circumstances, after the loan matures, the borrower returns the money plus interest, and the bank returns the collaterals. But if the borrower should become insolvent or is forced to default on the loan for any reason, the lender can recover its loss by selling the collateral that it holds.
Interest-rate volatility affects the market value of securities, even government securities. Thus the Fed's interest-rate policy affects the risk level of the repo market. So the money market is a game in which market participants guess the future decisions of the Fed, and their implication on market forces. Both the supply of money and the demand for money as affected by its price (interest rates) are set by the Fed based on macroeconomic theories that are ideologically derived, not scientifically validated. On this arbitrary monetary foundation set by a handful of appointed individuals is built free-market capitalism. On one level, the institution of central banking is politically independent; on a more fundamental level, central banking defies the most basic principles of democracy and free markets. For the lender, the advantage of collateralization is that it almost eliminates credit risk. For the borrower, the disadvantage is that collateral must first be found. The borrower's disadvantage and the lender's advantage are reflected in the price of the loan. The interest rate on a collateralized loan is below that on a non-collateralized loan. Within the same currency, the spread varies across maturities and according to the quality of the collateral and the counterparty, but a typical unsecured-secured differential is 0.1% to 0.5% for borrowers of equal credit rating.
Commercial-bank accounts with the central bank are subject to different rules about overdrafts in different countries. Some central banks prohibit overdrafts; at the end of each day no account may be overdrawn. Other central banks are less strict, specifying that every account must have a positive balance on average, where the averaging is conducted over a period of time known as a reserve period. Whichever the case, commercial banks need to avoid having an overdraft at the central bank, either on average or every day by borrowing money from another bank.
Bank-to-bank borrowing can only relocate rather than extinguish the aggregate overdraft in the banking system. The escape to systemic aggregate overdraft is to borrow money directly from the central bank or from the repo market when the Fed injects money into it through the monetizing of sovereign debt by lending against government securities held by non-bank entities. Government securities sold to private buyers caused money to be withdrawn from the economy. This money re-enters the economy when the government spends it. When the holders use such securities as collaterals to secure new loans in the repo market, new money is in fact created.
The discount rateThe greatest power bestowed on the Federal Reserve is the setting of the discount rate - the rate of interest charged by the Reserve Banks when lending to member institutions, collateralized by government securities. Raising the discount rate generally increases the cost of borrowing and tightens the money supply and slows the economy, while dropping it expands the money supply and stimulates economic activity, since banks set their loan rates above the discount rate, and not by market forces. Fed discount rates and Fed funds rate targets are inputs into market conditions, not outputs from market forces
The term "discount rate", although widely used, is an anachronism. Since 1971, Reserve Bank loans to depository institutions have been secured by advances. Interest is computed on an accrual basis and paid to the Fed at the time of loan repayment. The interest rate charged by a Federal Reserve Bank on short-term loans to depository institutions is referred to as the discount rate.
The discount rate is important for two reasons: (1) it affects the cost of reserves borrowed from the Fed and (2) changes in the rate can be interpreted as an indicator of monetary policy. Increases in the discount rate generally reflect the Fed's concern over inflationary pressures, while decreases reflect a concern over economic weakness or deflation. Discount-window lending, open-market operations to effect changes in reserves to set Fed funds rates, and bank reserve requirements are the three main monetary-policy tools of the Federal Reserve System. Together, they dictate the short-term cost and availability of money and credit in the US.
Usually, the discount rate is less than the federal-funds and other money-market interest rates. However, the Fed does not allow banks to borrow at the discount window for profit. Thus it monitors discount-window and federal funds activity to make sure that banks are not borrowing from the Fed in order to lend at a higher rate in the federal funds market.
During periods of monetary ease, such as now, the spread between the federal funds and discount rates may narrow or even disappear briefly because depository institutions have less of a need to borrow reserves in the money market. Under these conditions, the Fed may adjust the discount rate in order to reestablish the customary spread.
Discount-window loans are granted only after Reserve Banks are convinced that borrowers have fully used reasonably available alternative sources of funds, such as the federal funds market and loans from correspondents and other institutional sources. The latter sources include the credit programs that the Federal Home Loan Banks and the Central Liquidity Facility of the National Credit Union Administration provide for their members. Usually, relatively few depository institutions borrow at the discount window in any one week. Consequently, such lending provides only a small fraction of the banking system's total reserves. All depository institutions that maintain reservable transaction accounts or non-personal time deposits are entitled to borrow at the discount window. This includes commercial banks, thrift institutions, and US branches and agencies of foreign banks. Prior to the passage of the Depository Institutions Deregulation and Monetary Control Act of 1980, discount-window borrowing generally had been restricted to commercial banks that were members of the Federal Reserve System.
Changes in the discount rate generally have been infrequent. From 1980 through 1990, for example, there were 29 rate changes, and the duration of the periods between adjustments ranged from two weeks to 22 months. However, after those 22 months without a change, the Fed cut the discount rate seven times in the period of economic sluggishness from December 1990 to July 1992 - from 7% at the start of the period to 3% at the end. From May 1994 to February 1995, when the Fed was concerned about the threat of inflation, it raised the discount rate four times - from 3% to 5.25%. Changes in the discount rate often lag changes in market rates. Thus, even though the Fed pushed the federal funds rate down 25 basis points in July 1995, as of December it had not cut the discount rate.
Since 1980, the changes in the discount rate have been by either one-half or a full percentage point, although quarter-point changes were made in earlier years. The lowest discount-rate charged by the New York Fed was 0.5%, which was in effect from 1942 through 1946; the highest rate was 14%, which lasted from May to November 1981. In 1999, the discount rate was 4.5%, while the Fed funds rate was 4.75% and the three-month Treasury-bills rate was 4.27%. On September 13 this year, the discount rate was 4.5% (effective since August 9) while Fed funds rate target was 3.5% (also effective since August 9), the overnight repo rate was 3.45%, three-month T-bill was 3.45%, GE 30-44-day commercial paper was 3.68% and three-month LIBOR (London interbank offered rate) was 3.87%. Effective September 20, the discount rate was raised to 4.75%; the Fed funds rate target was raised to 3.75%. On October 19, the overnight repo rate was 3.70%, three-month T-bill was 3.785%, GE 30-44-day commercial paper was 3.90% and three-month LIBOR was 4.35%.
Today, while the spread between the discount rate and the Fed funds rate is zero, the repo market continues to expand, giving evidence that the borrowing is not being done by depository institutions. In other words, the credit market has largely run away from banks.
Fed fundsFederal funds include funds deposited by commercial banks at the Federal Reserve Banks, including funds in excess of bank reserve requirements. But Fed funds can be created at will by the Fed now that the dollar is a fiat currency, not backed by gold, its status since 1971.
Commercial banks may lend federal funds to one another on an overnight basis at the Fed funds rate, which is the most sensitive indicator of the direction of interest rates since it is set daily by the market in response to the Fed's open-market operation: the buying or selling of government securities in the repo market to meet Fed funds rate targets. Thus the real function of sovereign debt is to provide an instrument through the buying and selling of which the Fed can inject money into or withdraw money from the money supply without appearing to create or destroy money while actually doing so.
Notwithstanding conventional perception, sovereign debt is not incurred by fiscal deficits, which are created by government reluctance to raise taxes to cover expenditure. Fiscal deficits are government's way to inject money into the economy independent of the central bank, to avoid raising taxes, which is government's way of withdrawing money from the economy. A government fiscal surplus shrinks the economy in two ways: it drains money from the money supply and it prevents money from re-entering the economy through government spending. Thus fiscal deficits are not inherently bad for the economy, especially for an economy that is underperforming. It all depends on how the deficit spending is used. If it is used for increasing employment and improving infrastructure or education or health care, thus expanding the slowing economy, it is good; if it is used to make war or speculation, thus contributing to debt bubbles, it is bad.
The US money market is basically for short-term debt instruments generally collateralized by Treasury bills as default-free securities. Short-term interest rates are basically determined by the Fed's action in the money markets, in which instruments such as Treasury bills, commercial deposits, euro-deposits, commercial paper, and repurchase agreements are traded. It is in the money markets that the Federal Reserve conducts most of its transaction activity and attempts to meet its short-term interest-rate target.
Two key interest rates dominate the activity of the US money markets, the federal funds rate and the repo rate. These two rates, though closely connected, serve opposite functions. In normal usage, both are overnight rates based upon a 360-day year. The federal funds rate is the interest rate that banks charge when they lend reserves to one another. Money-supply calculations are dominated by consideration of bank demand and time deposits, which lie on the liability side of a commercial bank's balance sheet. Reserves make up a part of the bank's assets. Banks that are members of the Federal Reserve System are required to hold cash reserves against their deposit liabilities. The rules for the reserve requirements, once very simple, are now quite entangled.
Even if reserves were not a legal requirement, prudence would ensure that banks would hold a certain percentage of their assets in the form of cash reserves. It is common to think of commercial banks as passive receivers of deposits from their customers and, for many purposes, this is still an accurate view. This passive view of bank activity is misleading when it comes to considering what determines the nation's money supply and credit. Loan activity by banks plays a fundamental role in determining the money supply. As economists know, when a loan is made between parties, money is created. The effects of bank-created money are well understood by economists, but the effects of non-bank-created money are not.
The banking system as an institution is the venue through which the Fed, as a central bank, manages the money supply in the US economy. The instrument to effectuate this management is the Fed funds rate, which sets the price of funds when banks borrow from one another. The repo market is a venue through which banks and other market participants can bypass the Fed's money-supply policy targets. Thus Fed funds are institutional opposites of repo proceeds, each serving opposite functions. Fed funds are used by the Fed to control the money supply created by bank lending, while the repo market is where banks and other market participants can seek to free themselves from the Fed's control. In practice, it is commonly assumed that the Fed sets the federal funds rate while in reality, the Fed sets a federal funds rate target and tries to move the federal funds rate toward the target by influencing the repo rate with open-market operations.
Money supplyConcepts of the money supply are meant to identify the quantity of the one commodity (money) that is decreed as legal tender for settling financial obligations within the economy as a whole.
Like other economic concepts, there is no perfect conceptual definition of money supply. Most economists have come to accept some vague intuitive understanding of the notion. Measuring the money supply, in practice, is mostly opinion-based, and a variety of money-supply measures are currently in use. The most frequently cited measures are M1 and M2. M1 includes demand deposits and coin and currency held by the non-bank public. M2 is M1 plus commercial deposits. M2 is the most widely followed money-supply measure. M3 is equal to M2 plus time deposits over $100,000 and term repo agreements. A fourth category, known a L, measures M3 plus all other liquid assets such as Treasury bills, savings bonds, commercial paper, bankers' acceptances and Eurodollar holdings of US residents (non-bank).
Banks hold cash reserves in proportion to their deposit liabilities. If a bank's cash reserves increase, then the bank may be able to increase its demand deposit liabilities correspondingly. The total cash reserves of the commercial banking system serve to constrain the total deposit liabilities of the system. The daily ebb and flow of commercial and private transactions is constantly shifting cash reserves from bank to bank. Total cash reserves are also mildly affected as coins and notes held by the public are altered through daily transactions. Broadly speaking, the daily transaction activity does not affect the total cash reserves held by the commercial banking system, merely the distribution of those reserves among the various individual commercial banks.
If cash reserves of a commercial bank begin to decline, it will eventually be deemed, legally or prudentially, inadequate for the outstanding demand deposit and time deposit liabilities. The bank must then either rebuild its cash reserve levels or cut back on its deposit liabilities. If it cannot stem the decline in cash reserves, then the bank ultimately must reduce deposit liabilities. Since these deposit liabilities are the most important component of the money supply, the money supply ultimately must decline if cash reserves of the banking system fall sufficiently. A sufficient decline in commercial-bank cash reserves will ultimately cause a drop in the money supply. Conversely, expansion of the money supply is ultimately limited by the aggregate cash reserves of the commercial banking system. At least that is how it worked theoretically before the spectacular growth of structured finance (derivatives), the impact of which on the virtual money supply is not well understood even by experts.
The Fed is frequently mentioned as having a major influence on the US economy mostly because it has the ability to change the cash reserves of the commercial banking system easily and in essence immediately, through its open-market operation in the repo market. Normally, open-market operations are transactions that involve the central bank in buying or selling (with money) government securities. The Fed effects open-market purchases by buying Treasury bills or notes or bonds on the open market and paying central-bank money. It really doesn't make any difference what the Fed buys so long as it pays money that will be deposited in the private banking system. The point is that the money goes from the Fed, which has an unlimited amount, to the private banking system, which has a limited amount. Whether the money is given freely, dropped from helicopters or in exchange for Treasury securities is irrelevant because fiat money itself has no intrinsic value - the value lies in what fiat money can buy. And that value is determined by the aggregate money supply in relation to the aggregate amount of assets.
Prices go up or down only when the relationship between the money supply and the amount of assets changes. When more money is issued along with a growth of assets, there will be no asset-price changes, or no inflation even if the economy expands. When asset prices rise, it reflects a change in the money supply/asset relationship, meaning more money chasing the same number of assets. Thus when asset prices rise, it is not necessarily a healthy sign for the economy. It reflects a troublesome condition in which additional money is not creating correspondingly more assets. It is a fundamental self-deception for economists to view asset-price appreciation as economic growth. A housing bubble is an example of this.
When prices fall, economists call it deflation. Deflation is not always caused by the economy not having enough money. It can be caused by an excessive liquidity preference on the part of market participants. When that happens, a liquidity trap develops in which market participants withhold money waiting for still lower prices. The Japanese economy in the 1990s was the clearest example of a liquidity trap, in which even negative interest rates failed to cure deflation.
Financial markets are obsessed with Fed activities. It is common to hear or read that the Fed has eased or tightened. Wall Street thinks of a lowering of the federal funds rate as Fed easing. Little or no attention is paid to money-supply aggregates or Fed balance-sheet aggregates by Wall Street commentators. The error of Wall Street's view is that the Fed does not directly participate in any way in the federal funds market, which is strictly an inter-bank market for cash reserves.
The federal funds rate is determined, during the course of a market day, by supply and demand by commercial banks - the Fed funds rate is a restricted market-determined rate. It is not set by the Fed or by anyone. The Fed merely sets a Fed funds rate target. The reason that Wall Street monitors the federal funds rate target as an indication of Fed policy is that the federal funds rate closely tracks the repo rate that the Fed actively influences during most market days. Every business-day morning at 11:45 Eastern Time, the New York Fed announces what it intends to do in the repo market. Changes in the repo rate are normally quickly followed by changes in the Fed funds rate. Thus, indirectly, the Fed appears to be able to influence the federal funds rate through its impact upon the repo rate.
Monetarists consider the money supply, and even more so the growth rate of the money supply, to be the main instrument of Fed policy. Thus if money-supply growth is increasing, then the Fed is said to be easing. If money-supply growth is decreasing, then the Fed is tightening. This concept of easing and tightening has the virtue that the Fed is known to control, with varying degrees of precision, the growth rate of the money supply. In other words, whatever the path of the money supply, there is no question that the Fed has the instruments at its disposal to control (with some error) the growth rate of the money supply.
But the rapid growth of structured finance has created questions on the validity of this view. Money now, especially virtual money, is created quite independently of Fed action, and money creation has become much less sensitive to interest-rate fluctuations. This explains why the measured pace at which the Fed has been raising the Fed funds rate target has little direct or immediate effect on the housing bubble.
Non-monetarists subscribe to the view that Fed easing means the Fed lowers interest rates. They are often not very specific about how these rates are lowered or how the Fed should go about doing this. There are often periods (eg 1990-91) when interest rates drop but money growth falls. Non-monetarists (and Wall Streeters) view periods like these as Fed easing episodes, while monetarists argue that these are (implicitly) periods of Fed tightening.
There is no generally accepted empirical test of Fed easing or tightening when money growth and interest rates are moving in the same direction, as appears to be happening now. The Federal Reserve is active on almost any given day in the repo market. Commercial banks can sell assets to raise cash (possibly liquidating loans); or they can borrow the cash. The repo market provides a way to do the latter while making it appear that they are doing the former.
Repos increase systemic riskIt is a good question why a loan is called a repo. The answer is that the notion of a repurchase agreement was a fiction dreamed up to minimize the impact of such transactions on bank and broker-dealer capital requirements. If these transactions had been called loans, then banks (and broker-dealers) would be required to set aside cash (or perhaps other capital if a broker-dealer) against such loans. By inventing the fiction of calling what is actually a loan by some other name, banks and broker-dealers were able to bypass banking regulation and reserve less cash/capital against such activities.
Convertible bonds are another type of loan that can be incurred by corporations off the books. Repos obviously increase systemic risk in the banking system as well as in the monetary system, particularly when the daily repos volume has grown to $5 trillion and is rising by the week.
To raise cash, a commercial bank normally sells so-called secondary reserves consisting of Treasury bills and other short-term debt assets, before resorting to other more drastic measures such as the liquidation of loans, selling fixed assets of the bank and so forth. Or a bank can borrow reserves directly from other commercial banks in the federal funds market. The rate for such borrowings would be the prevailing federal funds rate that follows the law of supply and demand set by Fed open-market operations.
Or a bank can do reverse repos. Mechanically, it would send out secondary reserves to some third party in exchange for a cash loan. When the loan terminates, the bank will receive back its secondary reserves that had been pledged for the repo. Such transactions will occur at the prevailing repo rate. A default will saddle the creditor with a loss equal to the spread between the old and new prevailing repo rate. Or a bank can issue commercial deposits (cd) to the public (which will simultaneously increase reserves and deposits). Such issuance (sales, really) will be transacted at the prevailing cd rates corrected for the credit quality of the issuing bank. Or a bank can borrow from the Fed at the Federal Reserve Discount Window. This type of borrowing takes place at the Fed discount rate, an announced rate that the Fed changes from time to time.
Because banks are looking at these two alternatives as ways of raising cash, there will be a tendency for rates to move together. If the Fed funds rate is high, but the repo rate is low, then banks will be more likely to raise cash by doing reverse repos than by borrowing funds. This will tend to move those rates together, similarly with cd rates. Thus if the Fed is able to raise or lower the repo rate, this should have an impact on the Fed funds rate in the same direction (It also should impact the cd rates through similar reasoning). The Fed is likely to be able to control the federal funds rate if it can control the repo rate. Most observers seem to feel that the Fed can routinely control the daily federal funds rate because the Fed, in their eyes, can control the repo rate.
Until recently, it was not obvious that doing repos and reverses had any real impact, other than very temporarily, upon either the cash reserves of the banking system or the repo rate. These conditions have fundamentally changed. The Bond Market Association reports that the average daily volume of total outstanding repurchase (repo) and reverse repo agreement contracts totaled $5.47 trillion in the first half of 2005, an increase of 17.4% from the $4.66 trillion from the daily average outstanding during the same period a year ago. The repo market has become the biggest runaway credit window outside of the control of the Fed. Yet the prevailing consensus view remains that Fed activity in the repo market should in essence be the mechanism for targeting the federal funds rate in the manner outlined above.
John Kambhu, vice president of the New York Federal Reserve Bank, observes that convergence trading, in which speculators trade on the expectation that asset prices will converge to their fundamental, or normal, levels, typically stabilizes markets. By countering and smoothing price shocks, these trades can enhance market liquidity. However, if convergence traders close out their positions prematurely, asset prices will tend to diverge further from their fundamental levels.
Both stabilizing and destabilizing forces in the market are attributable to convergence trading. The swap spread tends to converge to its fundamental level more slowly when the capital of traders has been weakened by trading losses, while higher trading risk can sometimes cause the spread to diverge from its fundamental level. Although convergence trading typically absorbs shocks, an unusually large disruption can be amplified when traders close out their positions prematurely. Destabilizing shocks in the swap spread are associated with a fall in repo volume consistent with the premature closing out of convergence trading positions. Repo volume also falls in response to convergence trading losses. Kambhu explains that taken together, these results are consistent with the argument that while convergence trading tends to be a stabilizing force, the risks in trading, as reflected in repo volume, on occasion can lead to behavior that destabilizes the swap spread. That occasion will occur as unpredictably but surely as a devastating hurricane hitting New Orleans.
And the rate at which the central bank lends money can indeed be chosen at will by the central bank; this is the rate that makes the financial headlines. In the US, the central-bank lending rate is known as the Fed funds rate. The Fed sets a target for the Fed funds rate, which its Open Market Committee tries to match by lending or borrowing in the money market. Thus fundamentally, the money market is not a free market, but one dictated by the central bank with a particular preference for the resultant state of the economy. The so-called free-market capitalism operates through this command money market. Thus at the heart of the free-market ideology is a fiat money system set by command of the central bank. The Fed is the head of the central-bank snake because the US dollar is the key reserve currency for international trade. The global money market is a US dollar market. All other currencies markets revolve around the US dollar market.
When a government's Treasury issues sovereign debt, the money proceeds go to finance the portion of the fiscal budget not covered by taxes. When government runs a fiscal deficit, it takes money from the private sector by issuing sovereign debt and spends the money back in the private sector. Thus the important issue is not if the government runs a fiscal deficit, but how the fiscal deficit is spent. A fiscal deficit does not reduce the total money supply; it only increases the amount of debt. But monetary economists such as Hyman Minsky assert that whenever credit is issued, money is created. Thus the issuing of government bonds is the government's way of issuing money without the involvement of the central bank. This is why Federal Reserve Board chairman Alan Greenspan is always warning about the fiscal deficit.
Yet the notion that government borrowing crowds out private borrowing is controversial. Fiscal deficits do not even directly affect short-term interest rates, which are set by the central bank. If the government wishes, it can take money directly from the central bank, which is legislatively authorized to issue money by fiat as the sole legal tender in the nation. A country's fiat money enjoys currency because the government accepts it for payment of taxes.
Fiat money is in fact a form of tax credit, or sovereign credit. Sovereign debt instruments do have a market function: they provide the assets that a central bank can buy or sell in the repo market to meet its Fed funds rate targets.
Henry C K Liu is chairman of a New York-based private investment group.