> It's not just the size of Operation Gutt that is striking to the modern eye. It's also the oddity of the tool being used. Today, we control inflation with changes in interest rates, not changes in the quantity of money. To soften the effect of the global COVID monetary overhang, for instance, central banks in the U.S., Canada, and Europe began to raise rates in 2022 from around 0% to 4-5% in 2024.
It's a bit more interesting than 'the central bank sets interest rates'.
Simplified: the central bank decide on an interest rate that they want to see. By itself that decision doesn't do anything.
What happens next is that they buy and sell government bonds in the open market. The interest rate can be seen as the inverse of the price of bonds.
If the central bank wants to see a lower interest rate, they buy bonds with freshly printed money to drive up their prices, ie drive down the interest rate.
If the central banks wants to increase the prevailing interest rate, they sell government bonds from their inventory and essentially destroy they money they receive in return.
So even when the language of modern central banking talks about interest rates, they still change the quantity of money to implement that.
(This is all simplified, especially with the interest on excess reserves that was popular with the Fed for a while. And there's also repos and reverse repos etc.)
That's just one mechanism, but not the primary way in which the Fed controls interest rates.
The Fed is a large provider of short-term loans ("fed funds") to cover interbank exchanges.
It also is the lender of last resort and lends to banks directly ("discount rate").
By changing these rates, the FED can influence the rates the banks charge each other for loans, and down the line to consumers.
> That's just one mechanism, but not the primary way in which the Fed controls interest rates.
Yes.
> The Fed is a large provider of short-term loans ("fed funds") to cover interbank exchanges. > It also is the lender of last resort and lends to banks directly ("discount rate"). > By changing these rates, the FED can influence the rates the banks charge each other for loans, and down the line to consumers.
And for those rates to take effect, the Fed still has to actually make (or receive) those loans at the announced rates. They don't just magically announce a rate, and then everyone charges that spontaneously.
So it's still about moving money in and out of circulation.
(Or at least the Fed needs to be ready to make and receive those loans, and anticipation does a lot of heavy lifting..)
You are correct. Striking how many HN commentors are so often confident and yet wrong...
You are wrong. Striking how many HN commentators are so often confident and yet wrong in their assumption that everything is about the US.
> You are wrong.
Wrong about what? This is about the US Fed, in context of...well you can read the thread.
In context of a thread about Finland? Anyway, it is true that interbank lending is a primary method of controlling interest rates and this is not limited to the US. The commenter who mentioned it above, probably was talking about the Fed specifically maybe just because that's what they are familiar with.
> Central banks work this way
>> Here's a biggish counter-example, setting fiscal policy for 25% of the world's economic output
>>> Right! How could the other poster have possibly missed that?
>>>> There are other central banks. Ignorant Americans...
It's easy to see why many commentators on a US company's forum would make said assumptions.
Theres not much about the forums that makes it Y combinator specific.
Unless youre talking more broadly, in which case this is the internet. Do Tik Tok videos have to assume everyone is Chinese?
“The Fed” is the US Federal Reserve
The Finnish central bank is called the “Bank of Finland”
Yes and the parent was talking about an american companys forums, which I think is logically flawed.
When it comes to the government treasury markets by volume (which back the global financial system), it objectively is.
But that doesn't (directly) impact inflation in other countries.
Was he really that wrong? Isn't a bond just another debt instrument? It's not obvious to me, that there is any fundamental difference between the operations that both comments describe.
While the effects may be similar, they are fundamentally different mechanisms.
Also, this statement is incorrect:
> Simplified: the central bank decide on an interest rate that they want to see. By itself that decision doesn't do anything.
The Fed does in fact set interest rates and that decision directly impacts rates all down the line to mortgages and local loans. Intervening in the bond market is another tool that the Fed can use.
The Fed still has to make (and receive) loans at these rates, for them to have an effect on the market.
I think you're talking past each other.
A central bank doesn't directly set interest rates for your mortgage.
It can set rates at which it will lend to other banks, which in turn influences the rates banks will offer to mortgage borrowers, but this isn't necessarily so, see for example 2008.
Of course there are more contracts directly tied to the central bank rates, but thats just formalising the thing thats supposed to happen anyway.
Yes. Btw, I'm not sure how many contracts are directly tied to whatever rate the central bank announces these days? It used to be more common to tie contracts to eg LIBOR, which is or was a reported interbank lending rate, not a decreed one.
"What happens next is that they buy and sell government bonds in the open market."
This describes how central banks operated in the U.S. before 2008 and in Canada before the 1990s. The Federal Reserve and the Bank of Canada both set a target for the overnight interest rate, and then they hit that target by doing open market purchases (or sales) of bonds, effectively adding funds to (or draining funds away from) the overnight lending market. By changing the quantity of funds, they influenced the interest rate.
What changed is that both central banks introduced interest payments on balances that banks hold at the central bank. Previously, these balances earned 0%. With this new tool, they could directly set the overnight interest rate by adjusting the rate paid on these reserves, eliminating the need for regular open market operations.
You make it sounds like they aren't doing regular open market operations, but as I understand it... they still do?
https://www.newyorkfed.org/markets/domestic-market-operation...
The Fed still does purchases and sales, but not for the purpose of driving the overnight interest rates towards its target. The level of the overnight rate is not affected of any purchases or sales that the Fed does.
If the Fed makes or receives a loan at specific interest rates (even if only overnight), that's pretty much economicaly equivalent to buying or selling bonds.
For an international perspective: buying and selling government bonds is far from an universal mechanism for interest rate control (AFAIK when discussing central banks the US is almost always a special case)
For instance the Canadian, UK and European central banks provide systems for interbank short-term loans. It is almost entirely through these systems that they set their target rate.
For Canada the BoC doesn't do any open market operations to reach target interest rate (so almost only repos and reverse repos). Their target rate is in fact called the "target overnight rate" and only concerns overnight lending between Canadian financial institutions.
For the interested https://en.wikipedia.org/wiki/Interbank_lending_market#Monet... has more on it.
As far as I understand, the central banks intervene in this market by offering to loan or borrow above or below otherwise prevailing market rates. This has the effect of adding money to the system (or removing it). So that's pretty much the same mechanism as what I described.
They use an intermediate proxy like the 'target overnight rate' to help them decide how much money to add or remove to the system: exactly as much as needed to bring the market interest rate in line with their intermediate proxy.
Interest on reserves is very much still in place[0]. Open Market Operations haven't been a thing since shortly after the 2008 Financial Crisis.
https://www.federalreserve.gov/monetarypolicy/reserve-balanc...
Ahh…no. The Fed sets the interest rate directly. What you are talking about is yields on treasury bonds, which are manipulated via bond buying to force money into assets by artificially dropping the yield of those bonds, thus creating a more attractive investment in the stocks, assets, etc.
They were entirely wrong about mechanism of setting the interest rate through the discount and fed funds rate, but this description also isn't comprehensive. The feds buying of treasury bonds isn't just to push them down, but is also a mechanism for increasing the monetary supply through the expansion of the fed's balance sheet. This mechanism for increasing the monetary supply is also why the linked article doesn't appear to be accurate either, as they don't seem to understand that the fed does have the ability to manipulate the monetary supply through its balance sheet.
I explained this in my other reply. They buy the bonds with printed money, thereby injecting that money into the economy.
I described Open Market Operations, see https://en.wikipedia.org/wiki/Open_market_operation
> The Fed sets the interest rate directly.
So which interest rate does the Fed set directly, and how does that setting have any effect on the economy?
(I know they have interest on excess reserves. I already accounted for those in my original comment. I know, they are annoying and misguided. I was mostly talking about the system they used before, ie up until about 2008.)
The Fed directly sets the Overnight Rate, also known as the Federal Funds Rate. It's the rate at which banks can borrow from one another overnight to satisfy reserve requirements. This rate indirectly affects the interest rate of all other lending instruments because the higher cost of overnight bank to bank lending is passed on to the customers in the form of higher loan rates, credit card rates, mortgage rates, etc.
The Open Market Operations you described (completely different from the Overnight Rate) are a form of stimulus. Because money always seeks higher risk-adjusted returns, the Fed will buy treasuries, which drives down their yield. This makes them an unattractive investment, so the money goes where it can get a better risk-adjusted return. That's usually in the market. So by adjusting treasury note yields, they can stimulate the economy. Furthermore, those treasury bonds are bought with printed money, so this is effectively a way to inject massive amounts (trillions of dollars) of printed money directly into the economy. It's pretty crazy when you think about the power that these policies have.
The Fed can only 'set' the Overnight Rate by being ready to borrow or lend at the Overnight Rate. The Fed still has to stand ready to actually engage in these transactions.
As a counterfactual: we can imagine a world where the Fed 'sets' these rates by just announcing the rate but not engaging in any transactions and legislators pass a law that forbids banks from borrowing from one another at any other rates.
The interbank lending market would just not clear in that case, and you'd either have a glut of demand or a glut of supply.
The Fed sidesteps this by actually borrowing and lending in the interbank lending market. And that's economically equivalent to buying and selling very short term bonds, even if the legal form is different.
> Simplified: the central bank decide on an interest rate that they want to see. By itself that decision doesn't do anything.
I get the simplified qualifier and I see what you’re saying but it’s more accurate to say this is not true, it hasn’t really been substantially true since February 1994.
> What happens next is that they buy and sell government bonds in the open market
Again i feel it’s better to say exactly what happens, not a model or simplification. What happens next is banks move to the new interest rate without any OMO (open market operations) - the announcement effect as it’s called. There’s many reasons for this - not least that it’s futile to fight against the currency issuer.
Citation needed to support my statements above: https://www.newyorkfed.org/medialibrary/media/research/epr/0...
Yes, expectations and anticipation are very, very important in all markets, and especially financial markets.
But they only work, because the Fed can and will back up its announcements with large transactions at the announced prices and rates.
> There’s many reasons for this - not least that it’s futile to fight against the currency issuer.
You'd hope so, but not all currency issuers are so confident. I remember the Swiss central bank recently giving up on its exchange rate target to the Euro: they let the Swiss Frank appreciate against the Euro in the end, even though they control the printing presses and could always print enough Franks to drive their price down as much as they want to.
We saw similar reluctance to do whatever it takes for Japan since the 1990s and the Fed and the ECB during the 2010s, when they all failed comically to push inflation up to their targets, and came up with various excuses.
(The failure of the Fed is particularly funny, because one of the guys in charge had earlier written a piece telling the Japanese exactly how to get out of their own trap in the 1990s, but then did not practice his own preaching.)
> Today, we control inflation with changes in interest rates, not changes in the quantity of money.
That is also not quite correct, I believe. The FED can invest money created out of thin air any time it wants („fiat money“ — „there shall be money“), usually it buys government bonds to help the federal government run its deficit. Sometimes this scheme is called „quantitative easing“ which is a charming euphemism. This is what drove and still drives inflation directly and indirectly (through the effect of our fractional reserve system private banks amplify this about tenfold by lending out 90% of their customers‘ deposits).
The interest rates everybody is obsessed about are just the target rates for the interest earned or paid for for money in the account of private banks at the FED. It is just a secondary adjustment. At least during major crisis.
>> This is what drove and still drives inflation directly and indirectly
QE is fascinating.
On the one side of QE you have central banks, absolutely baffled by the fact they create all these reserves, we’re talking utterly un-relatable numbers for a human, numbers that belong in the field of astronomy rather than finance. And yet they fail to hit their inflation targets for over a decade.
On the other side you have a bunch of folks shouting loudly this amount of QE will cause ungodly amounts of inflation.
Now they shouted long enough that inflation eventually did show up but it’s not clear that it relates to QE. The onset of inflation correlates more closely with global energy supply shocks than with changes in QE policy or “printing” money in Covid stimulus. However, energy, despite being an input to almost everything we care about, isn’t really considered in the context of inflation by most macroeconomic models. The models are less than helpful in the face of “externalities”.
>> through the effect of our fractional reserve system
We don’t have fractional reserve in the USA, the UK, Aus etc and haven’t had for a number of years at this point.
> The onset of inflation correlates more closely with global energy supply shocks than with changes in QE policy or “printing” money in Covid stimulus.
Hmmm. [1]
> USA, the UK, Aus etc and haven’t had for a number of years at this point
Indeed, the US abolished the 10% reserve requirement in 2020. Crazy. I hope you guys also save in inflation hedges.
Canada (and other places) never had a reserve requirement, and they are doing fine.
Reserve requirements are mostly bullshit anyway. What you want is for banks to have enough loss absorbing capital. Reserves are almost meaningless.
What is the difference between a reserve and loss absorbing capital?
In a practical sense - reserves can’t be spent in the economy but you could withdraw your capital (accepting the impact on banking operations that would have) and use it to buy something. They’re different types of money. Reserves are only good for exchange by a few institutions and the central bank.
In a “model” sense for want of better phrasing that eludes me right now… One is infinite (via the discount window for reserves) and the other is very much finite.
In a hard legislation sense, capital is complicated, Basel regs recognise different types of capital and value them differently. But the tier 1 stuff is the equity you put into the bank to get it started or to support expanding its operations.
Yes. To oversimplify, capital is what's left after you take all your assets and subtract all your debt.
So you take all the investments that the bank has made (loans, but also their brand value, the office buildings they own and operate in, etc) and subtract all the deposits and outstanding bonds etc, and what's left over is their capital.
Another related way to measure that is to look at total market capitalisation. (But market cap makes economic sense, that's not the definition of loss absorbing capital that's used in the regulations.)
In the really olden days: reserves were physical gold or cash in your vault. (These days, it's a bit more complicated.)
About loss absorbing capital: companies usually finance themselves from two sources equity and debt. (Normal companies get their debt from their bank, or they issue bonds. Banks get some of their debt in the form of deposits.)
To give an example, supposed you have 2 billion dollars lying around, and you start a bank. You take another 8 billion dollars in debt (say as deposits), and you use that to make 10 billion dollars of loans and other investments.
If you investments gain in money, you keep the profit. Your accountant will count it as an increase in your capital, from 2 billion to, say, 3 billion. Basically, capital is just what's left of your assets after you subtract all your debt.
If your investments lose money, your capital shrinks. Say your investments are now worth only 9 billion dollars, but you still have 8 billion in deposits. Then you only have 9 - 8 = 1 billion in capital left.
There's never any money in the vault, ie no reserves, in our example. When a depositor wants their money, you sell some of your investments to cover that. As long as your total investments are worth more than your total deposits, this is fine.
(For convenience, real world banks keep some reserves around even when not legally required, so they can satisfy withdrawal requests directly, instead of having to sell some investments for every little withdrawal request.)
Obviously, the more capital cushion your bank has the larger a decline in the value of your investments you can absorb as losses, before your depositors and other creditors need to get nervous.
Back in Scotland's free banking era, when financial regulation was very light, banks over there routinely kept around a third of their total balance sheet as equity and two thirds as debt.
Today because of all the regulation we have accumulated and especially the too-big-too-fail expectations, banks need to be forced to even keep around 8% of their balance sheet as equity.
(It doesn't help that interest on debt can be paid with pre-tax money, but return on capital (ie dividends) comes largely out of post-tax money. So debt is cheaper.)
We're also currently in a period of Quantitative Tightening as we slowly undo the QE done during covid, which should have a deflationary effect. But who knows over what time period and with what magnitude.
> We don’t have fractional reserve in the USA, the UK, Aus etc and haven’t had for a number of years at this point.
What do you mean by that?
https://www.federalreserve.gov/monetarypolicy/reservereq.htm
EDIT hmm that doesn’t link directly to the relevant FAQ:
“Why did the Federal Reserve reduce reserve requirement ratios to zero percent?
For many years, reserve requirements played a central role in the implementation of monetary policy by creating a stable demand for reserves. In January 2019, the FOMC announced its intention to implement monetary policy in an ample reserves regime. Reserve requirements do not play a significant role in this operating framework.
As announced (Off-site) on March 15, 2020, the Board reduced reserve requirement ratios to zero percent, effective March 26, 2020, in light of the shift to an ample reserves regime. This action eliminates the need for thousands of depository institutions to maintain balances in accounts at Reserve Banks to satisfy reserve requirements, thereby freeing up liquidity in the banking system to support lending to households and businesses.”
You still have fractional reserve banking, even with zero mandatory reserve requirements.
Btw, banks still keep plenty of reserves around. See https://fred.stlouisfed.org/series/TOTRESNS for the US.
Not GP, but there are two senses in which I think it could be meant:
- Currency no longer has to be backed by some fraction of shiny objects,
- Banks do not need to arrange CB reserves before making loans -- they make loans and then secure the needed CB reserves.
That's still all fractional reserve banking.
Btw, in the US there's lots and lots of bank reserves: https://fred.stlouisfed.org/series/TOTRESNS
We don't control inflation with interest rates; we do some economic theatre with interest rates that some people believe controls inflation in a predictable way.
The evidence is very strong that we do actually control it, because in many countries you can see in the historical data when central bank targeting was introduced that the inflation rate drops fairly rapidly into the target band.
It's not a perfect control system because the cost is "NAIRU": non accelerating rate of unemployment. That is, economic growth and wage growth are constrained to avoid a wage-price spiral. And sometimes you get a shock from outside the system.
Please do show this very strong evidence that the effect is any more than the supply chains sorting themselves out. Even some within the CBs are doubting the causality.
Japan had the lowest inflation of any major economy post COVID, and yet persisted with essentially a ZIRP. There's a good argument that in our high reserves world, interest is actually inflationary.
>There's a good argument that in our high reserves world, interest is actually inflationary.
How's that working out with Turkey?
It will probably work out just fine as they become a key player in European LNG.
Turkey couldn’t have serviced their debt with rising interest rates and continued government spending on expansion.
Being inflation averse makes sense when you can’t reasonably make use of funds. It’s not clear how much of that is actually related to the business cycle and how much is related to MMT, regardless of what adherents would have you believe.
UK historical investigation: https://www.elibrary.imf.org/display/book/9781557758897/ch07... - written in 2000, but you can see on this graph https://www.macrotrends.net/global-metrics/countries/gbr/uni... how flat it is from 1992 to 2020. That's a very good record for any piece of policy. Inflation control works for controlling business cycle inflation. However, it's not perfect and the COVID+war shock resulted in unavoidable inflation.
It's a ""plant"" in the https://en.wikipedia.org/wiki/Control_theory sense.
(my original comment: "you can see in the historical data when central bank targeting was introduced that the inflation rate drops fairly rapidly into the target band".
The US graph is similar https://www.macrotrends.net/global-metrics/countries/usa/uni... - although the adoption of inflation targeting wasn't fully formalized, it was definitely used in setting interest rates from the 90s.
Now show the UK plot for the 2010s. Also, show the one for Japan. It's easy to cherry pick data to show whatever you want. It doesn't constitute strong data for a casual and reliable link between interest rates and inflation.
Plot the days when your air conditioner is on with the temperature of your room. It will have many concrete examples and a long-term correlation showing that actually, the air conditioner is associated with the temperature going up.
All feedback/control systems are like that.
Your argument would have more weight if the inflation predictions were accurate. Here's the prediction report for the BoE in Aug 2014: https://www.bankofengland.co.uk/-/media/boe/files/inflation-...
(It's worth noting that even with the assumption of the models used being useful, the spread on those inflation rates is wild).
Here's what actually happened: https://www.ons.gov.uk/economy/inflationandpriceindices/time...
It very rapidly hit the bottom end of the prediction range before jumping up again pretty high. All that time interest rates were held constant and low.
Given they claim feedback lags of two years or so, one wonders what the point of all this is... (one cannot run a control loop with control lags substantially longer than the time constant of the system; that's basically the recipe for an unstable control system, assuming of course the control system is doing anything).
There's an argument that it's inflation expectations that matter, but there are dissenters within the temple that disagree: https://www.federalreserve.gov/econres/feds/files/2021062pap...
The second link https://www.macrotrends.net/global-metrics/countries/gbr/uni... is from 1960 to 2023.
This is the result of demographic crash and stagnation more than anything else...
You mean interest rates don't necessarily control inflation?
Japan is the worlds first steady state economy and got that way due to a sudden stagnation in trade and then demographic crash over a generation.
They've tried just about anything (except unrestrained foreign immigration or forced breeding) to jump start the economy but nothing has worked.
Inflation is effectively zero...
They haven't tried just about anything. They haven't tried printing enough money.
As a counterfactual: if you can print arbitrary amounts of money without raising inflation, you can just gradually buy up all the assets in the world with newly printed money. As far as I can tell, the Japanese central bank does not own the US stock market or all the gold in the world or all bitcoin etc, yet, so they haven't printed enough money.
Btw, what makes you think Japan is a 'steady state economy'? https://fred.stlouisfed.org/series/NYGDPPCAPKDJPN says their real GDP per capita grew fairly steadily over the years. It's just the price level that has been relatively steady, but if that's your yardstick, than just about any economy that used to be on a gold standard would also fit that idiosyncratic definition.
Exactly, but read many macro textbooks and they will tell you that interest rates control inflation. You've eloquently described a counter example and made the point I was trying to make.
Correct, it's mostly theater , and people nerding out on the numbers. The true measure of inflation is this: For a single day one works (calculated over a lifetime), how many days can one survive without working which will pay off all of one's bills. The lower this figure, higher is the inflation.
>The true measure of inflation is this: For a single day one works (calculated over a lifetime), how many days can one survive without working which will pay off all of one's bills.
This rapidly falls apart when you try to actually calculate it. Whose income do you use? Is inflation lower for doctors than burger flippers? What do you use as the retirement age? Does inflation go down if the retirement age is raised? What counts as "survive"? Does that mean the price of smartphones don't count toward inflation because you can theoretically survive without them?
Huh? What does this have to do with inflation at all?
Unless I miss something, it's very much not a standard measure of inflation. That said, leaning on that "at all": If wages are stickier than expenses, then the gap between wages and expenses will represent recent inflation to some degree.
If inflation is stable (and low-ish), then stickiness doesn't matter.
Sticky prices are only important, when expectations are invalidated.
So to fix your sentence:
> If wages are stickier than expenses, then the gap between wages and expenses will represent recent unexpected inflation to some degree.
> Simplified: the central bank decide on an interest rate that they want to see.
This is incorrect.
Using the US central bank example, the Fed has a target for inflation. It uses interest rates to try and hit that target. If inflation is higher than 3% the Fed will raise rates to cool the economy.
The Fed sets interest rates directly because lending money to the Fed is viewed as "risk-free" (as the US government has never defaulted on a debt). So if the Fed offers a risk-free 4%, banks will need to offer more because they are not risk-free. So banks no longer really lend savings out for loans. They borrow money and lend it out at a higher rate (eg mortgage-backed securities).
So when you could get a mortgage at 2.5%, it was because the Fed was offering 0%. When the Fed offers 5%, mortgage rates will go up to 7-8%.
There is another mechanism that the government could use to control inflation: fiscal policy, specifically taxation. A criticism of monetary policy to control inflation is that it's indiscriminate. People will go out of business and lose their houses. Taxes only target profits.
So in 2021-2022, we should've just passed a windfall profits tax of 80%. That would've cooled off inflation real quick and given the governments funds to distribute to those most adversely affected. But that will never happen because the corporations and wealthy who own both parties will never stand for wealth redistribution to the poor.
They will however demand wealth be transferred from the government to the rich.
The false equivalence of a profits tax and redistributing wealth to the poor is quite funny.