Hello dear reader. Just two days after we last talked in early March, where I explained what a bank run is, guess what happened? A bank run, a big one that caused the failure of SVB! I promise I had nothing to do with it. If the world decides to speedrun through all financial catastrophes explained in Tailwit Capital, such that we have constant fodder for real-life examples in our humble newsletter, then who am I to stop that?
(By the way—last issue I used the analogy of planes landing or crashing, to represent the US and world economy. But did you know that heavy turbulence on actual planes—the horrible kind that makes you think you’re all gonna die, and comes with a real risk of injuries to passengers—is likely to continue getting worse and more frequent, thanks to climate change? Just a fun fact of the day for you.)
Here is a question. Do you ever wonder how it is that one day, a certain old dude says in the news that he and his crew decided that interest rates should be higher, and the next month, your mortgages’ interest rates are higher and your student loans’ interest rates are higher and your friends are all getting laid off (but for once, your savings account is paying more than 1%!)? How did that happen, exactly? Did he pass some sort of law? Are all banks just working for him, somehow? Dude seems powerful.
In Part 1, we talked about what the Fed is and why it came into existence, how the Fed is governed, and what its purpose is. In this installment, I want to talk about how exactly the Fed goes about doing one of the main things it is said to do, which is to “set interest rates.” But along the way, we’re going to have to talk about what interest rates actually are, what factors affect them generally, what Treasury bills are, how bonds are priced, what overnight loans have to do with all this, and finally, what the Fed actually literally does to influence interest rates. Plus, in the course of all the above, I promised to explain my claim from last issue that the Fed prints money.
Maybe you’re familiar with some of the above topics, especially the earlier bits, but if you keep reading, you’ll probably find something in here that you didn’t know. (Or, if you read this issue to the end and don’t find anything you didn’t already know, I want to get coffee with you! Email me!)
I’m aware that this issue is incredibly long and incredibly dense. But you’ve been waiting for months for this installment, and if we’re being realistic, I probably won’t get around to finishing the next installment for the next several months, so you have plenty of time to read this one again and again until it makes sense (or just because you miss Tailwit and it’s your comfort reading?!), or to read it a little bit at a time. “Reread a blog post / newsletter email?!” you say with an expression of undiluted horror. Yes. Sometimes we can read things more than once, because it helps us learn.
(Also, if you get stuck on anything and would like me to try to clarify, or if you otherwise have questions, please write in! You can reply directly to the email, if you’re a subscriber.)
Here’s the crux of it all. The Fed doesn’t actually print paper money, as in green dollar bills; that happens at the Bureau of Engraving and Printing. (Although it turns out the Fed actually does decide how much paper money to print, but that’s more of a logistical thing than a financial one, and not what we’re talking about here.) The Fed prints another, more powerful kind of money: it prints magic money. How it does so, is the subject of this issue. Why it does so, will be the subject of the next installment, and of future installments to come.
Where do interest rates come from?
Most people are vaguely aware that the Fed does something around interest rates. But what are interest rates? It’s often said that interest rates are the cost of money, but I think it could more fittingly be said that interest rates are the cost of time. Money exchanged for other money, in the same currency, traded in real time, is just worth its face value: I could give you $5 and you could give me $5 at the same time, and we would immediately be even, so there would be no point to our exchange. But if you want money today that you don’t have, and you want to pay it back tomorrow, then that costs a little extra money just to get it in advance. So as you’re returning the money, you’ll have to return a little extra, generally expressed as a percentage of the amount borrowed. If the lender didn’t need that money at all in the meantime, they can take all of that interest as revenue.
The question is: How much extra? How much money will it cost you to borrow the money? Like anything in the free-ish market, it costs whatever the lender wants to charge, that they can get a borrower to pay. Sometimes regulation limits these rates, to keep them from becoming extreme. But in unregulated markets, such as under-the-table cash loans from your local loan shark, the rate really is whatever the market (and/or the threat of violence) allows.
But what determines what the lender wants to charge or what borrowers are willing to pay? Put simply, supply and demand. It might be weird to think of money itself as being subject to supply and demand, since money is what you pay for other stuff with. Then again, when you borrow, you are not exactly buying money, but buying control over when you have money, which is almost a kind of luxury service, the way I see it.
More specifically, the things I can think of that affect supply and demand of money available to borrow are competition (within the same class of “products”) and opportunity cost (outside the class of products). Competition, meaning that as a lender you wouldn’t want to ask too much more than other lenders are asking for the same product, and as a borrower you can’t demand rates too much lower than other borrowers are willing to pay; otherwise a loan won’t happen for you at all. Opportunity cost, meaning that if a lender can make the same rate of return in a different way that’s “easier,” then they have less incentive to lend their money.
“Easier” has a specific meaning in this context. All loans come with the risk of default, which is when the borrower can’t pay part or all of the loan back, so the lender just has to write it off as a loss. Professional moneylenders factor this risk into the interest rates they charge, so that they can still make money overall if the expected percentage of borrowers default.
For professional moneylenders, the possible ways of making money with their money that are available to them typically have different levels of risk, which correspond to different rates of return: the higher the risk, the higher the reward. The idea is that they are supposed to do math, do research, call on their gut instincts, ignore their gut instincts, or any combination of the above, to convince themselves that when they choose to utilize a certain way of making money, it is optimal in terms of the risk/reward tradeoff, and that there are no deals available to them that are flat-out better than what they’re doing (higher reward for the same or lower risk would be better; the same or higher reward for lower risk would also be better). And, just as in all kinds of shopping, bad deals are everywhere and always will be—but you don’t have to take them. That’s the idea: to identify the good deals, and take those. Of course, some professional moneylenders are better at this skill than others. That’s a lot of what separates the winners from the losers in the finance world.
So let’s say you had been lending out money at a 4% annual interest rate—for mortgages, just to make this example more concrete—with some small risk of defaults, but nothing too crazy. If you found out there was a less risky way to make 4% annually on your money, you would want to immediately do that instead, and only offer mortgages if you can get a higher rate for them that you can’t get in the less-risky way. The less-risky option, because it’s more attractive (a better deal) overall, presents a form of opportunity cost that, if available to lots of lenders, will divert money out of the pool that would otherwise be available for lending out at 4% for mortgages. Borrowers across the entire affected market will have to settle for loans with higher interest rates, thanks to the presence of more attractive ways for lenders to make the same profit.
What if the above happened on a national scale? Not just in one region or for one type of loan, but affecting all regions and all forms of borrowing? What it would take for this to happen is the entrance of a way to make money with your money that a) unquestionably carries lower risk than most other things you can do with your money; b) offers interest rates high enough to compete with most other investments; c) is widely accessible to all kinds of different lenders/investors (as opposed to a small pool of insiders); and d) has a supply big enough that everyone who wants to buy into it can do so.
Picture it. Such an asset would immediately dominate the whole market for investments or loans that get you the same rate of return or lower. Instead of putting your money into riskier things, you’d just buy as much as you could of this asset, hang onto it, and reap the interest payments. And everyone who has money to throw around is thinking the same. As a result, lots of ventures that are comparatively too risky and/or are not rewarding enough won’t get funded. People trying to get a mortgage to buy a house will have to pay higher interest rates to get lenders to take them seriously, because all the lenders are too busy buying this awesome asset. So those people may decide they can’t afford a house now, and will wait until later to buy one. Same with people who need a loan to start a business, or keep a business running. And so on.
Now, what if that super-asset suddenly went away, or lost at least one of its key qualities, making it no longer a super-asset? You’d have to go back to investing your money into riskier things if you want to make a return—and if you’re a bank or investment firm, you must make a return, you can’t just hold cash and make zero returns if you want to keep your job. So more ventures will get funded, people can start business and buy houses, etc. If this asset is widespread enough, changes in its availability or the attractiveness of its terms can contribute to the economy booming or contracting.
Enter the Treasury bill
There is an asset that can, at times, meet all of the above criteria: US Treasury bills, or T-bills, or Treasurys. (Yup, it’s spelled “Treasurys” not “Treasuries”. This is the kind of thing that speaks directly to the spirit of Tailwit Capital: the duncey-looking spelling is in fact the one that true insiders use. Anytime I see a “Treasuries [sic]” it’s a thing of beauty.) They are also sometimes called Treasury bonds or notes, but don’t worry about “bills” vs. “bonds” vs. “notes”. When you buy these, you’re effectively loaning money to the US government, which it agrees to pay back on a fixed date N weeks or years into the future, and with annual interest payments at a fixed interest rate. This asset is widely considered the safest (lowest-risk) asset in the world (other than USD cash or bank deposits), because it’s guaranteed by the US government. It’s of course contingent on the ongoing stability of the US government and economy, so its safety depends on the degree to which everyone in the world continues to believe in that.
(Your bank is probably holding a ton of Treasurys anyway, so to the extent that it is doing so, they are similarly safe. Except that your bank account comes with FDIC insurance, which is good; on the other hand, Treasurys you hold directly are backed directly by the US government and don’t rely on any bank’s survival, which is also good. So both direct Treasurys and bank deposits are relatively safe, for different reasons. Just something I think it’s good to know in our present day of collapsing banks!)
So, Treasurys are considered the lowest-risk asset that money can buy. (Actually, they are often called “risk-free”, but calling anything “risk-free” freaks me out, so let’s stick with “lowest-risk”!) They are always in supply, as the US government is constantly issuing new bills, plus there are always people willing to sell theirs on the secondary market. And they are widely accessible for investors to buy: as an individual/small investor, you can buy them directly from the Treasury with only a $100 minimum, or you can buy them from the secondary market through retail brokerages. But they were not particularly attractive to invest in for most of the 21st century up until the past year or so, because they didn’t meet the 4th criterion I mentioned above: they had really low interest rates that wouldn’t be competitive with most other investments. So for most of this century, it was more like the second scenario I described above, where the super-asset isn’t available. People preferred to take on more risk for a significantly higher reward than what they could get with Treasurys. But right NOW, in 2023, the Treasury bill is a super-asset again.
But whether its interest rates are low or high, the humble, unassuming Treasury bill is the linchpin of the entire US economy, and of everything the Fed does. That’s why we’re going to spend time going over everything that’s relevant to know about them: how their interest rates get set, what determines their prices on the secondary market, how they are bought and sold, who is buying and selling them, what they have to do with the Fed’s interest rates.
Short digression: How to lose money on bonds
Bond pricing is a really weird beast, especially if you’ve never thought much about it before—as I hadn’t, before writing this. (For the purposes of this section, let’s consider Treasurys to just be bonds.) (Let’s also consider bonds to just be like loans that you can freely buy on the market.) Bonds aren’t like stocks, where your returns are correlated with how well a business is doing or how well people think it’s doing, and can swing wildly based on changes in those beliefs. With a bond, you know exactly the amount of money you are being promised, all in advance, including all the exact dates you’re supposed to be paid. If you “hold to maturity” (buy a bond, hold it, and don’t sell it until the entire term is over), you just get all the principal back, so you cannot lose money. So why would there be a “market” for a bond, where some people are willing to pay more or less than others for the same amount of future money? Why wouldn’t I just pay $100 now for $100 + interest later, and why wouldn’t everyone else?
The answer is that the real value or worth of a fixed number of future dollars totally depends on what exactly the future looks like. For one, it depends on inflation. If I buy a 30-year bond, but inflation over those years outpaces the interest I get from the bond, then I am losing money, in real terms. 30 years is a long time, inflation-wise. Do your parents or other older people like to tell you how much they paid for gas or milk in the 90s? (Or maybe YOU are that parent. Hello!) Today’s dollars can buy laughably little in comparison to 90s dollars. If you locked up some of your dollars in the 90s then you better have been promised a LOT of 2020s dollars. That’s inflation risk. Will 2050s dollars be worth more or less than 2020s dollars, in purchasing power? How much more or less? Nobody knows. So everyone will have a different idea of how many of today’s dollars they’d be willing to pay to get a certain amount of 2050s dollars. (And likewise for 2030s dollars, 2028 dollars, 2025 dollars.) That’s part of why there is a market for bonds, where they can be “auctioned” at different prices.
It also depends on what else I could do with that money in the meantime. If at some point the Treasury starts offering similar bonds with way better interest rates, that won’t cause me to lose money as long as I don’t sell mine, but it sure feels bad, because then I’m stuck with my alright bonds while everybody else can buy the new better bonds, and will slowly get richer than me while doing the same amount of nothing. If I’m just an individual investing my own money, then maybe that’s not a big deal (I’m not THAT competitive).. but if I’m a professional investor, then it is kind of a big deal, because then it’s my job to make as much money from money as my competitors are able to make, otherwise I’ll lose my job or my clients.
But this situation of being stuck with mediocre bonds can get really bad for me if some other event forces me to need cash immediately and I can no longer wait 30 years or however long it is, so I have to sell my bonds before they reach maturity AND while there are lots of better bonds currently on the market. In such a market, no one will want to buy my bonds for the same amount I paid for them; I’ll have to sell them at a heavy discount (aka, a loss). So then I neither get the future promised interest rate payments, nor do I even get my original money back. I just lose money. The risk of this situation is called interest rate risk.
(Aside: This is why SVB collapsed. It held a lot of its money in US-government-backed bonds that would’ve been fine IF they had been held to maturity, but were worth way less IF they had to be sold immediately, just because of the state of the market for those bonds in 2023. Since somebody pointed that out in public, the bank’s customers got nervous and started to pull their money out, which caused SVB to run out of money, even if it sold all its bonds. It was a self-fulfilling prophecy—if no one had pulled their money out then they would’ve been fine! But, it’s kind of a bank’s One Job to make sure it can survive a bank run, so the solution is to never get into such a precarious position in the first place. Plenty of other banks are, or were, in exactly the same financial position, by the way; it’s just that nobody has called them out about it, or started a panic, which is the reason they are still standing. Comforting, huh?)
So: There is a market for stocks because people hold different beliefs about what the future looks like for the company issuing the stock. There is a market for Treasury bonds because people hold different beliefs about what the future looks like for money itself. (As for corporate bonds, it’s some hybrid of the two, I guess!)
This ties into why I think bond prices are weirder to think about than stock prices. A share of a company only has one kind of dollar amount associated with it, which is its market value, i.e. the price it is selling for on the market. It represents a certain fraction of ownership of the company, which is the thing you’re buying. But a bond has two dollar amounts associated with it: its face value, and its market value. The face value is the amount that the bond issuer (borrower) promises to return, and the interest rate is calculated from that face value. (So, in loan terms, you could also think of face value as the principal.) If you buy a $100 bond at face value, you pay $100 and you are eventually owed $100 plus interest on $100. But later you can sell that bond on the secondary market for whatever price you can get for it. If you happen to be holding a bond that is now more attractive to people than when you bought it, you can sell it for more than $100; if the opposite, then you can sell it for less than $100. It’s like any collectible that can gain or lose value and gets auctioned off on eBay, except that the collectible is “future dollars.” (And if you try to auction it on eBay, you’ll probably get arrested.) So as a buyer, it’s possible to buy at a discount and, say, pay $95 to be owed $100 plus (the remaining) interest on $100. It sounds pretty weird to say (nay, to brag!) that you paid $95 to lend out $100. But that is in fact what’s happening.
The fact that there are two different dollar amounts is also the difference between interest rate and yield on a bond. The interest rate is the % you gain against the face value. The yield is the % you gain against what you actually paid for the bond. (You can think of it as a difference in what you use for the denominator of the fraction, if that helps.) If you paid exactly face value, then your yield will be the same as the interest rate. If you paid lower than face value, then your yield should be higher than the interest rate, and vice versa. (The math is a bit more involved than that, because the price needs to take into account how “far along” the bond is, i.e. how much of the total interest remains to be paid, and also because shorter-term Treasurys have the interest “built in” to the face value, but don’t worry about it.) High demand for a bond means higher prices means lower yields. Lower demand means lower prices means higher yields.
Meanwhile, back at the Fed…
Ok. Treasury bills come with interest rates when they are issued. The Fed, or FOMC, votes on interest rates—so we hear from the news. So the Fed decides on the interest rates for Treasury bills, and that’s how they influence the interest rates of all assets throughout the economy, right? Nope! Oh, if only it were that simple.
Treasury bills actually don’t come with predetermined interest rates. They come with a face value and a term (the time interval until maturity), and then the interest rates are decided by our two best friends, supply and demand. New bills come out for auction every week. There is a group of big institutions with lots of money, known as primary dealers—including some names you surely know, like your JP Morgan, your Wells Fargo, your Goldman Sachs, and some you probably don’t know—and these institutions place bids with the interest rates they would accept for these bills. The outcome of the auction is what determines the interest rate on that type of bill, for everyone.
In fact, it’s not true that the Fed sets any interest rates; not directly, anyway. What they vote on when they get together, is a target for the (short-term) interest rate (this is the interest rate for the shortest loans possible, which are overnight loans—yes, literally loans that are due the next day, on which more below); in other words, they vote on what they would like the short-term interest rate to be, which they will then try to manifest by indirect means. What means? That is where the weird gets weirder.
Overnight loans are a type of loan that banks give to each other, so that they can all settle their books at the end of each day, and no bank ends up in trouble just because they don’t have enough money on hand today but will have it by tomorrow. The interest rates on these loans depend on how much money is floating around in the system in general: how many banks have excess money to lend, how many are strapped and need to borrow money. Since the duration is so short, uncertainty/risk should be super low, and the interest rates are accordingly super low, so the interest rates on these loans are sort of the baseline for the rates on all other loans.
How can the Fed change the interest rates on these loans without directly setting the interest rates? By changing the thing those rates depend on—the supply of money! Okay.. but how? There are a couple of ways, but the one that the Fed used predominantly for many decades is called open market operations. (I think this may not be the primary method the Fed uses these days, but it is still one of the main methods and so is important to understand.) Here’s how it works, for the case where the Fed wants to lower the target interest rate by increasing the supply of money in the economy:
1. The primary dealers (mentioned above) buy Treasury bills, from the Treasury or from the secondary market.
2. The Fed intends to buy these Treasury bills from the primary dealers. (Perhaps the primary dealers even bought the bills with an eye toward reselling them to the Fed, if they are aware that the Fed is looking to buy.) It will do so by going through its favorite child, the New York Fed. (Because again, it’s New York!)
3. Every Federal Reserve branch has an account with the central Fed. This account contains a balance of electronic dollars which are just numbers in a database. So, to give the New York Fed money to go and buy Treasury bills, someone at the central Fed clicks a big “ADD MONEY” button and POOF! More money appears in the New York Fed’s account. Lots more money, like millions or billions of dollars. (Ok, the button part is a dramatization—I don’t know what the actual UI looks like—but the part about money magically appearing in the account is real.)
4. The traders at the New York Fed go and hit up the primary dealers to buy Treasury bills from them, using the newly created magic dollars. Everyone is happy.
5. The new dollars go to those primary dealers (which are big banks, if you recall), who then are flush with new cash and can go lend it out or do all sorts of things with it, which makes money easier to get for banks all over the country, which lowers the overnight interest rates, which over time starts to lower all other interest rates.
6. At the same time, those Treasury bills went to the Fed, who will hold them and not resell them anytime soon. That means the supply of Treasury bills available to be sold on the markets is a little bit lower. Which may raise the prices they fetch on the secondary markets, which would lower their yield, making them a slightly less rewarding investment. And the fact that there is a little bit less of the safest investment money can buy, also means that some of the total money in the world that would have gotten invested into Treasury bills will have to go towards slightly riskier investments. This is what’s known as “pushing money further out on the risk curve.” This effect is intentional, but we’ll return to this concept, because there’s much more to say about it.
7. The above steps are repeated until the actual overnight interest rate matches the target that the Fed decided on.
8. To raise the overnight interest rate by reducing the money supply, the process is reversed: the Fed sells Treasurys instead of buying them, then when the money it gets from selling them shows up in the account, someone at the Fed clicks a giant “DELETE!” button and the money is “destroyed.” (Button is a dramatization, but the rest is true.)
So. I’ve said that the total supply of dollars in the system affects interest rates, but I’ve also said that the supply of Treasury bills (and demand for them, and their going interest rates) affects interest rates. So which is it? Well, I think it’s both! Conveniently for the Fed, and for our understanding, in the process described above both seem to work in tandem toward the same outcomes. And it’s not an accident, as I said, but more that the Fed is feeding two birds with one scone. But how much of the change is due to money supply, and how much is due to the state of the Treasury market? I don’t know man, I’m not an economist! Does one of them mostly affect the other, or do they affect each other reciprocally? These are excellent questions. Questions we don’t have the answers to here at Tailwit.
Show me the money
Reader, I have a confession to make. When I was a kid, I used to play the original StarCraft, a sci-fi strategy game where you harvest resources to build little armies and attack other bases or defend your own base. And when you play in single-player mode, you can use cheat codes. You’d type in a special phrase, and it would magically have a certain effect on your current game. One of the cheat codes was the phrase “show me the money”. When you entered it in, it would instantly grant you an extra 10,000 minerals and 10,000 gas. I was never very good at harvesting or managing the resources, you see, but I still loved to do everything else, like build my units and fight with them. So I confess that I availed myself of “show me the money” whenever needed, which may have been often. I could also enter it again at any time for another round of 10,000 minerals/gas, and another, and another. It was my escape hatch for any mistake, any moment of mismanagement or oversight, sometimes even for any normal challenge that I just didn’t feel like tackling that day.
The concept of “show me the money” at the level of the central bank of the United States is weird to think about. You and I, as reg’lar folk, have to “settle our books,” in a way—by which I mean that if I get money, it must have come from somebody else losing money (not in a bad way—somebody could have paid me, for example), and if I lose money, somebody else must have ended up with it. For us, money follows the law of conservation (from physics); it can neither be created nor destroyed. (Unless you burn physical cash—which is illegal!) It’s the same for almost all banks, which is why they have to settle up overnight via overnight loans, or risk imminent collapse. But the Federal Reserve banks, ever since the US went off the gold standard (which we’ll get into sometime soon), are not subject to the law of conservation of money. Their money in their accounts with the central Fed can be freely created or destroyed. But since the Federal Reserve banks are the only exceptions, the only way for them to move the money to/from non-Federal-Reserve banks is to buy or sell other assets. They can’t just create/destroy money in other banks willy-nilly, so they create/destroy money in their own accounts, then transact as normal with other banks. This is what is meant when it is said that the Fed “prints money.” (Is this idea of money getting created or destroyed from nowhere breaking your brain or otherwise causing your sense of reality to start to crumble? Then perhaps you need to read, or reread, The total amount of money in the world is a figment of our collective imagination!)
In short, the Treasury sells Treasurys to primary dealers (which are private banks), who then resell them to the Fed, who ends up holding the bond. Which means the US government is somehow lending money to itself (?), yet new magic money ends up in the big banks, so it can trickle out to the rest of the economy, thereby lowering interest rates for all. Which, by the way, is all the opposite of what’s happening right now. Does it make sense yet?
Appendix: TL;DR from ChatGPT
explain how the fed buys treasury bills through primary dealers, in the voice of a brooklyn mobster
—
Yo, listen up, I’m gonna break it down for ya. So the Fed, they wanna buy some of them treasury bills, right? But they can’t just go to the bank and be like, “yo, give me some bills.” Nah, they gotta go through these guys called primary dealers. These are big shot banks and investment firms that the Fed trusts to handle this kinda stuff.
So, the Fed calls up one of these primary dealers and says, “Hey, we wanna buy some bills.” The dealer then goes out and buys the bills from the US Treasury at an auction. Then, the Fed pays the dealer for the bills, usually by crediting their account at the Fed.
Now, here’s the thing. The Fed wants to control the money supply and interest rates, right? So they can also sell these bills back to the dealers if they wanna decrease the money supply, or they can hold onto ‘em if they wanna increase it. It’s like a big ol’ game of financial tug-of-war, and the Fed’s the one pulling the strings.
So that’s how it works, my friend. The Fed buys treasury bills through these primary dealers, and they use ’em to keep the economy in check. Capisce?
Refs:
The Lords of Easy Money, Christopher Leonard.
Wikipedia on Treasurys, open market operations.
Much appreciation to the excellent question-askers and -answerers on Reddit, especially at /r/investing.