In this blog post, I want to explain really slowly that the quantitative easing is just a bond buyback program, it just reorganizes the debt, (more or less randomly) creates some losers and some winners, but to the leading order, it doesn't change the overall debt, inflation, or economic activity. Some important slogans that summarize the text below are
coins and banknotes are just another form of debtand
the government is just another type of a natural catastrophe.The markets work, they do adjust, and their ability to produce the prices optimized for the well-being of the people persists despite the existence of debt, money, governments, or other natural catastrophes. In fact, the invisible hand of the market "loves" to work with these factors and its ability to work with these factors is a necessary condition for the proposition "capitalism works" to have any beef.
Fine. What is the money? Money is objects or materials that guarantee that a person X – a physical person (human being) or a legal person (a company or an institution or a government) – that possesses the money will be able to use the money in the future to convince other persons, such as Y, to give or do certain things that are either needed for X to survive or that are making the life of X pleasant.
The amount of money that exist must either be conserved or the violation of the conservation must be controllable enough for X to be confident that his or her or their having the money will be useful in the future, too. In other words, money is a tool that improves barter by allowing the persons to delay their consumption.
This is a good thing because – to mention a trivial example – farmers don't have to try to grow and harvest corn in the winter which could be hard. Instead, they may sell it in the summer and save the money for the winter when their business isn't working well.
The agricultural example is a mundane one and an obvious one. You may invent dozens of related examples in which the delay of the consumption – and the amount of money that has to be saved – is more or less constant. But it's important to realize that the qualitative growth of the industry and progress of the mankind depends on the concentration of capital. Sometimes, people may decide to realize a more ambitious project than others did in the past. To do so, they have to accumulate lots of money which are needed for the dreams to come true. When they come true, the money may be repaid.
So the absence of synchronization between the human needs and the ability to produce was the primary reason why barter wasn't good enough and people needed some "savings" to be able to buy things they needed even at times when they don't produce things. Initially, they would use some "moderately useful things" that are "more or less conserved" such as marten pelts.
Croatia still uses marten pelts as its official unit of currency. Their word for a "marten" is "kuna". Many other nations have used less stinky but equally unsophisticated objects as their currency. A very popular choice was a yellow metal. America and much of the world was using this metal as recently as in the second part of the 20th century. Well, they were using "proxies" to gold but it was supposed to be "the same thing".
If we wanted to "smoothly" restore the gold standard today, the $100 banknote would have to say:
The owner of this picture of Benjamin Franklin may visit a building in Washington D.C. and demand to receive 2.156 grams of gold. If he won't get the precious metals, he may at least kick Ben Bernanke or the president of the U.S. into the buttocks 100 times.The value 2.156 grams is written there to reflect the word "smoothly" (I am talking about a "smooth transition") and the current price $1315 per ounce (=28.35 grams). Note that the gold price has dropped by 32% since the all time high of $1921 in September 2011. The second sentence is added to emphasize that you can never be quite guaranteed to get the promised gold but if you don't get it, the central bank at least promises you to be a moral winner.
What would happen if the banknotes were supplemented with this extra comment to make them "backed by gold" today? Well, it's easy, Barack Obama's buttocks would be doomed. Everyone would understand that it's much more likely for the government to run out of gold than to run out of pictures of Benjamin Franklin. If the artificial peg weren't imposed, it's very clear that the price of gold in dollars would increase once people would start to mine it from the reserves (in D.C., in my example). But because the government promises a constant gold price and not an increasing one, it's obvious that it becomes increasingly attractive to buy the gold for this price once the "run on the gold banks" begins. It would be inevitable that all the gold would be picked by the speediest holders of the banknotes (half a year of the U.S. GDP would be enough to deplete not only all the U.S. gold reserves but even all the gold that has ever been mined, about $8 trillion at current prices) while everyone else could only enjoy the pleasure of kicking into Bernanke's (or Yellen's) and Obama's buttocks. The gold peg would clearly be unsustainable.
OK, let me assume that the reader isn't completely stupid and understands that we can't back all the money supply by gold (or martens) today because there is just not enough gold (or martens) to match all the remaining wealth of the mankind. In more modern economies, the value of the money isn't pegged to an arbitrary metal; it is pegged to a basket of goods (or goods and services). So the banknote effectively says
The owner of this picture of Benjamin Franklin may visit the building in D.C. by Wednesday and get 50 two-liter bottles of Coke and one 64-gigabyte SD memory card. The numbers "50" and "1" will decrease by 0.2% every month.To avoid unnecessary junk, I have simplified the basket to 50 bottles of a soda for $1 and some memory card for $50. The second sentence indicates that the government is working hard to target the inflation rate. In reality, the basket that measures the inflation rate has to be much more diverse and complicated, there are many subtleties about it, and the U.S. central bank isn't really targeting "just" the inflation rate – to know what the banknote is promising you somewhat more accurately, you have to know something about the laws and the way of thinking of the central bankers.
But despite all these subtleties, the modern promise on the fiat money has two major advantages over the gold standard: it is actually possible to fulfill the obligation because those $100 are spread over many types of things (many more than 2 products from my example) and there's a sufficient amount of each product from the basket to back the money; and the Coke and memory cards are actually much more important for the people's survival and well-being than some abstract precious metal (there isn't any independent enough, permanent calculation of how much gold you need to survive in the year 2100, for example – because it isn't really gold that you need to survive).
Banknotes as debt; and why this interpretation may be irrelevant for the users
Even when we modernized the gold standard and replaced the gold by a basket of goods and services, the banknote retained a key property:
The banknote effectively says that the U.S. government is owing you something.If you own the picture of Benjamin Franklin, you will be able to get something good from the government. That's what the banknote – along with some promises written on it (or at least in the laws) – more or less guarantee. However, that's just the most straightforward interpretation. The point is that no one prohibits you from giving the banknote to someone else. You may give it to a shoemaker (or a clerk in a shop) who just gave you a pair of expensive shoes. If you give him the picture of Benjamin Franklin, he won't call police – which may be a good outcome for you because you will have shoes and you will avoid the prison tonight.
As you can see, even though the value of the money depends on some promises by the government, you may use the money in so many ways that the money has pretty much nothing to do with the government – at least from your practical viewpoint. You don't have to like the government. As long as you believe that there are mechanisms that enforce the promise (e.g. inflation targeting), the banknote has a value now and will have a more or less predictable value in the future which is why it is a useful unit for you to organize the savings and delay the consumption. In particular, the U.S. dollar is still the mankind's most popular and most practical unit of wealth. But be sure that if you plan to be buying things for your national currency, it's safer to have savings denominated in the same currency – to reduce the currency risk.
Bond buyback program
Fine. We want to analyze the simple question: What are the macroeconomic implications of the quantitative easing? By QE, I will mean the bond buyback program:
Some institutions of the U.S. government print $80 billion in new banknotes every month and use them to buy long-term bonds (U.S. bonds are known as the "Treasuries").The program is buying these bonds from commercial banks – which makes the program even more abstract than it will be in our example. But for the sake of simplicity, I will assume that the bonds are being bought from 80 million U.S. citizens, Joe Sixpack #1 through Joe Sixpack #80 million, and each of them sells bonds for $1,000 every month.
Will it produce inflation? Will it make the economy running?
First, we must understand what this bond worth $1,000 that each Joe sells every month is. It is a piece of paper that says
If and when the holder of this piece of paper visits the second building in D.C. in November 2020, he will either get $1,140 or will be allowed to hire James Bond to kick some government officials into their buttocks 1,140 times in total.Note that the rate $1 per kick stayed the same. The kicking isn't done by you in this case; it's done by James Bond. That's why these papers are called bonds. ;-) But otherwise the logic is analogous.
It's important to realize that what you get in the year 2020 isn't $1,000 but a little bit more. Well, I assumed that the yields (relevant for these bonds) are currently 2 percent so I added 7 (like 7 years) times 2% to those $1,000 – to get $1,140. A more accurate calculation would deal with the exponential function and I would probably be able to do it. ;-) But this would bring some technicalities (and just second-order corrections) which are not important for this discussion which analyzes the problem at the leading order.
Just to be sure, the yields mean that the fair profit that the buyers of the bonds expect is 2% per year. (You may find a more accurate figure; it would make the discussion below less transparent.) The bond contains a fixed number – $1,140 in our case – telling you how much the bondholder gets in 2020 (the maturity date). The number is written on this piece of paper and the interventions of the Fed (the buyback program) can't modify what was already printed. Because I was assuming the same yields as the inflation rate, the bond is still expected to buy 500 bottles of Coke and 10 memory cards in 2020 (excluding nonlinear corrections) but in general, the interest rates and inflation may be different.
Fine. The central bank prints $80 billion in fresh banknotes (e.g. 800 million pictures of Benjamin Franklin) every month and uses the money to buy the bonds from those 80 million citizens called Joe Sixpack. (American parents could be a bit more creative when they give names to their children.) What is happening? What will happen? What does it mean?
First, you must understand why the Joes are selling the bonds. These are the first owners of the bonds that decide that it's a good idea to get rid of them. Why? Because the bonds got a little bit less attractive – and it became a little bit better idea to sell them. Why? Because the central bank could only buy those 80 million pieces of paper by offering a slightly higher price for each paper than the initial price. So before the intervention, the price of one bond could have been $998 and after the intervention, it was $1,002. In average, the central bank bought one bond for $1,000. But the final price is slightly higher than the initial one. The owners of the bonds who think that the fair price of the bond is between $998 and $1,002 sold their bond at some moment during the interval when the price of the bond was increasing from $998 to $1,002.
Of course, sometimes it's a bit of an accident who just decides to sell the bond and there's a lot of noise. But the noise goes in both directions. You may replace the processes by the "mean value" which says that the bondholders know exactly how much they like the bond and when they should sell it.
Each Joe Sixpack gets $1,000 in average. Well, the buyback might be organized so that everyone gets the same price every month. It doesn't matter. However, the price of the bond does change. In my example, the price of the bond grew by $4 i.e. by 0.4% or so. Because its maturity date is 2020, it also means that the interest rate from now to 2020 decreased by 0.4%/7 = 0.06% or so. The final price is constant, at $1,140, so if the initial price increases by $40, it reduces the difference between the final price $1,140 and the current price. If you divide this difference by 7 (seven years left), you will get the annual yields and they will drop.
Great. So what has changed after this 1 month of buyback?
Those Joes who sold the bonds no longer have papers that say that they may get the 500 Cokes and 10 memory cards from the D.C. officials in 2020. Instead, they have papers that effectively say that they may get the same thing now. Will it push them to spend the money? The answer is No. If they would be finding the idea to spend attractive, they could have sold the bonds independently of the buyback program and spend the money, anyway!
These bondholders became holders of cash (more precisely, the percentage of their wealth they hold in cash has increased and the percentage in bonds has decreased). The price of the bonds has increased; the yields have decreased.
These changes produce some winners and some losers. Because the price of these bonds increased by 0.4% of their value, those who had the bonds before the operation benefited. They were lucky. (Perhaps they only gained 0.2% of the value because in average, the transaction occurred in the middle of the interval.)
On the contrary, those who were planning to buy these particular bonds – and save through 2020 – are losers. They will only be able to buy the bonds for a price that is 0.4% higher than one month ago; that's equivalent to the decrease of the expected annual interest rate between now and 2020 by 0.4%/7. Some of them will accept the bonds despite the decreased yield (despite the loss introduced by the central bank's intervention); some of them will cancel the plans to buy the bonds.
The government's often arbitrary acts are external events that you must treat as natural phenomena – as laws of physics of a sort. You can't really avoid them. In most countries of the world, it is actually illegal to shoot all members of the government along with all the government apparatchiks. You may see that the government's interventions are analogous to a natural catastrophe such as a hurricane. A hurricane may destroy some buildings in New Orleans; a government intervention may destroy a part of the expected returns of some savers. A hurricane may have winners as well – e.g. those who made a bet that at least 1,000 people would die in New Orleans. A government intervention may have winners, too. For example, a carbon tax may destroy most of the economy and starve millions of children to death but Al Gore will benefit.
An intervention by the government – including this buyback program – is a distortion or deformation of the market. But that doesn't mean that capitalism gives up once a similar natural catastrophe (government intervention) occurs. A hurricane is also a "distortion of the market" in New Orleans but capitalism is still the most efficient system that allows the city to recover.
What about the cash – the new banknotes – that suddenly belongs to the Joes? Has something changed about their desire to save i.e. about their lack of desire to spend?
The answer is No. Nothing has changed about their overall lack of desire to spend. This desire is dictated by their optimism about their future job, the future of the economy, and their biological instincts that make it irresistible for someone to spend the money. But they just don't have enough of this instinct so they're savers. In fact, we could even suggest that Joe – a natural saver, still having some long-term bonds – was just made sure that he will be poorer in 2020 than previously thought; that may encourage him to save even more. The others, "natural borrowers", were not affected all.
The only thing that has changed was that the Joes have determined that the 2020 bonds are no longer attractive enough for them after the 0.4%/7 decrease of the annual interest rates. But that doesn't mean that they were pushed to the alternative – to spend the extra $1,000. Nothing has changed about the attractiveness of spending.
What has changed is the attractiveness of saving in the form of 2020 bonds. And there is another thing that has changed, although many people fail to see it. What has changed is the attractiveness of holding the cash in saving accounts – or in short-term bonds. It has moved in the opposite direction than the attractiveness of the 2020 bonds. Why?
Simply because the government has used the cash to buy the long-term bonds. In the case of bonds, we saw that those who had the bonds before the buyback became winners and those who didn't have them but needed to buy them were losers. In the case of the cash that was used by the central bank to pay, the situation is opposite. Those who had the cash before the buyback are losers (well, yes, because the money has been printed) but those who plan to sell something and acquire money are winners!
So it's clear that if we remove the noise, the Joes (the market) will adjust and react exactly in the opposite way to the intervention by the government. They have sold the bonds, acquired cash, and it's a good idea to have cash. They place it to the most attractive saving account or buy some short-term bonds which is pretty much the same thing. I am not saying that every Joe will do exactly that; they may randomly do other things. But the expectation value of this extra noise is, once again, zero.
The long-term deposits or bonds may offer you higher interest rates than saving accounts or short-term deposits. But they have disadvantages, too. If you correctly account for these advantages and disadvantages, you will realize that already before the buyback program, there was an equilibrium. For the "effectively average" investors, holding the money in the saving accounts is equally attractive as holding the wealth in long-term bonds. Yes, I am assuming an "efficient market hypothesis", if you wish, but it is true, too. If one of the things were more attractive, people would rather quickly drift to it from the less attractive option. The equilibrium is a trivial fact. The question is not whether it's true. It is true. At most, we may ask what timescale is needed to achieve the equilibrium and what is the signal-to-noise ratio.
(Of course that if the timescale suggests too slow a thermalization and/or if the noise always remains high, we might effectively say that the markets are not efficient. But the sign of these inefficiencies isn't really predictable because these deviations from the right prices are irrational and irrationality has both signs and many flavors.)
So even though you may fail to see why people accept lower interest rates in saving accounts – or even lower interest rates in their saving accounts relatively to some analogous saving accounts with higher rates – it may be just to your blindness. People have some reasons, at least statistically. They find their bank more convenient, closer to their home, more trustworthy, and so on.
The central bank's intervention turned the previous holders of the long-term bonds into winners, the previous holders of cash into losers. It discouraged people from buying new long-term bonds but it encouraged them to sell things (well, long-term bonds) and acquire new cash. As a proxy, it discouraged them from having things similar to the long-term bonds and it encouraged them to get things similar to the cash (saving accounts). Because of that, the short-term interest rates are pushed up by the buyback program – they're "proxies of cash". They're not just proxies because the central bank is literally paying an interest rate on reserves.
In effect, the intervention has just redistributed the wealth in a random way, by producing some winners and some losers, and it has made some forms of savings more attractive and other forms less attractive. If you only look at the overall impact and you don't care who is a winner and who is a loser, the aggregate effect on inflation and the economic activity is zero, at least in the leading order.
To see this outcome, you must fully realize that the bonds and the banknotes are completely analogous pieces of paper. Both of them are papers in which Joes may keep their savings and wealth. Both of them are a form of the government debt. The intervention only changes the relative prices of these two things. But if you don't care how the people distribute their wealth into bonds of various kinds and/or saving accounts, and it really doesn't matter, the impact is simply non-existent.
Incidentally, there is a simple reason why it's good for the Joes to hold cash. They may just hold cash and wait for the moment of "tapering" when the interventions stop. At that moment, the long-term yields will start to increase again and the bonds will get cheaper. If the Joes "jump" into the long-term bonds again, they will actually recover the lost profit from not having the long-term bonds during the relevant part of the quantitative easing.
I will discuss the scenario in which there's no tapering at all at the end.
Analogy: two currencies
Imagine a country, like Denmark, that may use two currencies – e.g. the Danish euro and the Danish crown – simultaneously at some moment. (I chose a special Danish euro to avoid the complications resulting from the fact that the normal euro is also used by others and influenced by other countries' policies.) The Danish euro may be claimed to be the long-term currency that thinks about the far future while the Danish crown may be presented as a down-to-earth short-term currency. But the distinction is just on the paper, it doesn't really mean anything tangible. The difference between them is materially dictated exactly by policies similar to the buyback program (and whether they are allowed). The bond buyback program is analogous to an intervention by which the government changes the Danish euro-to-crown exchange rate. It's nice, creates some winners and losers, but it clearly doesn't change the overall debt of Denmark as counted e.g. in dollars and the overall amount of money it has to pay for interest rates.
What happens if you try to push long-term interest rates to zero?
I was trying to argue that if you count the cash as well as bonds into the "money", with weights representing how much Joes are holding each, the QE operation doesn't really influence the inflation rate, either. If that's so, isn't it a great idea to do as much quantative easing as possible? Isn't it a great idea to push the long-term interest rates to zero? (You can't really push them below zero because no one would buy the bonds: you may get better, vanishing, interest rates in the mattress.)
Instead of reacting emotionally, like "holy crap, this shouldn't be done!", one should rationally think what would happen if the policy changed in this way. Can't the U.S. government simply change its laws and print all the new money it needs, instead of borrowing money and issuing bonds?
First, if the policies changed in this way and if it were possible for the U.S. government to cover the new debt by the fresh printed banknotes, it would encourage the government to increase the debt almost indefinitely. Such policies would eventually lead to (hyper)inflation and the fall of the dollar's exchange rate. But I want to emphasize that the piling of the debt is an independent process from quantitative easing (done by someone else, by the way). Above, I was assuming that the budget deficit was kept the same during the bond buyback program.
Second, yes, you could push the long-term yields to zero. However, as argued above, this would inevitably lead to higher (positive) short-term rates. You should notice that there's something wrong about short-term interest rates that are higher than the long-term ones. This anomalous shape of the yield curve is known as the inverted yield curve and it is considered a predictor of recession. I would add that if the yield curve became inverted due to artificial interventions, this inverted curve is not only a predictor but the interventions were the cause of the recession. The recession would occur because the money stops circulating (at a certain timeframe) due to the negative interest rates. What do I mean?
Even without vague historical observations, it's possible to see what's wrong with this environment (described by the inverted yield curve). If the interest rates you get between 2013 and 2017 are positive and those between 2013 and 2017 are zero, those between 2017 and 2020 are expected to be negative. But negative rates can't really become popular because the mattress offers you better rates, namely zero, so the money will stop circulating in 2017.
So the intervention of the quantitative easing type just reshapes the yield curve and once it becomes inverted (so far things are OK because the short-term rates are still very low), the central bankers finally start to realize that it has the potential to produce negative rates for some timescale sometime in the future, and that's when they will stop the interventions, hopefully. ;-) If there's some "extra job" that the central bank should be doing aside from setting the basic rates, it's preventing the curve from getting inverted (kind of the opposite than what they're trying to do). But as long as they manage to avoid this predictable trap, the effect of this whole exercise is zero to the leading order.