A popular view is that the Fed has given up on monetary tightening and as a result the stock market should continue to trend upward over the months ahead. This view is based on flawed reasoning.

The reality is that the Fed possibly will give up on monetary tightening later this year, but currently the Fed is pulling quite firmly on the monetary reins via its on-going balance-sheet normalisation (that is, balance-sheet reduction) program. Moreover, the Fed’s on-going withdrawal of money from the economy is not being fully offset by the actions of the commercial banks, so the overall US money-supply situation is becoming increasingly restrictive. This is evidenced by the following chart of the year-over-year (YOY) change in US True Money Supply (TMS). The chart shows that in March-2019 the US monetary inflation rate made a 12-year low.

However, the unusually slow pace of US money-supply growth is not a good reason to be short-term bearish on the US stock market. This is partly because changes in the financial markets lag changes in the monetary backdrop by long and variable amounts of time. It is also because of a point that was covered in a TSI blog post about three weeks ago.

The point I’m referring to is that whether the overall monetary situation is ‘tightening’ or ‘loosening’ is not solely determined by the change in money supply. Instead, over periods of up to a few years the change in the demand for money (meaning: the change in the desire to hold/obtain cash as an asset) often will dominate the change in the supply of money.

In general terms, the change in overall liquidity is determined by the change in the supply of money relative to the change in the demand to hold cash or cash-like securities. As a consequence, it’s possible for the liquidity situation to be tight even if the monetary inflation rate is very high and/or rapidly increasing. A great example is the period from September-2008 to March-2009, when a large and fast increase in the US money supply was more than offset by a surge in the demand for money. Also, it’s possible for there to be abundant liquidity even if the monetary inflation rate is very low. A good example occurred over the past 3-4 months.

Although the supply side of the monetary equation tends to be dominated by the demand side of the equation over the short-to-intermediate-term, today’s unusually low monetary inflation rate is still significant. It means that only a small increase in the demand to hold cash could bring about another plunge in the stock market. To put it another way, due to the low monetary inflation rate the US stock market is far more vulnerable than usual to a short-term increase in risk aversion.

Taking a wider-angle view, the money-supply situation also leads to the conclusion that if a bear market did not begin last year (it most likely didn’t) then it will begin by the second half of next year at the latest. Other indicators will be required to narrow-down the timing.

[In a blog post last October I mentioned that a recent divergence between the gold/commodity ratio and the T-Bond price had bullish implications for the T-Bond. A strong rebound in the T-Bond soon got underway. Another divergence between the gold/commodity ratio and the T-Bond price has since developed, this time with bearish implications for the T-Bond. A discussion of the most recent divergence was included in a TSI commentary published on 28th March and is reprinted below.]

The gold/commodity (g/c) ratio and the T-Bond price tend to move in the same direction. As previously explained, this tendency is associated with what Keynesian economists call a paradox (“Gibson’s Paradox”) and Austrian economists call a natural and perfectly understandable consequence of the relationship between time preference and prices. The reason for revisiting the gold-bond relationship today is that a significant divergence developed over the past three months and such divergences are usually important.

The following chart illustrates our point that the gold/commodity ratio (the US$ gold price divided by the GSCI Spot Commodity Index) and the T-Bond price move in the same direction most of the time. It also shows that over the past three months the two quantities have diverged, with the g/c ratio trending downward while the T-Bond price extended its upward trend and moved to a marginal new 12-month high.

Given that the relationship between the g/c ratio and the T-Bond has a solid fundamental basis, that is, given that it’s not a case of random correlation, it should continue to apply. Therefore, we expect that the divergence will close over the months ahead — via either a rise in the g/c ratio to above its December-2018 high or a decline in the T-Bond price to well below its February-2019 low.

The divergence probably will close via a decline in the T-Bond price, because if there is a leader in this relationship it is the g/c ratio. For example, in each of the three biggest divergences of the past five years (the areas inside the blue boxes drawn on the above chart), the g/c ratio reversed course months in advance of the T-Bond. The g/c ratio also led the T-Bond by 2-3 months at the Q3-2017 top and by a couple of weeks at the Q4-2018 bottom. In other words, the recent performance of the g/c ratio is a reason to be intermediate-term bearish on the T-Bond.

One realistic possibility is that the T-Bond is now topping similarly to how it bottomed between December-2016 and March-2017. Back then, both the g/c ratio and the T-Bond turned up at around the same time (in late December of 2016), but whereas the g/c ratio trended upward throughout the first quarter of 2017 the T-Bond made a marginal new low in March before commencing an upward trend of its own. This time around the g/c ratio and the T-Bond turned down at around the same time (in late December of 2018), but whereas the g/c ratio has continued along a downward path the T-Bond has risen to a marginal new multi-month high.

[This blog post is an excerpt from a recent TSI commentary]

The Quantity Theory of Money (QTM) holds that the change in money Purchasing Power (PP) is proportional to the change in the Money Supply (MS). It’s a bad theory, because it doesn’t reflect reality.

There are three main reasons that QTM doesn’t work in the real world, the first being that money PP can’t be expressed as a single number. There is no such thing as the “general price level”. Instead, at any point in time there are millions of individual prices that cannot be averaged to arrive at something sensible. That being said, QTM wouldn’t work even if it were possible to determine the “general price level”.

The second reason that QTM doesn’t work in the real world is that new money never gets injected uniformly throughout the economy. A consequence is that different prices get affected in different ways at different times, depending on who the first receivers of the new money happen to be. For example, during normal times the commercial banks are responsible for almost all money creation, with new money entering the economy via loans to the banks’ customers, whereas during 2008-2014 most new US dollars were created by the Fed and injected into the financial markets via the purchasing of bonds.

However, even if there existed a single number that accurately represented money PP and new money was injected uniformly throughout the economy, the Quantity Theory of Money STILL wouldn’t work. The reason is that as is the case with the price of anything, the price of money is determined by supply AND demand. (As an aside, in the real world there is no such thing as money velocity.) In other words, the price (PP) of money never could be properly explained/understood by reference to only the supply of money. We’ll now expand on this point.

Over the very long term, changes in money supply dominate changes in money demand, where by money demand we mean the desire to hold cash as an asset rather than the desire to obtain money to facilitate current purchases. However, during periods of up to a few years the change in money demand often will dominate the change in money supply. A good example is September 2008 through to March 2009. During this period the Fed rapidly increased the money supply, but the Fed’s actions were overwhelmed by increasing demand for money. Furthermore, when prices suddenly started rising in March-April of 2009 it was not only because the money supply had grown, but also because the demand for money had begun to fall.

In relation to the above it’s important to understand that in addition to affecting the supply of money, the Fed and other central banks affect the demand for money. This is very relevant to the recent past. Over the past three months the Fed continued to reduce the money supply, but statements emanating from the Fed had the effect of reducing the desire to hold cash. The net effect was a general increase in ‘liquidity’ even while the Fed acted to reduce the money supply.

Unfortunately, there is no way to analyse the monetary situation that is both simple and accurate. In particular, there is no simple equation that indicates the real-world relationship between money supply and money purchasing-power.

I haven’t discussed gold’s true fundamentals* at the TSI Blog since early December of last year, at which time I concluded: “All things considered, for the first time in many months the true fundamentals appear to be slightly in gold’s favour. If the recent trend in the fundamental situation continues then we should see the gold price return to the $1300s early next year…” The “recent trend in the fundamental situation” did continue, enabling my Gold True Fundamentals Model (GTFM) to turn bullish at the beginning of this year (after spending almost all of 2018 in bearish territory) and paving the way for the US$ gold price to move up to the $1300s.

The following weekly chart shows that after moving slightly into the bullish zone (above 50) at the beginning of January, the GTFM has flat-lined (the GTFM is the blue line on the chart, the US$ gold price is the red line). Based on the current positions of the Model’s seven inputs, its next move is more likely to be further into bullish territory than a drop back into bearish territory.

As an aside, the bullish fundamental backdrop does not preclude some additional corrective activity in the near future.

The most important GTFM input that is yet to turn bullish is the yield curve, as indicated by the 10year-2year yield spread or the 10year-3month yield spread. This input shifting from bearish to bullish requires a reversal in the yield curve from flattening (long-term rates falling relative to short-term rates) to steepening (long-term rates rising relative to short-term rates). If the reversal is driven primarily by falling short-term interest rates it indicates a boom-to-bust transition, such as occurred in 2000 and 2007, whereas if the reversal is driven primarily by rising long-term interest rates it points to increasing inflation expectations.

As illustrated below, at the end of last week there was no evidence of such a trend change in the 10year-3month yield spread.

The US$ gold price could rise to the $1400s during the second quarter of this year as part of an intermediate-term rally, but to get a gold bull market there probably will have to be a sustained trend reversal in the yield curve.

*I use the term “true fundamentals” to distinguish the actual fundamental drivers of the gold price from the drivers that are regularly cited by gold-market analysts and commentators. According to many pontificators on the gold market, gold’s fundamentals include the volume of metal flowing into the inventories of gold ETFs, China’s gold imports, the volume of gold being transferred out of the Shanghai Futures Exchange inventory, the amount of “registered” gold at the COMEX, India’s monsoon and wedding seasons, jewellery demand, the amount of gold being bought/sold by various central banks, changes in mine production and scrap supply, and wild guesses regarding JP Morgan’s exposure to gold. These aren’t true fundamental price drivers. At best, they are distractions.

[This blog post is a modified excerpt from a TSI commentary published about a month ago]

Modern Monetary Theory, or MMT for short, is gaining popularity in the US. It is based on the idea that under the current monetary system the government doesn’t have to borrow. Instead, it simply can print all the money it needs to fill the gap between its spending and its income. The only limitation is “inflation”. As long as “inflation” is not a problem the government can spend — using newly-created money to finance any deficit — as much as required to ensure that almost everyone is gainfully employed and to provide all desired services and infrastructure. It sounds great! Why hasn’t anyone come up with such an effective and easy-to-implement prosperity scheme in the past?

Of course it has been tried in the past. It has been tried countless times over literally thousands of years. The fact is that there is nothing modern about Modern Monetary Theory. It is just another version of the same old attempt to get something for nothing.

Most recently, MMT was put into effect in Venezuela. For all intents and purposes, the government of Venezuela printed whatever money it needed to pay for the extensive ‘free’ social services it promised to the country’s citizens. The MMT apologists undoubtedly would argue that the money-printing experiment didn’t work in Venezuela because the government didn’t pay attention to the “inflation” rate. It kept on printing money at a rapid pace after “inflation” became a problem. Our retort would be: “Great point! Who would have thought that a government with the power to print money couldn’t be trusted to stop printing as soon as an index of prices moved above an arbitrary level.”

In essence, MMT is based on the fiction that the government can facilitate an increase in overall economic well-being by exchanging nothing (money created ‘out of thin air’) for something, or by enabling the recipients of the government’s largesse to exchange nothing for something. It is total nonsense, although there is an obvious reason that it appeals to certain politicians. Its appeal to the political class is that it superficially provides an easy answer to the question that arises when politicians promise widespread access to valuable services free of charge. The question is: “Who will pay?” According to MMT, nobody pays until/unless “inflation” gets too high.

And what happens when inflation gets too high? Well, according to MMT the government simply ramps up direct taxation to reduce the spending power of the private sector, which supposedly quells the upward pressure on prices.

Therefore, MMT can be viewed as a case of heads the government wins, tails the private sector loses. As long as “inflation” is below an arbitrary level the government can extract whatever wealth it wants from the private sector indirectly by printing money, and if “inflation” gets too high the government can extract whatever wealth it wants from the private sector via direct taxation.

The crux of the issue is that new wealth can’t be created by printing money, but existing wealth will be redistributed. It’s like when a private counterfeiter prints new money for himself. When he spends that money he diverts real wealth to himself while contributing nothing to the economy. MMT is the same principle applied on a gigantic scale.

That being said, MMT does have its good points, just not the good points that its proponents claim.

As happens when money is loaned into existence under the current system, the application of MMT will affect relative prices as well as the so-called “general price level”. The reason is that the new money won’t be injected uniformly across the economy. However, it’s likely that the price increases stemming from the monetary inflation will be more uniform and direct under MMT than under the current system. In other words, under MMT the effects of monetary inflation should be reflected much sooner and to a far greater extent in the CPI than is the case with the current system.

That the application of MMT would lead quickly to what most people think of as “inflation” is a benefit, because the link between cause (monetary inflation) and effect (rising prices) would be obvious to almost everyone. A related benefit is that MMT would short-circuit the boom-bust cycle.

Booms happen when the Fractional Reserve Banking (FRB) system (with or without a central bank) expands credit and in doing so creates the impression that the quantity of real savings is much greater than is actually so. This prompts excessive investment in long-term business ventures that would not look viable in the absence of misleading interest-rate signals.

We assume that under MMT the commercial banks still would be lending new money into existence, but the temporary downward pressure on interest rates from the surreptitious money creation of the banks would be more than offset by the upward pressure on interest rates from the blatant money-printing of the government. The boom phase therefore would be very short, perhaps even barely noticeable. In effect, MMT would bypass the boom and go straight to the bust. Again, this would be beneficial because it would expose the link between cause (the application of a crackpot monetary theory) and effect (economic hardship for most people).

MMT is such an obviously silly idea that any economist, politician, journalist or financial-market commentator who advocates it should not be taken seriously. However, that they are being taken seriously opens up the possibility that MMT will be implemented in the not-too-distant future, with the ‘benefits’ outlined above.

It’s likely that by the middle of the next decade, most new cars, trucks and buses will be Electric Vehicles (EVs). As a consequence, it’s a good bet that over the next several years the demand for both gasoline and diesel will shrink dramatically while the demand for electricity (to recharge EV batteries) experiences huge growth. Part of this demand growth will be satisfied by nuclear power, which is why uranium is an indirect play on the EV trend.

The uranium price might have begun to discount the aforementioned shift in demand in that it has been quietly trending upward since around April of last year (a weekly chart is displayed below). I say “quietly” because the rally has been accompanied by very little in the way of speculative enthusiasm. On the contrary, the rally that began last April has been accompanied by widespread scepticism.

This is an important sentiment change. Whereas every multi-month up-move in the uranium price prior to last year was greeted as if it heralded the beginning of a bull market, almost everyone has dismissed the most recent rally as just another counter-trend bounce. Being bearish on uranium has become easy, but bull markets begin when it’s easy to be bearish.

I’m not convinced that a uranium bull market is underway, but I do think that for the uranium-mining sector the intermediate-term risk/reward is skewed decisively towards reward. The reason is that the mining stocks could achieve large price gains with or without a genuine bull market in the underlying commodity. All it would take, I think, is a move by the uranium price into the $30s to convince many speculators that a major trend reversal had occurred and prompt aggressive buying of uranium-mining equities.

It used to be that owning shares of the Global X Uranium Fund (URA) was the simplest and surest way of participating in a uranium-mining rally, but that is no longer the case due to the changes that were made to this fund last year. As outlined HERE, during the second quarter of last year the index that URA tracks was changed from the Solactive Global Uranium Total Return Index to the Solactive Global Uranium & Nuclear Components Total Return Index. Thanks to this change, seven of URA’s current top ten holdings have no correlation with the uranium price.

Nowadays, owning the shares of Cameco (CCJ) is the surest way of participating in a uranium-mining rally.

It looks to me like CCJ is not far from completing a 3-year base (see chart below). I like the idea of gradually building up a position on weakness while the basing process continues.

[This blog post is an excerpt from a TSI commentary published on 24th February]

Based on the par value of maturing securities on its balance sheet there was scope for the Fed to withdraw as much as $43B from the financial markets on 15th February. A week ago we noted that this ‘liquidity drain’ had no effect on the stock market, possibly because the effect would occur on the next trading day (Tuesday 19th February) or because the Fed chose to withdraw a lot less money than it could have. Now that the Fed has issued its latest weekly balance sheet update we can see why there was no effect from this potential bout of “quantitative tightening” (QT). We can also see that the general perception of what the Fed is doing has deviated in a big way from what the Fed actually is doing.

During the 7-day period from 13th to 20th February the Fed’s securities portfolio fell by $31B. In other words, the Fed implemented $31B of a potential maximum $43B of QT. Over the same period, however, the amount of money in the US federal government’s account at the Fed fell by $44B. This means that there was a $31B withdrawal of liquidity by the Fed in parallel with a $44B injection of liquidity by the government, resulting in a net liquidity ADDITION of $13B. No wonder there wasn’t a noticeable negative effect on the stock market from the Fed’s actions.

The difference between a Fed liquidity injection and a government liquidity injection is that whereas the Fed can inject new money, the government can only recycle existing money (the government returns to the economy the money it previously removed via borrowing or taxation). Government liquidity injections therefore are not inflationary, but their short-term effects can be similar to Fed liquidity injections.

Note that at 20th February the government had about $330B in its account at the Fed. This means that the government currently has the ability to inject up to $330B into the economy, but depending on the size of its desired cash float it may or may not make additional injections in the short-term.

Also note that notwithstanding all of the ‘dovish’ talk that has emanated from the Fed over the past two months, the QT program has continued. From 2nd January to 20th February the Fed removed $63B from the economy as part of its “balance sheet normalisation”. The pace of the liquidity removal is slower than the Fed’s self-imposed $50B/month limit, but it is not correct to say — as some pundits have said — that the Fed has stopped tightening. The Fed is still pulling on the monetary reins.

That the Fed is still tightening means that there is a substantial mismatch at the moment between perception and reality. The general perception is that the Fed is now either on hold or preparing to loosen, but, as mentioned above, this most definitely is not the case. Consequently, bullish speculators in the stock, commodity and gold markets are getting ahead of themselves.

It’s possible that the Fed will end its monetary tightening within the next few months, but that’s only going to happen if there’s another pronounced shift away from risk. To put it another way, the more the markets discount an easing of monetary policy the less chance the easing will occur. In fact, if the stock market extends its upward trend into May-June then the Fed probably will resume its rate-hiking in June.

Like every other financial market in world history, the gold market is manipulated. However, anyone who believes that manipulation of the gold market is an important influence on major gold-price trends does not understand the true fundamental drivers of the gold price.

To paraphrase Jim Grant, gold’s market value is the reciprocal of confidence — in the banking system (including the central bank), the economy and the government. In other words, gold should do relatively well when confidence is on the decline and relatively poorly when confidence is on the rise.

By comparing the gold/commodity ratio with measures of monetary and/or economic confidence it can be shown that gold generally does exactly what it should do. There are periods of divergence, but these tend to be short (no more than a few months) and barely noticeable on long-term charts.

The point outlined above can be illustrated by comparing the gold/commodity (gold/GNX) ratio with the IEF/HYG ratio, which I’ve done in the following chart.

The IEF/HYG ratio is fit for our purpose because it is a measure of what’s happening to credit spreads, and because the economy-wide credit-spread trend is one of the best indicators of economic confidence. Specifically, the IEF/HYG ratio increases when credit spreads are widening (indicating declining economic confidence) and decreases when credit spreads are narrowing (indicating rising economic confidence).

Therefore, it is fair to say that the following chart compares the gold/commodity ratio with the reciprocal of confidence in the US economy.

Lo and behold, the two lines on the chart track each other quite closely.

[This blog post is an excerpt from a commentary posted at TSI about three weeks ago]

TARGET2 is the system set up in the euro-zone to clear inter-bank payments. The Bundesbank (Germany’s central bank) describes it as a payment system that enables the speedy and final settlement of national and cross-border payments. The problem is that often there is no “final settlement” under TARGET2. Instead, credits and debits can build up indefinitely.

To understand the issue it first must be understood that although the 19 countries that comprise the euro-zone use a common currency, the euro-zone isn’t really a unified monetary system. It is more like 19 separate monetary systems, each of which is overseen by a National Central Bank (NCB). These NCBs are, in turn, overseen and coordinated by the ECB. TARGET2 is the means by which money is transferred quickly and efficiently between these 19 separate monetary systems. The transfer may well be quick and efficient, but, as noted above, it often doesn’t result in final settlement.

Further explanation is provided by the Bundesbank, as follows:

“…both the Bundesbank and the Banque de France will be involved in a cross-border payment transaction made in settlement of a German export to France, for instance. That transaction begins when the French importer’s commercial bank in France debits the purchase amount from the importer’s account and submits a credit transfer in TARGET2 to the German exporter’s commercial bank in Germany. The Banque de France then debits the amount from the TARGET2 account it operates for the French commercial bank and posts a liability owed to the Bundesbank. For its part, the Bundesbank posts a claim on the Banque de France and credits the amount to the German commercial bank’s TARGET2 account. The transaction is concluded when the commercial bank credits the amount in question to the account it operates for the German exporter.

At the end of the business day, all the intraday bilateral liabilities and claims are automatically cleared as part of a multilateral netting procedure and transferred to the ECB via novation, leaving a single NCB liability to, or claim on, the ECB.

Viewed in isolation, the transaction used as an example above leaves the Banque de France with a liability to the ECB and the Bundesbank with a claim on the ECB at the end of the business day. These claims on, or liabilities to, the ECB are generally referred to as TARGET2 balances.”

The example given above by the Bundesbank refers to a German export to France, but the same process would apply when someone transfers money from a bank deposit in one EZ country to a bank deposit in another EZ country. For example, the electronic wiring of funds from a commercial bank account in Italy to a commercial bank account in Luxembourg would leave the Banca d’Italia with a liability to the ECB and the Banque Centrale du Luxembourg with a claim on the ECB.

The process described above means that there is never any net clearing of cross border payments at the NCB level. Unless the money flowing in one direction (into Country X) equals the money flowing in the opposite direction (out of Country X), credit/debit balances will build up and there is no limit to how large these balances can become.

As illustrated by the following chart from Yardeni.com, this is not just a hypothetical issue. The NCBs of some EZ countries, most notably Germany and Luxembourg, now have huge positive TARGET2 balances, and the NCBs of some other EZ countries, most notably Italy and Spain, now have huge negative TARGET2 balances.

As at October-2018, the central bank of Germany was owed 928 billion euros by the TARGET2 system, while together the central banks of Italy and Spain owed 887 billion euros to the TARGET2 system. Is this a problem?

The system is so strange that there doesn’t appear to be a clear-cut answer to the above question, at least not one that we can fathom. It could be a huge problem or it could be no problem at all.

The Bundesbank is sitting there with an asset valued at almost 1 trillion euros that will never pay any interest and cannot be collected. At first blush this appears to be a huge problem. It implies that at some point the asset will have to be written off, perhaps leading to a very expensive bailout funded by German taxpayers. But then again, due to the way the current monetary system works it may well be possible for TARGET2 balances to grow indefinitely with no adverse consequences. That’s why we haven’t devoted any commentary space to this issue in the past.

If we were forced to give an answer to the above question it would be that rising interest rates, burgeoning government debt levels and private bank failures will become system-threatening issues in the EZ long before the TARGET2 balances pose a major threat.

Bank reserves are a throwback to a time when the amount of receipts for money (gold) that could be issued by a bank was limited by the amount of money (gold) the bank held in reserve. Under the current monetary system bank reserves have no real meaning, since it isn’t possible for a dollar in a bank deposit to be genuinely backed by a dollar held somewhere else. The dollar can’t back itself! However, it is still important to understand what today’s bank reserves are/aren’t and how changes in the reserves quantity are linked to changes in the economy-wide money supply. Remarkably, these bank-reserve basics are misunderstood by almost everyone who comments on the topic.

The simplest way for me to deal with the common misunderstandings about bank reserves is in point form, so that’s how I’ll do it. Here goes:

1) Bank reserves aren’t money, that is, they are not considered to be general media of exchange and are not counted in the True Money Supply (TMS). Instead, they provide ‘backing’ for part of the money supply.

2) A corollary of the above is that banks can’t use their reserves to buy things outside the Federal Reserve system.

3) Banks can lend their reserves to other banks, but the banking industry as a whole cannot expand or shrink its reserves. In other words, the banking industry has no control over its collective reserves. The central bank has total control.

4) Bank reserves can be shifted around within accounts at the Fed, but the only way that reserves can leave the Fed and enter the economy is via the withdrawal, by the public, of physical currency from banks. For example, when $100 is withdrawn from an ATM, $100 is converted from deposit currency to physical currency. This doesn’t alter the money supply, but it causes the bank to lose a $100 liability (the bank customer’s deposit) and a $100 asset (the physical currency held in the bank’s vault). When the quantity of physical currency held in a bank’s vault gets too small, the bank will replenish its supply by withdrawing reserves from the Fed in the form of new paper dollars. Although it may appear that this imposes some sort of limit on the supply of physical dollars, the Fed stands ready, willing and able to meet any increase in demand. This is further discussed in point 5).

5) Under the current monetary system, reserves effectively are created out of nothing. To be more precise, the Fed creates reserves when it purchases bonds and other assets. Since there is no limit to the dollar value of assets that can be purchased by the Fed, the banking system will never run short of the reserves it needs to meet the public’s demand for physical currency. Also, the Fed can remove reserves whenever it wants by selling bonds and other assets.

6) Except for the siphoning of reserves in response to the public’s increasing demand for physical currency, it is accurate to say that reserves at the Fed stay at the Fed until they are removed by the Fed. A corollary — as already mentioned in point 3) — is that the commercial banking industry cannot draw-down its reserves.

7) The Fed pays interest on ALL reserves, not just so-called “excess reserves”. In any case and as outlined below, for all intents and purposes all US bank reserves, with the exception of the relatively small portion required to meet any increase in the demand for physical currency, are now excess and have been for the past few decades.

8) The way the US monetary system now works it is fair to say that all reserves are excess. The reason is that the quantity of bank reserves has no bearing on the amount by which banks expand/contract credit. In effect, the US now has a zero-reserve fractional reserve banking system. That’s why it was possible for the greatest expansion of bank credit in modern US history, which took place during 1990-2007, to happen while the commercial banking industry had almost no reserves. During this period total bank credit rose by $6 trillion, from $2.5T to $8.5T, while bank reserves at the Fed dwindled from $64B to $40B.

9) Further to point 8), bank lending doesn’t ‘piggy-back’ on bank reserves. It possibly did 40 years ago, but it hasn’t for at least the past 25 years. Hopefully, economics textbooks eventually will be updated to reflect this reality.

10) An implication of points 7) and 8) is that interest payments on reserves are neither an incentive nor a disincentive to bank lending. When a bank makes a loan to a customer it doesn’t lose any reserves and therefore continues to collect the same interest-on-reserves payment from the Fed.

11) The sole purpose of paying interest on reserves is to enable the Fed to hike the Fed Funds Rate during a period when the banks are inundated with reserves, without having to massively reduce the quantity of reserves. This was discussed in previous blog posts, for example HERE.

12) When the Fed was ‘quantitatively easing’ many pundits wrote that it was adding to bank reserves but not the money supply. This is wrong. When the Fed buys X$ of securities as part of a QE program it adds X$ to bank reserves AND it adds X$ to the economy-wide money supply. I previously described the process HERE.

13) By the same token, now that the Fed is ‘quantitatively tightening’ it is not just removing bank reserves. When the Fed sells X$ of securities as part of what it refers to as its balance-sheet normalisation program it removes X$ from bank reserves AND it removes X$ from the economy-wide money supply. In essence, it’s the process I described in the above-linked post (point 12) in reverse. That’s why the balance-sheet normalisation program is vastly more important, as far as monetary conditions are concerned, than the rate-hiking program.