Monetary Policy: Part 3 - The Stable Decades (cont.)

The Federal Reserve spent the four decades after WWII targeting the level of interest rates, and in particular the overnight fed funds rate, with what could only be considered disastrous results. The 50's were marred by recession after recession, four in all. The 60's saw modest inflation result in government-imposed price controls followed by a resurgent inflation when they were dropped later. And the 70's saw a serious recession followed by the infamous "stagflation" when a stagnant economy was paired with near-double-digit inflation and long-term treasury rates bordering on 10%. The early 80's saw nearly 20% short rates, 14% long rates, and another serious double-dip recession. This track record hardly qualifies as a success for the Federal Reserve in that era.

But Stability Reigned in the Operating Rules

Yet, that was the time frame during which the Fed's operating rules remained relatively stable. It was also, not coincidentally, a time when a growing number of both monetarists and investors became interested in monetarism as a theory that explained what was happening to both prices and economic activity.

This was a time when so little changed in the system's banking rules that patterns began to appear, patterns from which deductions could be made and conclusions drawn. In fact, by 1980 a few prominent economists were beginning to home in on what was really going on with monetary policy. Then the introduction of NOW accounts blew it all out of the water.

Lessons to be Learned from the Stable Decades

At this point I'm switching to first person for a bit because this is a personal story of discovery. By 1975, as bond portfolio manager at the Tennessee Consolidated Retirement System in Nashville TN, I had become seriously interested in monetary theory. At that time the Fed of St. Louis branch of the Federal Reserve was the keeper of the Monetary Base, a measure of bank reserves, currency in circulation, and an adjustment factor that they calculated to account for the occasional (then quite minimal) changes in reserve requirements from time to time. They put the Monetary Base forward as a leading indicator of both money growth and the economy, a concept that I found interesting and useful.

Then, sometime in early 1976 I had a Eureka moment. Inflation by then had accelerated significantly and people were finding that cash didn't go as far as it used to go. Or, more accurately, the cash in one's wallet didn't last as long as it needed to before the next regular infusion. It suddenly dawned on me that the currency we all carry, or did then before debit cards at least, was an exceptionally stable amount and was increased only reluctantly, and in response to inflation that had already occurred. That was the insight, the Eureka! For at that time currency in circulation was a large component of the Monetary Base, but the Monetary Base was considered a leading indicator of money growth, economic growth, and inflation. But, clearly, people were reacting to inflation by increasing the money in their wallets, not the reverse. Virtually no one rationalized that inflation would be high over the next year so they'd better prepare for it by starting to carry more cash.

Assuming you buy that rationale, which I obviously did, the first question that arises is: "What is a lagging indicator of inflation (cash in your wallet) doing in a leading indictor of inflation (the monetary base)?" The obvious answer, to me at least, was that it doesn't belong there at all. So I took it out.

That left a much smaller number that consisted only of Total Reserves plus the Fed of St. Louis's adjustment factor, a number that I began to refer to as Effective Reserves. And that was a far more interesting number.

The Role of Effective Reserves and the Economy

In a future series of posts I might reanalyze the data from 1946 to 1980. I did it years ago and later discarded the original data. Today, that data is not readily available due to the many seasonal adjustments that have been applied over the succeeding years. I've looked at the current data and some of what I'll now describe can still be discerned, but it's not as convincing as the originally-reported data. I'll leave it at that.

What I found was that whereas the twists and turns in the M1 Money Supply (which was followed closely by most monetarists at that time) would indeed presage both economic turns and inflation, it was far less sensitive than Effective Reserves.

In fact, almost without fail for over 35 years any economic slowdown was presaged by a plateauing and sometimes a downturn of Effective Reserves that began almost exactly six months earlier. And if Effective Reserves turned back upward at some point (as it inevitably did) the economy would begin to recover almost exactly six months thereafter. There was a point in the late 1970's where you could count on a surprising drop in retail sales, for example, just by noting that it had been nearly six months since Effective Reserves topped out.

Effective Reserves and Inflation

What I found with inflation was equally interesting, and informative. If Effective Reserves and the GNP Deflator (its name at the time) were each indexed to 100 in the late 1940's the Deflator then tracked Effective Reserves with a lag for the next 40 years, with one notable exception. That was during the price control era of the early 1970's. President Nixon implemented price controls when inflation was running what later turned out to be a relatively modest 4% or so (as I recall) and inflation naturally slowed, or at least prices stopped rising.

However, Effective Reserves did not stop their steady rise and so the spread between the Effective Reserves line and the GNP Deflator line began to widen, with Effective Reserves outpacing the Deflator. This marked an incipient inflation, as it turned out. When price controls were eventually removed inflation picked up, eventually rising faster than the Effective Reserves line, for it had some "catching up" to do. Thus, ironically, the Carter inflation of the late 70's could actually have been a hangover effect from the Nixon price control era. In the end, the lines converged once again.

Volcker than took over the Fed and tightened viciously, driving interest rates on fed funds to nearly 20% because that was the theory the Fed operated under. Rates had to go up to get inflation under control.

But inflation was already under control according to the Effective Reserves measure. It had already plateaued and the Deflator had nearly caught up to it. According to the Effective Reserves measure, inflation was conquered and the economy was in trouble. The aftermath consisted of a serious double-dip recession and the commencement of a sustained drop in the inflation rate for almost 20 years resulting in the Fed of the early 2000's having to worry about possible deflation.

Question: Why Would Effective Reserves Track with the Deflator in a Growing Economy?

That is, why wouldn't a growing economy require a larger money supply to keep prices stable instead of falling. This question nagged me a lot during the 70's and early 80's and the best I could come up with, since the two measures tracked so reliably, is that the banking system has a productivity factor just like any other industry.

So, if economic growth required, say, 3% more money each year just to hold prices stable, and the banking industry got, say, 3% more efficient at pushing people's money around from place to place, perhaps it would even out. That's obviously speculation on my part, but it seems reasonable, especially since we had 40 years of relatively stable banking rules so any adaptations would be gradual, rather than sudden reactions to changing rules.

Conclusion

The evidence indicated to me that turns in Effective Reserves presaged both economic turns and changes in the rate of inflation. So if the Fed wants to keep inflation under control while not causing undue declines and accelerations in economic conditions, they should have simply targeted Effective Reserves rather than the fed funds rate.

By targeting, I mean that they should have sought to keep them stable, not rising or falling, either of which tends to influence economic growth. Instead, let the economy react to real factors, rather than fed-induced ones, and worry mainly about keeping inflation in check.

In the next post I'll describe what I learned about money growth in response to Fed actions, and how it is that I believe that maintaining a stable money supply was a trivial matter under the operating rules of the stable decades, although that was rarely achieved due to the theory the Fed was operating under at that time.

Continue to Monetary Policy: Part 4 - Solving the Mystery of Money Growth

Monetary Policy: Part 2 - The Stable Decades

This next few posts discuss the operation of monetary policy from WWII through to 1980 when NOW accounts were suddenly introduced. These were the stable decades in terms of the operating environment, although they were anything but stable in terms of the results. The instability of results had nothing to do with the operating system, but everything to do with the operators, for the operators simply didn't understand how the system actually functioned.

The Perceived Role of Interest Rates

For most of the Fed's existence up through 1980 the economic theory relied upon, and espoused, by the Federal Reserve held that the level of interest rates govern economic activity. This remains true to the present day, by the way.

Thus, if the economy was slowing, the Fed would cut interest rates, while if it was growing too fast and inflation was beginning to surge, it would raise interest rates. Actually it would cut or raise the interest rate, that rate being the rate of interest paid on federal funds in the interbank market. The theory essentially states that rising rates will slow economic activity whereas falling rates will encourage economic activity. Of course, this is true as the activity in the housing market often demonstrates. But it's the wrong theory upon which to base the operation of monetary policy. It's a reactionary policy, and therefore too slow, leading inevitably to large swings in economic activity and inflation.

The Essential Role of Interest Rates

There is only one interest rate that really matters to monetary policy, and that rate is zero percent, for that is the one and only rate that should be allowed on all transaction balances, or what we know as demand deposits, or checking accounts. It's also the rate paid on cash, incidentally. No one earns interest on the cash in their pockets, nor should they be permitted to earn interest on balances in their banks that they intend to be used for immediate purchases.

Why is this important? Why is paying interest on checking a terrible idea? Because it renders the operating mechanism of a fractional reserve system useless. And how does it do that? It makes holders of checking account balances indifferent to how much they hold in checking and how much they hold in savings.

When interest was only paid on savings deposits, people rationally minimized their checking account balances, even when interest on savings was relatively low. Some interest is better than no interest, after all. Similarly, they minimized the cash in their pockets. Importantly, this remained true regardless of the level of interest rates on savings. Again, some interest is better than none and a lot of interest is even better yet, but checking accounts and cash were always minimized by their holders. And as long as they were minimized at the outset, the level of interest rates on savings and other alternatives mattered little to monetary policy. The key to an operational fractional reserve system is the minimization of cash and demand balances by their holders.

Where the Fed Erred in the Stable Decades

Because both cash and checking balances earned no interest, the fractional reserve system in place since the Fed's founding could operate as planned. The problem is that the Fed didn't operate it as planned. Instead, because they followed the economic theory that put the cart of interest rates ahead of the monetary horse, and decided to manage interest rates instead of required reserves. But doing so was painfully slow-acting and often just plain wrong. Time after time, the Fed could be shown to have been tightening (raising rates) well after the economy had already gone into recession.

It generally takes four to six months before a recession is even suspected, much less acknowledged. Acknowledgement might take several quarters. Similarly, the Fed can be shown many times in the past to have been easing well after the economy had already been heating back up. The operating policy using interest rates as a lever of policy suffered from too much lag time to be truly effective. In fact, because of that lag time the Fed really never knew if it had tightened enough but would worry that it had and pause in the effort. This often prolonged the entire process until the Fed finally got ahead of things (or so they thought) and the economy would finally slow (with the Fed still in tightening mode several months after.)

What those operating the Fed at the time failed to realize is that they possessed all they needed within the reserve requirement mechanism itself. Interest rates needed no consideration whatsoever, other than the need to keep short term rates somewhat positive at a minimum, a trivial exercise given their control over the fed funds market.

The next post or two will seek to explain how monetary policy actually operated in the stable decades. Without the Fed's knowledge, it was exceptionally effective at producing the results obtained during that time. That the results were less than optimal is no fault of the operating system, for it performed exactly as programmed by the operators at the time.

Continue to Monetary Policy: Part 3 - The Stable Decades (cont.)

Monetary Policy Theory: Part 1 - Fractional Reserves

The next few posts will describe the operation of monetary policy prior to late 1980 when NOW accounts were first widely authorized. NOW accounts enabled banks to begin paying interest on demand deposits, that is, on what were essentially just checking accounts. Prior to the authorization of NOW accounts, checking accounts could not draw interest. NOW accounts were a concession to the rapid growth in money market accounts during the high-interest era of the late 1970's.

How a Fractional Reserve System Works

Under a system of fractional reserves each bank is required to hold a certain percentage of its deposit base in the form of reserves at the Federal Reserve System. For example, if the only form of deposits were checking accounts (demand deposits) and the reserve requirement was 10%, then a bank with demand deposits of $1 million would have to deposit $100,000 at the Fed in a "Required Reserve" balance.

Now assume that only demand deposits are permitted and that the reserve requirement is 10%. If total demand deposits in the system are $10 billion, then the required reserves at the Fed would have to be $1 billion. But the Fed can change the level of reserves in the system at will. All it has to do to add, say, $10 million in reserves, is purchase that dollar amount of securities from the banking system. The bank sends in the securities to the Fed and receives a $10 million increase in its reserve balance at the Fed in return. Because those reserves are not currently needed to be held against deposits they are considered "excess reserves."

And, since the bank doesn't need that additional $10 million as required reserves, it lends out the money either to a customer or by lending it in the fed funds market. As the money circulates from bank to bank, and from customer to customer, and as the economy grows, eventually the total demand deposits in the banking system will rise by $100 million and the $10 million in excess reserves would be converted to required reserves. Thus, by adding reserves to the banking system, the Fed will have allowed the overall banking system (and perhaps the economy) to grow. This is known as a Fed easing.

Effecting a Tightening

To reverse the process, perhaps to cool off an overheating economy or to suppress a rising inflation rate, the Fed withdraws required reserves from the banking system. It does this by selling securities from its portfolio to a bank in the system. Say the Fed wants to reverse the above action. It then sells $10 million of it securities to a bank in the system. The bank takes in the securities and to pay for them has its balance of required reserves at the Fed reduced by $10 million.

Now the banking system has a deficit of required reserves. If the Fed refuses to accommodate that deficit, interest rates will tend to rise and banks will make less loans because they are scrambling for reserves. Ultimately, the deposits in the banking system will diminish to the point where the lower level of reserves is sufficient to meet the reserve requirement.

In theory, this is how a fractional reserve system is supposed to work. That is, to keep the money supply, and consequently the price level of goods and services, stable, the Fed simply provides a constant level of reserves.

Complications

The first complication is that all deposits aren't demand deposits. In fact, savings deposits used to be commonly called time deposits to distinguish them from demand deposits. Each holder of a savings account received clear notification when opening an account that the money couldn't necessarily be withdrawn on less than a few day's notice. Demand deposits, in contrast, could be withdrawn (demanded) immediately.

Because time deposits were judged to be a more stable component of the deposit base of any bank, the Fed set a lower reserve requirement on them. In our present example, let's say that's only 2%. When interest rates were low as was the case for most of the Fed's existence up until the late 1970's, savings balances did remain quite stable, growing along with the economy and not reacting to the level of interest rates.

Once interest rates started rising significantly in the 1970's people started looking for ways to maximize the income on their checking balances. Mutual funds began offering Money Market Funds to people in response. In the beginning, they only permitted the withdrawal of relatively large sums at one time, say $500 or $1,000, so that people wouldn't use them in place of their traditional checking accounts.

However, in time the banks managed to convince the regulators to let them compete with money market funds. That led to the introduction of NOW accounts which were essentially interest-bearing checking accounts with some modest restrictions. Once they came onto the scene most people converted their checking accounts to NOW accounts and many blended both their checking and savings accounts into one NOW account instead.

The Problem Posed by Differing Reserve Requirements

Because NOW accounts were really a blend of demand deposits and savings (time) deposits, they more or less destroyed the fractional reserve system in place until that time because of the differing reserve requirements on demand deposits and savings deposits. Suddenly savings deposits weren't stable anymore. Instead they were mixed in with interest-bearing checking accounts. Furthermore, as will be discussed later in more detail, checking accounts (demand deposits) were no longer minimized by their holders since they now earned interest. That, as will be covered in the next post, had major negative implications for monetary policy as it had been conducted up to that time.

The next post will discuss the period between WWII and the introduction of NOW accounts and will describe how the Fed once had a monetary tool that worked exceptionally well, but let it slip out of their grasp with the introduction of NOW accounts because they never actually understood how it worked. In fact, if they had understood, NOW accounts would never have been authorized and money market funds would have had their usage in daily transactions restricted.

Continue to Monetary Policy: Part 2 - The Stable Decades