Those are helpful examples of the expectations mechanism Hellestal but I would argue that the exchange rate one is far more awesome.
In the first case the Fed is creating an immediate arbitrage opportunity with a very short time horizon. Also, AFAIK (correct me if I’m wrong) you are comparing 2 periods with different open market desk policies. In the early 1980s, money supply was targeted and interest rates were allowed to vary. Then things evolved towards strict interest rate targeting. (What the heck happened in the very late 1990s? They had been targeting the Fed funds rate for over a decade then, but the market seems to have settled down further? ) Anyway, as you noted this is a 1-step expectations story which is rather different. (Also, if push comes to shove, the Fed can AFAIK make a special purchase or sale of bonds in the middle of the day should the rate get out of line. Again, IIRC, AFAIK.)
The exchange rate story is more interesting, because the monetary authorities aren’t the 20 megaton Godzillas that they are in the interbank market. A 10% change in the exchange rate is a big deal. Another example might be the decline in the dollar during the mid 1980s after the G-5? announcement. And there are a lot more moving parts here: we have flows of goods, assets and speculation driving the market with the biggest but by no means absolutely dominant player being the government. Ask George Soros. Again, good example.
Let’s make everything sane and call it a 3-4% inflation target or a nominal GDP path of 5.25% extrapolated from 2007. I agree about the 10 year bond. Measurements of velocity are basically quarterly figures… where is the nominal growth going to come from? Plant and equipment investment requires longer lags. Do you think people will start buying more cars after they observe auto rates increase? Is this an inventory investment story? This is where I see a lot of hand waving and others less kind than me characterize as an appeal to the Inflation Expectations Imp (close relative to the Confidence Fairy).
To be clear, I favor unconventional Fed Reserve policy because I see very little downside to it. And I don’t doubt that Ben can call Great Expectations from the vastly deep. But so can you. And so can I. But will they come? My take is that we really don’t know until we try, and then we have a nontrivial econometric/economic history challenge. [1]
There’s a huge difference between asserting that a given policy will change prices in the financial market and saying that they will change goods market behaviors. Especially when the linkages between the 2 work in the opposite direction: we would expect higher interest rates to slow down growth, not speed it up. Though I concede that both bond prices and goods prices respond positively to higher aggregate demand, and that the latter has some relationship to manipulable expectations, but to an unknown magnitude.
Heh. There’s a certain twilight zone aspect to the OP. It’s not the sort of policy that I would advocate or I would want to risk.
[1] I’m channeling discussions at the Brad DeLong’s and Scott Sumner’s blog.
It’s about the broad macro target versus their direct policy lever target. The policy lever is always interest rates/monetary base, regardless of whatever broad macroeconomic target they have which changes over time.
Pretty much from the historical beginning, central banks have used interest rates as their lever. That naturally includes the brief Old Monetarist experiment (which was, by the way, much more doggedly pursued by the Bank of England than the Fed, which had a more practical mindset). But they don’t just target the money supply directly. I mean, hell, that is literally impossible. They don’t control it. M2 is private-bank money, not central bank money. This is, according to your own favorite Goodhart arguments, part of the reason why that particular macro target didn’t work. Private banks control it directly – it’s their own creation – which means they can manipulate it directly.
You can go to the graph and switch to the monthly view to see they really did have their thumbs on the rate in a broader sense. Or hell, I can link it myself. You can read old business news about the raising and lowering of rates, too. The only big shift historically was a change from mostly relying on the “discount rate”, i.e. the rate for borrowing direct from the central bank, to instead influencing the interbank rate by buying and selling stuff on the market.
The volatility was truly about banks gouging each other for funds to meet their reserve requirements. I’m failing to come up with the name, but a while back we had a poster right here on the boards whose former job was to try to model this stuff so that the bank didn’t get screwed by other banks. Eventually, the Fed decided to end that circus. I’m not sure precisely why.
Yet for most of the history of central banking, we were dealing with fixed exchange rates. Monetary authorities really did manage to do it for a long long time. It eventually failed, sure, but the Swiss example doesn’t tell us that fixing the exchange rate is merely possible, because, well, we already knew that, thanks. History is real clear on that particular topic.
The example tells us about the extraordinary influence of expectations on a market. That is what is fascinating for me personally. The instant arbitrage based on a press conference. You’re right, of course, that foreign exchange is a monster market, but the size by itself isn’t what captivates me.
This is one jigsaw piece. I admit it is not the whole puzzle.
The next step is linking the markets together. Simultaneously.
Two starting comments from me here:
This is a totally legitimate question, so I want to say, that there is no reason for you to believe me on this, and that I wouldn’t believe me either.
All of the evidence is on my side. Even if I you don’t find my specific argument for the mechanism plausible, then you would still need to come up with some alternative explanation for what I’m going to try to explain. This is not something that can be dismissed as inconsequential. We actually see it happening.
The mechanism is… not so crude as you seem to describe. It’s very simple, actually, just hard to build up the foundation of thought to accept it. I’ll try to get to it in more detail, but I can’t talk about it directly because it really does seem to need prep-work.
And I would say that they have tried and that we have seen it happen in front of us.
As I already pointed out in post 11, if you take the yield for the 10-year regular Treasury, and you subtract the yield from the 10-year TIPS, we actually do get market inflation expectations, right there, plain as day.
This isn’t kinda sorta quasi maybe inflation expectations. It is honest-to-god literal inflation expectations. TIPS protect against inflation. Ordinary Treasuries do not. The difference between the two is the risk premium that ordinary Treasuries have to pay because of inflation risk. It’s not 100% perfect, there’s a tiny bit more going on, but it’s really a genuine measure. It is right there, an equation we can rely on:Treasuries - TIPS = inflation expectations
So let’s look at it again, in more detail this time. Here it is for the last five years. Open a new tab for that and then come back and look at these dates:
QE1: Nov 2008 to Mar 2010
Jackson Hole QE2 speech: Aug 2010
QE2: Nov 2010 to Jun 2011
Twist: Sept 2011 to late 2012 (but then QE3happened)
That by itself should be a wake-up call, but there’s more if you want it. We can, for instance, look at equities. They all say the same thing. This is the real data. I don’t need to rely on a mere argument that the Fed successfully influences inflation expectations, because I can literally link to the graph.
The task for me isn’t to show that the Fed influences inflation expectations, even at the zero lower bound. I have already done that. Anyone who knows the significance of that spread should see it immediately. When Krugman speaks here of David Glasner’s paper on the relationship between expected inflation and asset prices, they’re referring to the same data, just in another context, to demonstrate (correctly) that our current economic problem is with demand. The strange relationship between equities and inflation expectations means something is really going on here.
What remains is to explain how it works, the “transmission mechanism”. That requires many more puzzle pieces. What I’m trying to do here, I guess, is break that fancy theory down into easy individual pieces so that we can fit together in a way that people can make sense of. We can see Fed action pushing on this. You rightly want to know how.
I think I can do that. It might be a little long, though.
I think I’ll start with the difference between monetary base and government debt. They’re both assets on bank books, but they work very differently. It might be worth investigating how for starters.
First and most obvious, the price of a dollar is always a dollar. If I have a dollar, and there’s a major recession, at the end of the recession my dollar will be worth one dollar. If we experience boom times, at the end of the boom my dollar will be worth one dollar. It’s always worth that amount because that is what it literally is. In order for the value of my dollar to be different, the price of everything else in the economy has to change. This is important because prices don’t change instantaneously in all markets at once, and they don’t change at the same speed, which means there is an opportunity to make a good deal in a purchase by taking advantage in a sale before the slower-moving prices change.
Second and also important, but a little less obvious: there is a cost to holding a dollar. Even if the inflation rate is zero, there is a very real cost. Dollars have a yield of zero percent (or 0.25% if we’re talking excess reserves). The opportunity cost of holding a dollar is giving up the yield you would have otherwise received by holding an asset with a genuine rate of return. This means that dollars are inherently faster-moving, higher-velocity instruments than bonds. There is some pain to holding them, like they are hot to the touch, even at 0% inflation. We might not ordinarily think about this in our everyday lives, but big financial institutions are well aware of how much money they aren’t making when they’re holding dollars. And the higher the inflation rate, the hotter they are. They can easily become a hot potato, where people receive them and try to dump them quickly.
These two facts together mean that people have interesting options available to them when they’re holding dollars. This is big stuff. We need a stupid example to demonstrate.
We can have two people holding assets. One has 1000 cash, the other a government bond currently priced in the market at 1000. The bond has a rate of return, the cash not, so we already know that the cash is more painful to hold and therefore much more likely to move first. The person holding the bond, in contrast, is more likely to be satisfied with it as an investment.
(We naturally realize that banks actually are sitting on a lot of cash at the moment, but we should already see why that is so ridiculously weird. Something is off. Eventually we’ll figure out why.)
Suppose that we start with zero inflation but the world wakes up tomorrow morning totally convinced, for whatever reason, that prices will be noticeably higher next year. What happens? First, the 1000 cash is still worth 1000. In addition, it’s value is still mostly stable. People expect prices to be higher next year, but all those sticker prices on the shelves haven’t changed yet. That plasma screen will be more expensive a year from now, and if you act today, you can snap it up before that happens. You can take advantage of the prices that change slowly by purchasing before the change happens.
The bond is very very different. The price of a plasma screen is a bit sticky and won’t change overnight, but the bond market turns on a dime. The very moment that everyone in the world wakes up expecting higher inflation, the price of the bond will drop below 1000.
The price changes immediately, incorporating the new expectations about inflation next year. This is to say that while the pain of holding cash is a small bite every single day, 365 days a year, the pain of holding a bond can be a big damn bite that happens in a single instant, and then is finished. So what does the bond owner do? The price is down, nominal yield up – the owner took a loss – but the real yield (adjusted for inflation) is the same, so the owner still has no additional incentive to sell. Done is done. It took its bite and went away. If there is a one-time change in inflation expectations, there is no reason to fear any more future pain. You’re free to keep the bond, and even expand your bond holdings if you’re convinced this is as bad as it’s going to get. When all that big damage is packed into a single tiny moment, you’re free to stick with what you’ve got.
This has big significance for monetary policy. The simple act of changing out a bunch of bonds for a bunch of cash means you’ve removed from the markets an asset that people like to sit on like eggs, and replaced it with a huge stack of fuel, the entirety of which could potentially ignite simultaneously. It’s gotta be simultaneous. If one dollar is losing value, then all of them are. And naturally, the bigger that stack is, the more likely it is to reach a critical mass and start the chain reaction.
The question is, can the chain reaction proceed on its own, or are there any cooling rods in place? As it happens, a central bank with a big balance sheet is our cooling rod here.
You are a bank. You had a bunch of bonds. The central bank showed up and offered to buy your bonds for cash.
All else equal, you’d much rather have the bond yield than the cash, but the people buying have some serious weight. They can make a sweet offer that you’re willing to accept. So now you have excess reserves instead of bonds, and you’re wondering what to do with this potato that has not yet become hot.
You are a bank. You now have cash. What are your options? Three things. 1) You can do nothing and sit on it. 2) You can extend a loan. 3) You can buy securities, such as more Treasuries or MBS. The process for the latter two is incredibly similar.
DEBIT new asset, CREDIT deposit liability.
The question is, when would you want to do these things? The process is similar for both, but the incentives are subtly different. What about extending a loan? You’d be happy to extend a loan if you feel the rate of return would be good, given your likely future cost of funding, and that you’re very likely to be repaid. If we are at 2% inflation and the Fed is talking about how they’re a bunch of pansy-ass weaklings who want to keep inflation well “anchored”, then extending a whole bunch of loans is obviously not a good idea. If you are your bankster buddies are all extending loans, economic activity increases. If that happens, inflation increases. If that happens, the Fed gets all worried about inflation and possibly yanks back on the chain. If rates go back up, then you are hit with a double whammy. The economy stagnates, decreasing your chance of getting paid back, and your own personal cost of funding goes up. Loans are terrible ideas. Better to sit on the cash.
What about buying another security? Well, why the hell would you do that? In general terms, you’d rather hold the security than cash. But to lure you into selling your old security, the central bank had to make a super-sweet offer. They paid you enough that you were willing to part with the old security. Fair enough. So right now, there does not seem to be much incentive buying a new one. It looks like, in this one very strange situation, the one asset really is similar to the other asset. Why is this situation strange? Because the potato isn’t hot right now. Should that heat rise, the two things that seem temporarily similar become extremely different again.
Next puzzle piece.
The central bank grows some big brass gonads. They commit to a higher inflation target (a higher nominal spending target would be very similar in the short run, but we’ll use inflation for convenience). They make a very simple announcement: “We’re going to buy 100 billion of assets every single week until the 10-year Treasuries/TIPS spread is 4%.” What they are saying, of course, is that they want 4% expected inflation, and they will increase the potato heat until it starts jumping. (We’ll also say that they’re practical, they’ll be a little bit lenient if the less liquid TIPS market suffers some manipulation. They’ll ask Treasury to issue more or something.)
Woof. Crystal clear statement. This is how the Swiss roll.
What happens? Before the press conference is even over, the holders of longer-term Treasuries take a huge hit. There should not be much doubt about this. The drop happens and it happens immediately. (Unless the Fed steps in to support the market with specific purchases, which could totally happen but we will ignore for now.)
You are a bank. You’re sitting on a fat stack of cash. This is the same question as before. What do you do? You still have three options: 1) You can do nothing and sit on it. 2) You can extend a loan. 3) You can buy securities, such as more Treasuries or MBS. The question is the same, but the market situation is totally totally different. Option #1 is no longer palatable. You are not going to sit on cash that not only does not offer a yield, but that has an active negative yield because of the higher inflation.
How else has the situation changed? Equity markets just moved up. Longer-bonds just moved down, so their yields are up. Long-term funding for housing might look to be in some trouble, but just wait a sec… Interest rates are higher, sure, but only because of inflation. That means real rates haven’t changed for long-term lending, only nominal. There will also be upward pressure on housing prices. So while rates look higher in nominal terms, they aren’t in real terms, and the asset itself is even more valuable. You haven’t hurt long-term borrowing too much, or at all.
Do you loan? Do you extend business credit? Well maybe you don’t, maybe you’re still worried. But it’s notably more likely for you to loan, and on fairly good terms for borrowers, even with the higher inflation. First, cash is poison. Also, your short-term funding seems much more secure. Asset prices are also more secure. The chance of repaying also seems more secure. So right here, with lending, the money should move.
What if we insist that lending still looks bad, at least for some banks? The other option is to dip into securities again. And this is totally doable. An increase in inflation expectations will cut into long-term bond prices, but only the once. Their new price has already incorporated the change in situation, so there’s no reason they take another hit. Meanwhile, cash is still poison. Too hot, hot potato. It must be passed on. So even if more loans look bad, then the banks will definitely load up on securities again. They will rebalance their portfolios, getting rid of one asset whose value is constantly decreasing, and buy another good asset that offers some return.
However. This time they are not transacting with the central bank. They are out in the world, buying these bonds from big investment companies and other institutions. Those investment companies take the cash. Why? They aren’t going to sit on the cash either. They’re not stupid. They are selling their bonds in order to buy other assets. The money is a hot potato, and it’s passed from institution to institution as they all rebalance their portfolios, looking for good opportunities in this new situation.
Money is bouncing around. It must. It’s way too hot for anyone sensible to want to hold on to it. That must absolutely happen. The Fed can influence the TIPS spread, just as it can influence the interbank market or even foreign exchange to some degree. The money will move. Everywhere the money goes, it must get passed on to somewhere else. No one receives money to keep it, they’re all just rebalancing portfolios and then passing it right along.
We still haven’t reached the real economy. But we are so close. This is a lot of typing, but let’s try to finish.
I started this whole sequence of events with the banks, because the banks are the ones to transact with the central bank. Everyone knows that central banks buy assets from regular banks, and so people are looking for a chain of logic that starts with the banks and then somehow leads to the real economy. I have attempted to oblige. I’ve given one logical step which I don’t think anyone can deny. Then I’ve given another logical step which I don’t believe anyone can deny. Then another. Then another. I’ve given the sequence of events that shows money moving. This is the whole idea behind “transmission mechanisms”. We can see clearly the chain of events that leads to money movement.
Hey, news flash: the banks aren’t the only ones that have money. Everybody has money. You have money. I have money. That bum has money, fifty cents he just received panhandling. God damn everybody has money. You now what the banks have? New money. You know what the economic difference is between new money and money I stuck in my mattress five years ago? Not a god damn thing. It’s not just the big sophisticated financial institutions that are all trying to rebalance their portfolios which is moving cash around. Everybody, everywhere, has hot potatoes in their hands.
All money moves. It all moves for the same reason at the same time.
Some people are slow and won’t appreciate what’s happening, not right away, but there are plenty others who know a good deal when they see one. If my mutual fund just increased in value, but that plasma screen I’ve really been wanting is still the same price – and I’m expecting this good deal to not last very long – then what am I personally going to do? You’d better believe I’m going to grab some of my fund and buy the damn TV today when I can get it for a song.
All money gets hot together. When money moves, it’s not just bank money that moves. It is all money, everywhere, that everyone holds. So do the banks lend money? Of course they do. Everybody has started spending their money together, because spending is better than losing the value of the cash. I’ve been describing this like a chain of events, Step-A leads to Step-B leads to Step-C, but only because that’s what our narratively-biased minds want to hear. And in fact, I have explained such a series of events accurately, since we need that logic to see how it works. But once we finally understand that logic, we can apply it everywhere, to everyone’s money, simultaneously. If we can see the logic in following the cash move from a central bank, to a bank, to a bond dealer that wants to unload bonds for cash, to another bond dealer, to an institutional investor, to a brokerage, and finally to me as I cash out my own fund in order to buy the TV – if we can see that chain just one time, then we can see it again and again. We can see all the lifeblood moving around the whole body instead of trying to grasp one single chain of causation.
And we can see that it applies to absolutely everyone. That single chain is kind of an illusion, the mental lure of human storytelling. In fact, everything happens at once. For the next link in the causal chain, we find that everything happens at once yet again the next day.
I want a TV. Asset prices go up. I cash out. I buy a TV.
A large corporation with a massive amount of cash on hand decides they don’t want to lose the value of their cash to inflation. You do realize that corporations are sitting on enormous dragon-hoards right now? They truly are. They decide to invest instead of losing the cash inflation. Higher equity prices provide incentive for this.
A bank has massive excess reserves. They have a borderline loan opportunity. Higher inflation pushes them over the edge, not because the loan is necessarily great in their view but because it’s better than suffering the inflation loss of the reserves.
Another bank has massive excess reserves. They find someone selling securities. They buy the securities. After a series of transactions where various firms rebalance their portfolios, a guy decides to sell his assets to buy a new Jet-Ski.
Commodity prices are higher from higher inflation. Borrowing is cheap. A mining company starts a new project, hires some workers.
The US dollar is down on the foreign exchange markets. There is more demand from Europe. A factory expands production, hires an additional shift of workers.
A Jet-Ski dealer is having a surprisingly good month. She decides to take an early vacation to celebrate the turnaround in economic fortunes.
A tourist destination is having a surprisingly good month. The restaurants and gift shops in the town are doing good business. They hire more workers.
And on and on and on. It’s all interconnected and it’s all happening at once.
What is the true “transmission mechanism” of monetary policy? It is, simply, the circulation of money. And money circulates more quickly when it’s hot. It can be described as a chain of events, if that’s what we need to hear the first time to get our minds around the process, but it applies generally. It is everybody, all of us, realizing collectively that our dollars aren’t going to be worth as much in the future as they are today. And we respond to that naturally, as all human beings do.
I’ve been dealing with this so long it’s hard for me to understand how other people miss this. After it finally clicks, it’s ludicrously obvious. We even have stories about it, in extreme situations. A joke from Keynes talking about Austrian inflation in his “Tract on Monetary Reform”:
This is so utterly obvious in extreme situations that people, somehow, miss how obvious it is in less extreme situations, too.
The principle is the same, exactly the same. It is only a difference of scale. The transmission mechanism of monetary policy is simply money moving, which it would naturally, effortlessly do if there were a higher price of it sitting still. Our problem is a lack of “aggregate demand”. You know how aggregate demand is often defined? GDP. Which is to say, how much money changed hands for new goods and services. I normally specify “nominal” when I’m talking about this stuff, but GDP is naturally nominal until it’s been adjusted. And nominal GDP is just the same as the money supply times the velocity of money. Nominal GDP is just money moving.
Our problem is that money wasn’t moving, which caused trouble in the labor markets which have extreme price stickiness. The solution is to get money moving again so that the labor markets can clear.
And when things get too hot, and the TIPS spread starts creeping up to five, that’s when they pull back the chain. Asset prices decrease, loan opportunities decrease, and the economy cools again.
I don’t really have any more time. Big project for the end of the year.
If something is unclear, as is probably the case, I can check back here next week. Not daily, just can’t do any more like this, but I can check in once a week.
I agree reserves and currency are different from commercial bank deposits. Commercial bank deposits can be redeemed for currency, but currency itself is irredeemable. Commercial banks can be bankrupted. The Fed has a balance sheet, but can’t really be bankrupted. The Fed’s liabilities are legal tender.
But where is the evidence that Fed liabilities cause inflation?
Comparing base money to the price index doesn’t seem to show a connection.
And as far as Fed liabilities not being real liabilities, that’s true, but the point is that when the Fed issues liabilities, it also takes private sector assets (takes them from the private sector).
Let’s say that you’re right, and that when you look at, say, the stock market, you can see a correlation between the Fed announcing new purchases, and rising stock prices. (I haven’t looked.) Why is there not similar empirical evidence of the Fed’s effect on the prices of real goods and services?
Okay, seriously don’t have more time here but I did say I would answer this question. So really seriously last one this week. I’ll come back in a week.
There is zero reason to believe in a direct connection between the monetary base, or any definition of the money stock, and inflation. Zero reason.
Velocity is not stable. Money demand is not stable.
When they create new money, they can’t say for certain exactly how much money people will want to sit on, and how much will move. If a new religion suddenly took hold in the world that indicated that we should burn money in order to send cash to the afterlife to support grams and gramps, we could naturally see no direct connection between the increase in the monetary base and price movements. Some of the money is getting a fire-transfer to heaven. We couldn’t say exactly how much, though, because we wouldn’t know exactly how many new adherents to the religion, or how fervent each individual was to support their ancestors. We can’t possibly know that. There are literally hundreds of millions of us, and even more if you count foreign holders. You can’t predict velocity in such a situation.
And if people decide to try to save suddenly, all at once, you get an extremely similar effect. People hold their money instead of spending it. How many savers are there? How religiously fervent in their saving? We don’t know. We can’t know. Less extreme changes in velocity/money demand happen all the time, too. There is no way, none, to rely on a correlation between the money stock and inflation because there is no way to see directly how much people want to hold.
To see Fed influence on inflation, you can’t look directly at any individual money stock because you can’t perfectly predict how much people will want to hold. You absolutely have to look at broader markets that indicate people’s thinking, like the 10-year TIPS spread.
If you can’t see the Fed influence on inflation from the 10-year TIPS spread, then I’d suggest a new monitor with higher resolution.
We were in actual deflation when the Fed was hitting the gas with QE1. Core inflation was extremely low. Then it came up again. Then they said they weren’t satisifed, and tried again. Then it went up to roughly 2%. Everything they do influences the inflation rate, it just naturally takes the real economy a little bit longer to catch up with the expectations.
Let’s say a strange force, beyond our ken, wipes out all the gold on earth. Maybe it reverse-alchemies the stuff into lead or something. The point is: no more gold.
And I personally spring into action. I happen to be the only person in the world researching alchemy, and the mysterious vanishing of gold leaves evidence that is picked up on my own equipment and no one else’s. Only I understand what happened. What’s more, I can reverse it. I can bring gold back into the world. It’s actually pretty cheap to do, once you know the devilishly clever secret. It’s just like minting coins or printing paper money. I can make and unmake gold as easily as hitting buttons on my computer. It remains forever my secret alchemical gold recipe.
So I start making money.
Gold is a classic monetary asset. Politicians, sniffing out an opportunity to appease the gold bugs in their constituencies, give me permission to play around with my equipment. I make pennies that look like pennies, but with a tiny dollop of the yellow stuff inside. I make nickels with five times that amount, dimes with ten times the penny amount, and so on. I play with paper, making paper banknotes with gold thread weaved in, the one dollar version having exactly 100 times the penny mass of gold, the five dollar bill with five times that amount, and so on, all in perfect proportions. My version of Franklin is super shiny.
At first people hoard the stuff, but I make so much of it that eventually people accept my pennies as pennies. They accept my dollars as dollars. Banks decide to accept it, too, same as the stuff that comes from the Bureau of Engraving and Printing. Everything is working swimmingly, exactly as before, except most of the currency in circulation is me-printed and me-minted rather from the gummint. Then I get even more ambitious and manage to take over bank reserves. Gold has that magical effect on people’s minds, so I’m able to swing it.
I supplant the Fed. The US starts using Hellalchemical gold currency.
Voila. The US is (technically) no longer using fiat money. A dollar is defined as a precise amount of gold, isn’t it? Yep. People who are holding a paper bill are actually also holding gold? Yep. There is a real physical thing that’s going on. Even if I die and my machine breaks down, that new gold will still have value in and of itself, just as gold does for us now. It could still be used as money, even after I’m gone, if that’s what people wanted. (Incidentally, though, the same is true for paper banknotes. The czarist ruble was still used as money for many years after the Bolshevik revolution. It still had value, too. There were no more czars to print more.)
Now here is the crazy-ass part.
If we were going to set up the accounting procedures to keep track of this gold creation, how would we do it? There are actually two ways. I could bump my assets every time the machine starts humming and new gold is created (the corresponding double-entry is a credit to equity on my books). Makes sense, right? This metal is valuable, and if I can create more, then the actual accounting procedure should be to reward my asset column every time the machine sings its golden hymn.
There’s another option, though. I could do absolutely nothing when I create the gold. It doesn’t show up on the books when it’s created, like it’s not even there, like it doesn’t exist. This precious metal, sitting in the vaults, would not appear as even the tiniest blip on the accountants’ radar at all. That is, it would not show up until I used this asset to purchase something like government bonds. Then and only then, when the gold is in circulation, would it show up on the books. And it would show up as a liability. The corresponding double-entry would be a debit to assets in the form of the new bonds or whatever that I bought with my newly alchemized gold (thus equity would not change).
The question at issue is: Which makes more literal sense?
That has an absurdly easy answer. If we were talking about a private company, the new gold would show up as an asset when produced. There is absolutely no question about this, it is how mining companies actually operate when they pull precious elements out of the ground. As it happens, though, real central banks do the exact opposite. When they create new monetary base, new artificial gold, they really do mark it down on their books as a liability when it goes into circulation.
Our current system is in all practical respects exactly identical to the gold alchemical system. If this machine actually existed, we could replace all of our current dollars with new gold dollars and the effect on the economy would be… nothing at all. (With the exception of gold miners, who’d have to find something different to dig out of the earth.) As far as money goes, there would be no effect. It would work exactly like our present system.
We treat our monetary base from central banks, our real-world alchemy, like a liability on the books instead of the asset that it actually seems to be. There are legitimate historical reasons for this. I’m not arguing that the accountants should change their system. I’m just pointing out the utter arbitrariness of standard accounting definitions when we’re dealing with money on this level of abstraction. Double-entry bookkeeping was not created for central banks, and the notions behind it do not fully apply. We could alter our present system, changing around the columns, and… nothing would fundamentally change. So long as my own personal central bank continued to create and destroy money according to the same criteria the Fed uses now, it wouldn’t matter in the slightest which columns get debited and credited.
The point is, we need to be very very careful when we’re talking about the monetary base. The base really is special. It’s different. It’s very much like an artificial gold which can only be mined by one central bank.
The accounting stuff is meaningless. There is no reason to call it a real liability. It’s not true that “all money is debt”. The monetary base is not debt, not a liability, in any normal sense. And in fact, calling it an asset would be more strictly correct. Crediting equity instead of liabilities would be more strictly correct.
The monetary base, the M0, central bank money, is legitimately a different breed.
On my own accounting ledgers of my personal central bank, the “equity” would be enormous. And for any real company, yes, the owners can take out that equity and use it for their own selfish purposes. Obviously, I could do that. But here’s the thing: real central banks can do this, too. They do it at regular historical intervals, and we experience it as hyperinflation. The accountants can call it a “liability”, but a true liability could not be abused in that manner. So yes, it can literally be considered equity. It’s just that equity does not have to be taken out of a company. Equity belongs to the owner, and the owner can do whatever they want with it, and that includes leaving it in the company forever. Which is actually what I would do. I wouldn’t need gold toilets or Lamborghinis or anything ridiculous like that. I’d pay myself a Ben-Bernanke-like salary, and I’d be a humble central banker for the rest of my life. No committee necessary, except for advice. I can decide policy myself, thanks.
And it would be a fairly interesting job, too.
That actually seems to be the real-world progression in central banking, too, in broad strokes. As it happens, the Bank of England was originally a private corporation. Complicated story but eventually it became so… public spirited, in a way, that they went ahead and nationalized it in 1946. Which is to say, again, that the only real difference between my own central bank, and the current central bank, is that mine would have massive equity instead of massive liabilities. A superficiality only. The mechanics of money creation could work the same way, despite the fact that my company would have 3 trillion in equity on the books.
Well, there’s one more difference. My central bank would be slightly better managed. Slightly. (Bernanke knows what he’s doing, but plenty of other people on the FOMC don’t.)
Hellestal, you say your equity is $3 trillion, but why stop there? Why not $9 trillion? Or a trillion trillion? I don’t want to belabor the point, which I’m sure you understand, but if you can make gold-money in any amount, for free, there’s really no point in putting a number on the value of your stake. The value is whatever you want it to be. Or rather, it is limitless. At least in terms of its value in gold-money. It’s infinite.
And it’s value, should the technology escape your hands, is 0. Because people won’t pay you for it, if they have the technology themselves.
From your point of view, as the owner of infinite gold, the value of all that gold lies only in what it can buy. You could, if you wanted to, create mountains of the stuff, and count it, and call it your net worth. But since you can make it in any amount, whenever you want, there’s no point in storing it up, is there? And in fact, there’s no point in counting it. If you lose it, you can instantly and effortlessly make more.
Except that you wouldn’t want to do that - lose it, I mean. It’d be ok if it simply disappeared. That’d be no loss to you. But you wouldn’t want people to be able to be able to help themselves to it, when you weren’t looking. Because people won’t pay you for something they can take for free. In fact, the more of your gold leaves your hands, the less your gold-making power is worth. You could even say that every gold-dollar that escapes you reduces your buying power, by one gold-dollar. Because what you can buy with gold-dollars is not infinite. And neither is the rest-of-the-world’s demand for your money.
So if you want to keep track of things, one could argue (I would argue) it’d make sense to count each-gold dollar that left your hands as a dimunition of your buying power - a debit. A liability against the value of your gold-making machine. Because, as I said, while your power to make gold-dollars is infinite, the rest-of-the-world’s interest In buying them from you is limited.
Now, on the other hand, should you buy something with gold-dollars, it’d make sense to count the thing you bought as an asset. Because, what else could it be?
Now there is a caveat to all this. Or several of them. One is that you couldn’t count a thing as an asset, if you consume it. If you buy a hamburger with gold-dollars, and eat it, well, you can’t really count it as an asset anymore.
So, if for whatever reason, you wanted to keep a balance sheet, and show people your assets were at least equal to your liabilities, you wouldn’t want to buy only hamburgers. You’d want something that didn’t get eaten up, or decay, or depreciate. In fact, if you wanted a “profit” (so you could buy some hamburgers) you want something that paid interest. MBSs, for example, or government bonds.
And here’s another caveat. If you buy government bonds, you’re doing something sort of strange. You’re exchanging something you can create, in infinite amounts, whenever you want, for something the other party (the government) can create, in infinite amounts, whenever it wants. And you are (if you’re the Fed) actually part of that same government!
Which makes the whole thing sort of strange.
One could even argue, I think, that the distinction between Federal Reserve liabilities (dollars) and Treasury liabilities (bonds) is not all that significant. Yes, bonds pay interest, and dollars don’t. And dollars can be spent immediately. But that is the difference between an interest-bearing savings account and a “free” checking account.
If the Fed were to purchase all the Treasury bonds - or at least all the ones people were willing to sell - the effect would be that of moving the money from interest-paying savings accounts to non-interest paying checking accounts.
Maybe I just don’t understand buying US debt as an inverstment option but isn’t the goal to buy low and sell high? If the interest is so low then the price must be pretty high right? Where can the price go except drop?
Hellestal - Let me comment on what I said before. Official bookkeeping shows government bonds as liabilities against the government, and federal reserves as liabilities of the Fed.
But if you consolidate, counting the Fed as part of the government (which it is) the Treasuries disappear (because they represent money the government owes to itself), leaving just the liabilities. You’ve said that the liabilities of the Fed - reserves and currency - are not real liabilities. I agree with that. It’s just fiat money. All the accounting rigmarole is legerdemain designed to cover what’s really going on - the government is printing money. Just like it always has, and as it must, in order for the economy to function.
So to be clear, where we differ on this is that you focus on the Fed as the prime mover. I think it’s the Treasury - when it issues new bonds it creates “money” in the form of what is effectively savings accounts, with no corresponding real liability. The Fed is part of the process, but the Treasury (I’d argue) is the prime mover. The Fed takes something that was already pretty liquid (US government debt) and turns it into something that’s perfectly liquid - reserves and currency. That’s not insignificant. But in a low inflation/low interest environment with high unemployment, you need to look to the Treasury to restore lost savings. Not the Fed.
Hellestal - Thanks for your hard work, and please feel free to respond at your leisure. (Or not at all, as I didn’t locate any serious errors in the argument.) I’ll quote the following, if only because it was such a pleasure to read.
I’m inclined to think that main street responds more tepidly than Wall Street and the bond market. But I can see that the process should occur anyway. Moreover, I hadn’t earlier made the conceptual link between “Anticipated inflation” and “Hot money”, at least not in the manner described.
Caveat 1: I think it’s reasonable to hypothesize that a 4% inflation announcement by the Fed might not be sufficient to overcome the sorts of structural problems occurring in the midst of a full blown financial crisis. But this applies to 1931-33 and 2008-09, not so much in 2012. I can imagine a battle between the facts on the ground and the Fed’s words, with the latter losing the contest.
Caveat 2: IMHO. 4% anticipated inflation won’t make anybody buy a plasma TV. 8% might though. And a 4% Fed announcement would arguably be sufficient to make a CEO/CFO sweat a little. Which would be sufficient to jump start the process. Plausibly.
Caveat 3: If this process occurs, it should be measurable after the fact.
Caveat 4: I’d like to know where the growth is going to come from if it doesn’t come from residential construction. (Arguably, it would come from exactly that though if done in 2012-13.) It’s my understanding that the effects of monetary expansion are not uniform across SIC codes. No citation, just an impression, alas.
Caveat 5: I’ve wondered whether there’s an odd discontinuity between paying 0.25% on excess reserves and paying 0%.
Caveat 6: I noticed you didn’t mention the stock market. Good to hear, as I understand the effects of equity prices on corporate investment are modest, though not zero.
Observation: Banksters and straight-laced types tend to dislike inflation. But in this story it truly does seem to be a necessary evil. Aggregate demand frameworks don’t make the same use of it. Can this story work with a 1-2% PCI inflation “Comfort zone”? I can’t see how.
What I’d really like to see are empirical treatments of the above. Which is hard to present as we’re discussing unconventional monetary policy that has never been tried c. 2011, AFAIK. And if current efforts don’t work, monetarists could fairly note that nominal GDP was not targeted, nor was inflation pegged to 3%, never mind 4%. Just ask Kocherlakota, who promises cold feet if inflation tops 2.25%.
I should have just numbered the above: never mind the word “Caveat”. Sheesh.
Overall, Hellestal isn’t presenting an “Growth Expectations” story so much as an “Inflationary Expectations and Hot Money” tale. I think there’s a difference. Imagine you are a Vice President of a bank, setting loan standards. If you set them too low and a downturn hits, you could lose your job. But if you set them too high and there is a wave of prosperity… you could also lose your job. Wall street pundits talk about the stock market’s balance between fear and greed. But in some cases fear can be on both sides of the ledger.
Now imagine the Fed has a 1-2% PCI comfort zone and an implicit 5% nominal GDP growth target. It’s early 2009 and GDP is growing by 4% and falling. You know the Fed will loosen. But weighing costs and benefits, it would make sense to retrench: the risks of underestimating growth aren’t high when you believe the Fed will keep inflation capped below 2.5%. But create a possibility of 3-6% core inflation and the calculation shifts. Targetting GDP levels could build such fears of setting loan standards too high into the job.
Hoo-boy. Creating a mean expectation of 4% inflation would be politically costly. I’m not saying it couldn’t be done: recessions are costly as well. But the serious-but-innumerate pundit crowd will stroke their chins, declare their concerns and preen for greater austerity. These aren’t objections: they are however complications.
If most of my balance sheet is non-gold assets, which it would be, then there is every reason to put a number on the value of my stake. It’s what my company is worth if the machine breaks or I somehow can’t repair it.
The very same thing is true if I want to be the world’s only gold mine, rather than a central bank. Would you seriously recommend I start marking the gold I sell as my liabilities just because the gold on market next year will be priced less than the gold I’ve already sold? “I profited by putting another $100 billion of gold in circulation this year. Oh! I should remember to mark that gold I’ve sold into circulation in my liabilities column, because the same amount will only be worth $93.7 billion when I sell it next year!” Is that the argument you’re making? Because there’s not an accountant in the world who would actually suggest such a thing.
If I need to keep track of how much gold I’ve put into circulation (and yeah, maybe that would be nice), I can keep a running tally on the back of some envelop in my office. I can open a spreadsheet. I can ask the Wall Street Journal. I can phone the London dealers. But that information does not belong on my balance sheet.
Gold possessed by other people does not belong on my balance sheet.
You already know this, but I’m going to to emphasize this point yet again because it’s important: It is not a liability. I don’t owe them anything. They can’t demand interest from me. They can’t demand coupon payments from me. They can’t demand that I redeem their current form of money for something else. They can’t request that I roll-over the obligation for another month, because there is no date of maturity and no obligation anywhere here. There is no obligation for alchemied gold, just as there is no obligation for monetary base.
You might think a different way of accounting might be personally convenient to you if the alchemy machine were in your possession and the central bank under your personal control, and that is totally okay. But it is nevertheless a personal convenience that is contrary to the standard meaning of the word.
If savings accounts were insured by the government past the FDIC limit, that might be a better analogy. If you could only cash out your savings account if someone else bought into it, that might be a better analogy. If the total amount of savings accounts was fixed by the federal government, that might be a better analogy. If the dollar balance of savings accounts changed overnight based on a change in world market expectations, that might be a better analogy.
However, none of those things are true. Which is why that is a completely terrible analogy.
The Fed is the prime mover of the nominal economy.
There are many other real things out there in the world that really matter, but nominal rigidities in the markets exist, which means the nominal economy matters, too. The monetary authority has a very narrow range of influence, but it is stupendously powerful, Old Testament wrath of god powerful, within that narrow range.
Government bonds are a real liability. The issuer has clearly defined obligations. If the issuer does not meet those obligations, there is default.
You can argue that the Fed wouldn’t allow that to happen by issuing its own paper to pay the debt, which means the liability is effectively the same as the liability of money: none. The problem with that is that they don’t control the price of bonds in the market. If they are going to pay this liability with brand new dollars worth less than old dollars, then the price of the bonds drop because they are no longer meeting their original obligation. The obligation exists, and a failure to meet that obligation can causes the price to change immediately. This is one of the chief differences I pointed out above at some length. Money is always worth what the number on the paper says it’s worth, but the same is not true of bond. It changes its prices almost instantaneously in response to expectations.
The next step the government can take is that all the bonds could be paid with this new money. But if that were to happen on mass scale, then the potato becomes nuclear reactor hot. You’ve just exploded the money supply, with no way to remove it from circulation. The potato core melts down.
It’s the government walking up to me with guns and telling me to run my gold machine or else. It’s disaster.
This is possible. It happens sometimes. It is not advisable.
The Fed does not have to buy government bonds.
Right now the Fed is buying other securities, not government bonds.
I do not even have to make a hypothetical here. They are already acting in a way contrary to your “Treasury is the prime mover” theory. They can make more money on their own. They don’t have to wait for other people to act first.
This is the single most interesting macroeconomic question in the world right now, for me personally.
All true, but I still think the politics here is a lot easier than the politics of fiscal expansion which is a main reason why I talk so much about it. A president can appoint the right people and then act surprised by the results, even as they smile behind closed doors about the unemployment rate going down much more quickly.
Agreed that you should value your non-gold assets as assets. That would be the market price. I’m asking about the value of your gold-making machine. If you value it at the amount of gold it can make, then its value is infinite. If you value it at the amount you can buy, then its value is finite. At any point in time, people are willing to give up only so much of their real world productivity in exchange for your gold-money. The amount is less than all of it - probably substantially less. If you opened the doors and let everyone have all the gold they wanted, at some point they’d have all they wanted. Then your machine would be worthless. So I’m arguing that the value of your machine is not infinite (even though it can produce infinite gold) but finite. If demand for gold-money is 100 in March, and you sell 5 on April first, it follows that the value of your machine has declined by 5. You’d debit your balance sheet by 5, therefore, while crediting it with the value of whatever you bought with those 5 coins.
Now the Fed doesn’t actually calculate the value of its gold-making machine, AFAIK. And it doesn’t list it as one of its assets. So it’s, in effect, an off-sheet asset. But whatever its value (many many trillions, I would think), it’s marginally reduced by the value of the dollars the Fed produced.
No. I’m arguing the value of the stuff you can buy is marginally reduced by the value of the stuff you have bought.
But that’s how central bank accountants do central bank accounting, isn’t it? They mark the dollars as liabilities and the things they buy as assets.
I agree: dollars are not liabilities on the Fed in the sense of being a debt. They can only be exchanged for other dollars. The Fed doesn’t have to give you gold or goods and services for them. I’d only argue that the way the Fed treats them on its balance sheet makes logical, consistent sense. It could do things differently, if it wanted. It could get rid of the “liability” side of its ledger altogether. Or it could try to estimate the value of its gold-making machine, and debit dollars against that asset.
But what it actually does makes sense, and is consistent with how banks have always done things.
The only drawback is that it gives some people the false impression that central banks have to maintain a net-positive balance sheet, and that something dreadful would happen if they didn’t.
A better analogy would be between checking accounts and certificates of deposit.
I’m following Linus’ 2 threads with interest. As per above, don’t worry about immediate replies.
Heh. I was alluding to Fed politics, I think. Their secondary directive is price stability, their tertiary directive is long run potential to increase production and promote effectively the goals of maximum employment, but their prime directive is to maintain independence within the Federal government. They absolutely will not announce a 4% inflation target this year (though they should IMHO, or at least a 3-4% band). They could just barely plausibly do so in 2014, after they had laid down a lot of groundwork.
1b. OTOH, Fed governors have fixed terms anyway. Oddly, I suspect many of them wouldn’t mind saving the world: either way they have a cushy academic appointment in their future. And central bank independence has a somewhat stronger international consensus behind it than it did during the 1970s say.
For years I believed that we could never experience another Great Depression, because we understand the appropriate policies and any administration that failed to act would get thrown out of office. I was wrong. It’s in the interest of the minority party to cause the current administration to fail, and in the US it is almost child’s play for them to obstruct without accountability. Both sides understand the appropriate policies now, but they share no interest in carrying them out. So I agree that the efficacy of unconventional monetary policy is a very salient question.
Drilling down, the optimal real rate was something like -6% during 2009. So a 4% inflation announcement can’t possibly send us rapidly to full employment. It might however turn depression to recession though again this is an empirical matter.
We might be able to learn something from current conditions, as assurances that the Fed won’t raise rates until late 2014 assuming inflation remains under control… GAH! provide some sort of test. This arguably warms up the M2 to some extent. Though Kevin Drum has pointed out that some of the recovering numbers precede the Fed announcement. You just don’t get clean experiments in economics. Frustrating.
Having another policy lever would be a boon for humanity, regardless of the politics. I don’t get off on $800 billion stimulus packages, never mind $1.5 trillion ones. If smaller ones are possible or if they can be skipped altogether, so much the better.
That sounds like a liquidity trap argument. (Treasury bonds are always liquid, though now their interest return is close to cash. So they are closer substitutes, except cash has no interest rate risk.) As I understand it the response goes like this: someday there will be a favorable economic shock. If it happens, the Fed will accommodate it (or so they are announcing). This announcement makes cash in pocket heat up somewhat, and increases the desire for non-cash assets like machines, research, equity and non-government bonds. Or that’s what Krugman argues anyway, if I understand him correctly: nonconventional policy gets traction via the possibility that a future shock will be accommodated.
Money also changes its real value, because of inflation. It’s not subject to changes because of interest rates, because it pays no interest.
Whenever the Treasury pays off debt, it does so using Fed “paper”. When the Fed buys Treasuries, it’s issuing new paper to cover the debt. It’s standard operating procedure.
As far as the Fed not controlling the price of bonds in the market, I’d argue they do control it. They can purchase, if they want, any amount of bonds in the market. They control bank reserves, which means they control the interest rate at which banks can borrow and lend, and therefore the risk-free interest rate. Treasuries are the safest investment, so their rate is the lowest. Treasuries go up, not down, when the Fed issues new paper.
The Treasury’s obligation is to pay, in dollars, the price printed on the bond. It has no obligation to pay with old dollars rather than new dollars, or (except for TIPS) to pay with inflation-adjusted dollars. And if fact, investors don’t get to control the risk-free rate of interest - the Fed does that. And why should they? If they want a better rate of return, they should take some risks.
My point from the beginning has been that that doesn’t fit the literal meaning of these terms as we typically define them. Central bank accounting is different.
This is consistent with history.
It’s not consistent with what these words typically mean when we’re talking about other enterprises.
There is a extremely high degree of arbitrariness in modern central bank accounting, and in my opinion, the current label tends to blind some folks (not everyone, but some) into believing things about central bank accounting that aren’t quite true. The Hellestal Central Bank would do things differently, more in spirit with the literal meanings.
It’s a matter of personal judgement which you consider to be a more useful and descriptive system. I’m only making the argument that my way would work just fine.
You’ll have to pardon the lameness of my reply after a long week, but I can’t see how that could possibly be considered a better analogy.
I pointed out four notable differences: bank deposits aren’t insured past a certain limit, total bank account balances aren’t constant such that no one can cash out without a new person taking it over, total bank account balances aren’t fixed at a certain level by federal choice, and bank account balances don’t change instantaneously in responses to changes in economic expectations. These are all notable differences and they add up. Now you offer as a new analogy CDs, which not only fails the previous four criteria I already laid out, but fails yet another test in that it is not as fully liquid as bank account balances or bonds.
We need to engage directly with the differences, rather than unhelpfully exaggerating similarities that don’t really exist.
Our desire to possess money is absolutely subject to changes in interest rates because higher interest rates make holding money more expensive, all else equal.
As for inflation, I’ve already discussed that at some length in post 22. Again, I don’t intend to be too lame here, and that is not a short post, but to get a better response from me, you need to engage directly with the differences I’ve already pointed out.
The short version is: prices don’t change instantaneously for everything else in the economy. For some things, yes. For most other things, no. That means even in a relatively high inflation environment, money provides an opportunity to maintain its present purchasing power if purchases are made today, before prices change, rather than tomorrow when all the goods have new stickers with high numbers. Bonds are completely different in that respect. Yet another difference.
There’s a whole yield curve out there, and there is no possible way for a central bank to control all of it. It’s like pushing on a balloon. You can’t push down on one place without pressure building somewhere else. They are normally very careful and only push on the short-term stuff, the very safest place to push.
They cannot possibly purchase any amount of bonds in the market. This is whole reason I stepped into this thread in the first place. If they put their thumb down way too hard on the yield-curve balloon, the whole thing pops. The whole system comes apart. As soon as they lose a big enough chunk of their balance sheet, they lose their ability to undo the choices they previously made. That by itself is an important constraint on their present actions.