Systemic Risk and the Federal Reserve

If I hadn't worked at the Fed for 36 years, I wouldn't know its history either. Several misconceptions muddy the current debate about the future role of the Fed as systemic risk regulator. The following statements, frequently heard in some forms, are very misleading.

1. The Fed was created to conduct monetary policy; so let it stick to it's knitting.

2. The Fed doesn't need to take on a new role on top of those it already has.

3. The Fed had a large role in creating the current mess, so it shouldn't be rewarded with new assignments.

The Fed was not created to conduct monetary policy. Monetary policy as we know it today hadn't been invented yet. The Fed was created to stabilize the financial system and prevent or ameliorate financial panics.

Following the Panic of 1907, Congress established a national monetary commission to study the financial system and prevent episodes such as the panic. (Sound familiar?)

Much debate ensued, but eventually the study and the debate led to the Federal Reserve Act, which President Wilson signed on December 23, 1913.

One of the main problems that made the banking system unstable was that banks held their reserves with each other. Bank A deposited reserves (its assets) into Bank B (its liabilities). Bank B's reserves (assets) were deposited in Bank C (liabilities). And so on.

This arrangement worked well when it worked well. Problems arose when some banks, for whatever reasons, had to withdraw its reserve deposited in other banks. That caused those banks to have a reserve shortage and led them to withdraw reserve from still other banks, leading to a multiple contraction of reserves from the whole banking system. The scramble for reserves, if large enough, could and did create disruptions, even panics.

The framers of the Federal Reserve System substantially cured this problem by having banks deposit their reserves with a central entity, the Federal Reserve Banks, rather than with each other. That simple solution meant that a shortage of reserves in some banks didn't spread to the other parts of the banking system. Now you know how the Fed got its odd name: The Federal Reserve System.

Now, if the bank or banks with an initial shortage of reserves needed some temporary liquidity, they could borrow them from the Fed "on good collateral." Back then, most bank loans were called discounts because if someone borrowed $1000 they would get, say, $950 and repay $1000, the difference being interest. Instead of borrowing outright from the Fed, banks would usually discount their notes from customers. Since those notes were, themselves, called discounts, they were "rediscounting" them and their interest rate was called the "rediscount" rate. Over time, direct borrowing gradually replaced rediscounting customer paper and the "rediscount" rate became the "discount" rate.

The ability to discount credit worthy but illiquid paper at the Fed was designed to provide flexibility to the banks total reserve base. Bank borrowing from other banks just moved existing reserves around in the system without creating any new reserves to fill the shortage. Bank borrowing from the Fed's discount window left other banks undisturbed and actually increased the reserves of the borrowing bank as well as the system as a whole. This process added much stability to the banking system and prevented problems from spreading from bank to bank. It was not really monetary policy as we know it today. It was, if you will, the prevention or reduction of systemic risk.

Monetary policy emerged gradually in discount operations because the Fed could encourage banks to borrow from it rather than from other banks by lowering the discount rate. Shifting bank borrowing from other banks in the Fed funds market (the market for fund on deposit at the Fed, hence, Fed Funds) to borrowing at the Fed's discount window eased reserve pressure on the system as a whole because the latter involved a net expansion of reserves while the former was just a shifting of reserves.

On the other hand, the Fed could discourage expansionary borrowing from the discount window and encourage banks to rely more on the Fed funds market by raising its discount rate relative to the Fed funds rate. In the early years of the Fed, these discount rate changes were the only tool of monetary policy, but they were ancillary to the stability aspects of centralizing bank reserves. 

As indicated above, the Fed was not created to conduct monetary policy. It was created to provide stability to the banking and financial system through the centralization of reserves.

As I recall the story, open market operations, which became the principal tool of monetary policy, grew out of the recession of 1921. Bank discounting declined because loan demand had declined, and the Fed needed some earning assets to pay its bills. It bought some government securities, primarily for their income.

The Fed noticed the monetary implications of its purchases and sales of government securities. Purchases tended indirectly to expand bank reserves and stimulate bank lending and money creation. Sales or maturing securities had the opposite effect. So, informally at first, open market operations, or monetary policy, was invented in the 1920s. In the "reforms" of the 1930s, the formal trappings of the FOMC Committee, its membership, etc. were formalized by Congress.

Also in the 1930s, the Fed was given authority to change bank reserve requirements as a means of affecting bank reserves. As I recall, the reserve percentages that prevailed in the law, were made the lower end of a range and twice those levels were made the upper end of the range. Thus, reserve requirement changes were made the third tool of monetary policy.

Unfortunately, the Fed at the time made a mistake using its new reserve authority to "mop up" excess reserves in the banking system. As I've written here recently, the reserves were only "excess" in a legal or regulatory sense. Under the circumstances of the mid-1930s, the bankers didn't regard them as excess and contracted credit.

I mention this episode again because banks have excess reserves on their balance sheets currently and we don't want to make the same mistake again.

Comments (3)

Trackback URL | Comments RSS Feed

  1. Greg says:

    Thanks! I never knew how the Fed worked and had incorrectly assumed it was created specifically for monetary policy. Perhaps I should forward this to Ron Paul?

  2. GAry says:

    My wife is thinking of retiring and using the TRS for her medical since she works for a school district in Texas. If she retires after the health care act is passed will she be able to move from her school district plan to the state retirement medical plan or will she be forced to go to the government medical plan


  3. Mahesh says:

    Ramanan: I don’t hold views that the repos fund the Treasury purchase Sorry, I did not mean that. I tuhoght you meant that no repo are needed in the aftermath of the auction : no need for the repos to settle an auction . My fault entirely.Re. TT& program. There are three kinds of participants: collectors, retainers and investors. Collectors channel tax payments to the TGA; retainers retain payments on an interest bearing main account subject to caps (A about $80M; B about $250M;C about $8B, daily) *And* required collateral; investors, in addition to retaining, can accept Treasury transfers with one or the same date prior notification.There are tens of thousand of collectors, about a thousand of retainers, and about two hundred investors.Every day before 9 am EST, the Treasure and the Feds independently estimate the amount of tax receipts from collectors, treasury disbursements, proceeds from new bond sales, interest/principal payments, etc. At 9am, they compare there notes and decide on cash management actions.If the end-day estimated balance exceeds the $7B target, the treasury, with prior notification, invests with the investors provided the caps and collateral requirements are satisfied. Sometimes, the investor capacity is at the limit and more Fed repo action is needed to compensate.In the case of a predicted balance shortfall (<$5B), the treasury "calls" on the retainers or/and investors to transfer the required amount to the TGA.At 9:30 am, the fed OMO starts to conrtibute its cash management share, by repo'ing with the primary dealers, based on the earlier treasury/fed estimates and anticipated retainer/investor actions.After 2001, the treasury introduced dynamic account balancing intraday to shift excess cash, on an hourly basis, to the investor's interest bearing "main account" subject to caps and collateral as usual.That last facility is perhaps closest to you mental model of shifting cash between the treasury and the investor depositary institutions.The interest on treasury investment is determined though an action I mentioned earlier."The Bank of Canada"Don't know about BoC arrangements."The Treasury can decide what to do with the cash later. It can do a repurchase agreement with some financial market player etc."The treasury does not do repos. The Feds do.What the treasury does, I tried to describe above to the best of my recollection.