Here's a simple example for explaining what this means. Imagine I'm the government, and I sell bonds where I promise I'll pay you back $100 in 5 years or 10 years. Because I'm the government, and I can print money to repay any debt, these bonds are considered very low risk - sometimes they're even called risk free.
The market bids on these bonds, and the market price is public information. Let's say that people are paying $90 for a 5 year bond, meaning they pay $90 today and get $100 in 5 years. Knowing this, we can work out that this is equivalent to a yearly interest rate of 2.13%.
Now we expect that the computed interest rate for the 10 year bond should be higher than the 5 year bond. People want to be compensated for locking up their money for longer. So let's also say people are paying $80 for a 10 year bond, which works out to an interest rate of 2.26%.
These numbers fluctuate up and down every day, and they generally rise and fall with the prevailing interest rates. Here's the key question - what happens if the 10 year interest rate falls below the 5 year rate?
There are a lot of possible reasons, but this often indicates that investors think a recession is around the corner. The reason is that investors are betting that the prevailing interest rates will drop. If interest rates drop, then future bond purchases may have much lower interest rates than those bought today. It turns out you might like the idea of getting 2% interest for 10 years rather than 2.13% for 5 years and then 0.5% for the next 5 years.
Interestingly, the US has never had a recession that wasn't preceded by an inverted yield curve, but not every yield curve inversion has been followed by a recession.
> the US has never had a recession that wasn’t preceded by an inverted yield curve
This is incorrect. Most recessions in the post WW2 period have been preceded by an inversion of the 3m/10y UST curve, but it did not precede a few recessions in the USA during the first half of the 20th century.
Can't you essentially fabricate this statistic by controlling for the amount of time that counts as "before", assuming the inverted curves happen more often than recessions?
When large number of people are betting there is recession, there will a recession simply because people are collectively contracting their consumption in anticipation.
From what you've said, it's my interpretation that if the 10 year interest rate falls below the 5 year interest rate, then it means that bond purchasers speculate that money in 10 years will have more purchasing power than money in 5 years.
the idea is that you’re evaluated on your purchasing power ten years hence, but in one case you have to liquidate your position five years in and find another position with the same risk characteristics.
I'm actually hoping the fed refuses to lower rates. Let things dip a little rather than try to regulate the economy so much. Them jerking around rates was part of the trigger for 2008. Long term I think rates need to be higher anyway. So here's to hoping the speculators are wrong!
But it's also a bit philosophical, because any causal claim implicitly invokes a counterfactual. Like, if a bus drives off a cliff, was it the fault of the cliff or the bus driver who didn't respond quickly enough in trying to steer away? To me, common sense says that sometimes it's the cliff's fault and sometimes it's the driver's fault - the difference depends on how early the driver 'should' have known.
Yeah if by caused you mean could have acted more aggressively and quickly which would have reduced the depth and breadth of the recession than I've heard and agree with that.
It's just a weird use of the word caused that gives people the wrong impression a lot of the time. If I say "Fireman Rick caused that house to burn down" most people will think Fireman Rick is an arsonist and not necessarily that he didn't act swiftly and quickly to put it out. And then if you say most houses burning down were caused by firemen many people might get the wrong impression that if we got rid of the fire department less houses would burn down instead of more.
P.s. I know Milton Friedman is to blame for this terribly tortured use of the word.
Yeah, I totally agree. But on the other hand, if Fireman Rick's job is to put out the occasional small wildfire, and one day a small controllable fire starts but Fireman Rick says not "not my problem, I'm gonna stay home and play video games" and then that fire burns down half the state, there's a sense in which the fireman is to blame by choosing to abandon his job.
Or to give it more of a HN spin: if Rick's job was to run a website, but hackers broke in and stole all the plaintext passwords, it could be construed as partially Rick's fault for negligently storing the passwords in plaintext, even though the hackers were the ones directly responsible for the hack.
Yeah the Fed is screwing up massively when it comes to crises. They're keeping inflation too low(so they spend too much time trapped at the zero lower bound), and they're not as willing to engage in unconventional policy when the risks are clearly asymetric(a great recession is way more damaging than accidentally creating a little too much inflation temporarily)
FAS157 was the trigger for the recession. Look it up. Hindight is 2020 looking back. It's pretty evident. But yes interest rates higher and the risky investments almost certainly blew things up even more. FAS157 was the spotlight
Inflation has been way too high. Not uniformly, but chiefly housing where new money is introduced into the consumer realm.
In an inflation-neutral environment, one would expect the prices of manufactured goods to be trending slowly downwards. After all, producing something less expensively is exactly what economic competition leads to.
I think this is a combination of low interest rates making the opportunity cost of owning collectibles much lower and rising inequality driving up the purchasing power of collectors.
Well houses have gone up, but the price of housing (rent) have only gone up a bit.(except for certain markets driven by regulation and fundamentals)
If the government increased interest rates until inflation was zero unemployment would be twice what it is right now. And that seems like a high cost to pay for a neutral monetary base.
Inflation is an average. Sure, you can disagree on how to compute the average (what gets more weight) but then it gets pretty subjective and then the question is: too high compared to what?
I don't have the quote, but Alan Greenspan actually said the rapid lowering and raising of rates was part of the trigger. The financial environment was also not good, but they set it off:
Note the massive drop (stimulus) in the early 2000s. That pushed debt levels up and encouraged the crappy loans and all that. Then the rapid increase set it off. Home prices vary inversely with interest rates, so raising them rapidly cooled the market and put all that cheap debt under water.
Because we have 9 rate cuts before 0% interest, the ECB is at negative yield and so the BoJ?
The fed funds rate was around 4.5% in July of 2007 before the shit hit the fan. We need room to cut interest rates to spur growth when the next recession hits. 2.25% isn't a very big margin.
Because of the aging of the population and the consolidation of wealth the natural interest rate have fallen quite a bit. And you're right I think that's a big risk for the next recession.
But the only way to a higher nominal interest rate while keeping the same real rate is too raise the inflation rate to 3 or 4%.
> >2. Political dysfunction could cause the U.S. government to default notwithstanding its ability to print money
> This risk is likely also going to be true for the Dollar
It can't be, because it doesn't apply to the currency, which while it is structurally created by creation of debt is not a debt the government is obligated to pay and on which it can fail and default.
> so bonds don't carry any more risk than USD in this regard
Even if there was a default risk in the dollar, bonds necessarily have more risk since they are fully exposed to every risk to the currency they are denominated in and also their own default risk, which even if the currency had such a risk would sit on top of it.
I found the example to be very clear. What exactly are you confused about?
Here is the summary from the GP.
>[An inverted yield curve] often indicates that investors think a recession is around the corner. The reason is that investors are betting that the prevailing interest rates will drop. If interest rates drop, then future bond purchases may have much lower interest rates than those bought today. It turns out you might like the idea of getting 2% interest for 10 years rather than 2.13% for 5 years and then 0.5% for the next 5 years.
A simple example is 3 or 4 sentences, not 5 paragraphs.
He didn't address what it means for the economy. What happened last time the yield curve inverted? The last 10 times? The headline implies recession by referencing 2007, is there a valid statistical correlation?
> A simple example is 3 or 4 sentences, not 5 paragraphs.
Where is that rule written? Where did those particular numbers come from?
> He didn't address what it means for the economy. What happened last time the yield curve inverted? The last 10 times? The headline implies recession by referencing 2007, is there a valid statistical correlation?
You want the answer trimmed down to 3 or 4 sentences, but you also want answers to multiple new questions. Which one would you rather have?
>Because I'm the government, and I can print money to repay any debt, these bonds are considered very low risk - sometimes they're even called risk free.
This is absolutely not the reason they're considered low risk. If the USG tried to repay its debts by printing money, debtholders uninsured against inflation would lose tons of money. The expectation that they're _not_ going to print money to repay debt is part of why it's so safe.
> If the USG tried to repay its debts by printing money, debtholders uninsured against inflation would lose tons of money.
No, they'd lose anticipated value of the debt repayment compared to other goods and services if the gov did that without other compensatory policy, but get all the money they expected.
> The expectation that they're _not_ going to print money to repay debt is part of why it's so safe.
You are confusing “why dollars are safe” with “why US government debt is safe”.
The expectation that the money supply will be managed, in total, for stability and reasonably low inflation is why dollars are safe. (The expectation of not monetizing the debt, to the extent it is relevant at all, is relevant here.)
The expectation that the US government will not in any case default is why, among dollar-denominated instruments, US government debt is considered almost as safe as actual dollars.
You're acting as if default risk is the only relevant risk to creditors. Inflation risk is just as real. You can't just wish it away by saying it's "why dollars are safe". And then there are other risks on top of that - liquidity, reinvestment, etc.
> While the 3-month to 10-year spread “has a relatively decent track record of predicting recessions, it suffers from a timing problem,” said TD Securities U.S. rates strategist Gennadiy Goldberg. “Its inversion can suggest a recession occurred six months ago or will occur two years from now.”
Not just "relatively decent," but perfect over the last seven recessions at least. When the 3-month treasury yields more than the 10-year treasury, the economy has been in recession within two years 7/7 times.
There are a lot of leading economic indicators, but this one beats them all in terms of reliability.
There's one complicating factor that I don't see getting the attention it deserves. Unlike previous cycles, the Fed has been actively manipulating long-term yields for over a decade. That used to be taboo, but not anymore. The Fed holds trillions of dollars of long-term treasuries, accumulated under its QE/X programs.
The ownership of that debt by the Fed puts downward pressure on long-term yields. Recently, the Fed started allowing its long bonds to mature through its "quantitative tightening" (QT) program. That pressure (which really ramped up a few months ago) may have temporarily pumped up yields, and in so doing delayed a yield curve inversion signal. The status going forward of QT and the Fed's gargantuan balance sheet is anybody's guess. My bet is the Fed is done with QT until at least after the 2020 election. If things really start going south, watch for the Fed to begin buying stocks directly like the Bank of Japan has been doing for years. That and negative-yielding treasuries.
Either way, this is unprecedented territory. Signals could be quite confounded for the foreseeable future - which means policy makers are winging it. They almost certainly will make the wrong decisions.
As an investor, I find this rather a good thing. If I have to constantly afraid of recession then I would tend to just wait and sit on cash. If Fed is doing all these interventions including directly buying stock with their trillion dollar war chest + printing press then may be recession can be avoided or at least softened out. I also worry that recession may happen just because people are expecting it to happen even though economy is pretty sound. In that case, Fed intervention can avoid things falling fast just for short time and collective fear can be subsided back to sanity.
> There's one complicating factor that I don't see getting the attention it deserves. Unlike previous cycles, the Fed has been actively manipulating long-term yields for over a decade.
Yeah, very few are talking about what has been going on in the back end by them and how it distorts the curve gymnastics, but even the fed warned last year that it (long end inversion) might not be a reliable signal as it once was (because of their dealings) implying that things may come faster than people may expect if that was what they were only looking at.
The big problem is that people are now starting to realize is that even though QE can provide liquidity, it can't provide solvency: more debt just makes it worse when banks try to lend out their fresh reserves they were able to get from their trade in for liquid assets from the fed, especially if the top line of companies stays the same or starts to deteriorate, even setting aside companies that have been skirting along on negative income for so long because they were able to keep rolling over debt at ZLB.
>The status going forward of QT and the Fed's gargantuan balance sheet is anybody's guess. My bet is the Fed is done with QT until at least after the 2020 election.
QT is still on, just tweaked now:
"
- Begin tapering the “runoff” of Treasury securities in May.
- End the runoff of Treasury securities on September 30.
- Continue shedding mortgage-backed securities (MBS) at the current maximum of $20 billion a month, essentially until they’re gone.
- After September, reinvest MBS principal payments into Treasury securities.
Chair Jerome Powell said during the press conference that the balance sheet will by then be “a bit above $3.5 trillion.”
- The balance sheet will remain at this level even as the economy grows, thus slowly shrinking in relationship to GDP.
- The Fed may sell MBS outright to speed up the process of getting rid of them.
- No decision has been made on the delicate issue of the maturity composition of the balance sheet – which would require buying short-term bills for the first time in years to replace longer-term notes and bonds." [0][1]
>They almost certainly will make the wrong decisions.
Yes they will, and plenty of traders are betting on it. It's amazing that some people think CB's have control of this ship… but hey, someone needs to be on the other side of the trade.
Most of the commentary on the signaling of an inverted yield curve is all parroted by everyone else as “ominous”. Catherine Wood is the only person I’ve seen to go against the grain on this one [0].
Essentially, she notes that the yield curve has been inverted roughly 50% of the time in the late 19th century and early 20th century. Furthermore, during the times it was inverted, it marked periods of strong growth!
It’s an interesting take as everyone else conveniently (more like lazy) are not doing the research and going back as far back as late 18th century. Her point of view is contrarian and refreshingly unique!
> Essentially, she notes that the yield curve has been inverted roughly 50% of the time in the late 19th century and early 20th century.
Actually, she says specifically 1800-1929, which is an interesting choice because AFAICT the “3-month" security (the 13-week T-bill) referenced when discussing the 3m/10y inversion (the one discussed here as a key indicator) was first regularly issued in 1929. [0]
The behavior of the markets when either or both of the instruments in the analysis were not regularly issued is obviously not comparable. (Further, when they aren't both regularly issued, assessing how much of the time an inversion was in effect is, at best, creative extrapolation, and more likely outright fiction.)
> It’s an interesting take as everyone else conveniently (more like lazy) are not doing the research and going back as far back as late 18th century.
So lazy of everyone else to only study a measure as far back as the basic premises of using the measure are valid and the components that make it up actually exist.
A lot of the people in Finance are surprisingly superstitious. I had a boss once who would not conduct trades if he saw a totally unrelated bad omen.
The treasury yield curve is a measure of expectations. It's a market marker that is the result of traders and reflects the overall view of the economy.
Having said that, it's also taken on a mythical status where now that it has inverted, people will start planning for a recession which may in turn lead to happening much sooner/if at all. It's a self fulfilling prophecy in a way where the expectation of a recession leads you act in a way that leads to that outcome.
In December it was the 3 and 5 year which inverted, and today it was the 3 month and 10 year. Typically I’ve understood the biggest one to watch is the 2 and 10 year though.
I'm not sure that the yield curve isn't us being fooled by randomness. Everyone is looking at every metric for predictive ability. The odds that one of them went 7/7 have to approach 1.
I am surprised this is not happening more frequently. An inversion between the 1 and 5 year bonds just means that the expected average yield for each of the next 5 years is less than the yield for next year. Statistically that should be the case 50% of the time. Treasure bonds are liquid, so buying a 5 year one does not mean that you have to hold it to maturity.
There are a few factors that cause one to expect a monotonically increasing, concave yield curve [0].
Firstly, one must consider the term premium [1]. In short, you expect to be compensated more for lending money for a longer period of time. Consider lending money for two years. You could lend your money in two single year-long increments, in which case you would compound your gains from the first year when you lend out the second year. If you lend in a single, two year-long commitment, you rationally expect, all else being equal, to earn as much as you would have by lending as described previously. This drives the general upwards slope (or contango) of the yield curve.
Next, consider a credit component. When you lend someone money, how likely is it that you are paid back? If you expect there is default risk, you factor that into the interest rate you demand. As the tenor of the bond increases, the risk of default increases. However, the extra interest you demand for credit risk also decreases as time goes out. Why? What are the chances that someone defaults on a loan between 5 and 10 years? Probably greater than the risk that the party defaults between 10 and 15 years. This drives the concavity of the yield curve.
Finally, the movement in the curve is driven by expectations of future interest rates, as determined by Federal Reserve policy. Fed typically acts at the short end of the curve. In its tool chest, fed can manipulate rates like the fed funds rate, IOER, and ON REPO. As the economy improves, fed will raise rates. When economic outlook declines, fed will lower rates. If you expect that rates will be higher in the future, you will demand higher yields for longer dated bonds, since your invested money will earn less of a premium to interest rates in the future relative to where you invested today. On the flip side, if you expect interest rates to decline, you will want to lock in your money now, so you purchase longer dated bonds, since you do not expect to be able to get as high a yield in the future. This action at the long end of the curve, coupled with fed policy at the short end, ultimately drives yield curve fluctuations.
>Statistically that should be the case 50% of the time.
Only if the risk premium is the same.
>Treasure bonds are liquid, so buying a 5 year one does not mean that you have to hold it to maturity.
Here you claim that the risk premium is the same because it's liquid. However the term is not the same so you'll be hit with a bigger effect when short term interest rates change on a longer time horizon instrument.
Where are you getting this 50% from? Also, you're ignoring that there is more price risk in long bonds even if you can sell them, so there should anyway be a term premium (upward-sloping curve) even if the expected interest rate over five years is constant.
Hmm for some reason I thought this was already the case actually.
I was looking at Ally CD rates a few weeks ago, and I saw an inversion there and I just chalked that up to the CD rates being dependent on the yield curve.
There was a less extreme yield curve inversion a few months ago, which is likely what you're referring to. But it wasn't so far reaching as the current inversion, which puts the 1 year Treasury over the 10 year.
It may have predicted ten out of the last seven recessions. That is, sometimes the yield curve inverts and the recession doesn't happen. So it's not quite a self-fulfilling prophesy - though it still may be a sometimes-self-fulfilling prophesy...
Lending money for longer term is disadvantaged vs. shorter term so borrowing is impacted. This leads to less stability and in the past has been a fairly good indicator of an upcoming recession.
Past performance is not an indicator of future returns.
When investors buy bonds, specifically US Treasury securities, the relationship between the yield and the time to maturity (2,3,5,7,10,30 years) is generally upward sloping because, as the logic goes, the longer dated securities should have a higher yield because the investor wants to get paid more since they are parting with their money for longer.
When we see yield curve inversions, usually beginning with just a small 'kink' on some point in the curve, that means that the yield at a point further out is lower than a shorter dated security. When the yield curve inverts like this, it means that demand for shorter-dated bonds is higher than longer-dated bonds, and because the duration of longer-dated bonds is higher than shorter-dated bonds, when demand falls for longer-dated bonds, those investors will instead buy shorter-dated bonds. Can also mean that the market believes that a recession is coming, so there's a flight to safety out of equities and into shorter-dated treasuries.
I think you have the relationship between yield and demand backwards. The yield of a bond and its price are inversely related, so an inverted yield curve (lower yields at longer durations) implies the opposite of what you said: there is higher demand for longer-dated bonds relative to short-dated ones. One way to interpret this is that investors expect poorer asset returns in the short term, so they are willing to lock in a lower yield for a long period of time instead of, for example, taking significant losses in the stock market.
> When the yield curve inverts like this, it means that demand for shorter-dated bonds is higher than longer-dated bonds
That is true for a booming economy, but not for the inverted yield case.
It all comes down to where investors think the Fed is headed when it comes to interest rates.
For a booming economy the Fed puts up rates in an effort to slow growth.
For a lagging economy the Fed cuts the rate in an effort to stimulate growth.
Now consider the investor wanting to invest in bonds hoping for the best possible return.
In a booming economy you would only invest in short term 5 year bonds because chances are future interest rates will be higher than what you are currently getting.
But for a slowing economy it might be better to invest in the 10 year bond, to lock in a long term interest rate only because in the future interest rates may well be lower.
Now consider the supply and demand forces that also come into play.
When the demand is high the price goes up and the yield goes down and when the demand is low the price goes down and the yield goes up.
So when you have large numbers of investors wanting to get out of 5 year bonds (i.e. high selling) and into 10 year bonds (i.e. high buying) you get this inverted yield situation.
Perfect in layman’s terms. The Federal reserve keeps spewing the economy is so strong and nobody in the bond world is buying it, thus locking in the long term rates now (they’ll never admit a recession is coming, or blindly don’t know).
It’s interesting to note the now 60% of bonds are now inverted, it was just 50% a day ago!!!
Even if one was to say this is just one indicator look at Canada -50% inverted and housing bubble currently
Here's a simple example for explaining what this means. Imagine I'm the government, and I sell bonds where I promise I'll pay you back $100 in 5 years or 10 years. Because I'm the government, and I can print money to repay any debt, these bonds are considered very low risk - sometimes they're even called risk free.
The market bids on these bonds, and the market price is public information. Let's say that people are paying $90 for a 5 year bond, meaning they pay $90 today and get $100 in 5 years. Knowing this, we can work out that this is equivalent to a yearly interest rate of 2.13%.
Now we expect that the computed interest rate for the 10 year bond should be higher than the 5 year bond. People want to be compensated for locking up their money for longer. So let's also say people are paying $80 for a 10 year bond, which works out to an interest rate of 2.26%.
These numbers fluctuate up and down every day, and they generally rise and fall with the prevailing interest rates. Here's the key question - what happens if the 10 year interest rate falls below the 5 year rate?
There are a lot of possible reasons, but this often indicates that investors think a recession is around the corner. The reason is that investors are betting that the prevailing interest rates will drop. If interest rates drop, then future bond purchases may have much lower interest rates than those bought today. It turns out you might like the idea of getting 2% interest for 10 years rather than 2.13% for 5 years and then 0.5% for the next 5 years.
Interestingly, the US has never had a recession that wasn't preceded by an inverted yield curve, but not every yield curve inversion has been followed by a recession.
> the US has never had a recession that wasn’t preceded by an inverted yield curve
This is incorrect. Most recessions in the post WW2 period have been preceded by an inversion of the 3m/10y UST curve, but it did not precede a few recessions in the USA during the first half of the 20th century.
[0] http://bonddad.blogspot.com/2011/07/yield-curve-and-pre-ww2-...
Can't you essentially fabricate this statistic by controlling for the amount of time that counts as "before", assuming the inverted curves happen more often than recessions?
I am guessing that "before" means that the inversion persisted until the economy was technically in recession.
Not quite - see https://fred.stlouisfed.org/series/T10Y3M for reference. The grey areas are recessions, and values below 0 are inversions.
To address the GP’s point, the relationship could be fabricated, but the data don’t indicate that that’s happening.
When large number of people are betting there is recession, there will a recession simply because people are collectively contracting their consumption in anticipation.
This doesn't account for recessions before the creation of the Federal Reserve.
From what you've said, it's my interpretation that if the 10 year interest rate falls below the 5 year interest rate, then it means that bond purchasers speculate that money in 10 years will have more purchasing power than money in 5 years.
Is that interpretation correct?
...money in five years in five further years.
the idea is that you’re evaluated on your purchasing power ten years hence, but in one case you have to liquidate your position five years in and find another position with the same risk characteristics.
I'm actually hoping the fed refuses to lower rates. Let things dip a little rather than try to regulate the economy so much. Them jerking around rates was part of the trigger for 2008. Long term I think rates need to be higher anyway. So here's to hoping the speculators are wrong!
Why do you think rates need to be higher? Inflation has been to high or unemployment has been too low?
And I've never heard anyone say the fed caused the 2008 recession. It's pretty much universally agreed upon it was finance blowing up.
Quite a few macroeconomists have said the Fed caused or depeened the recession: http://bruegel.org/2016/02/blaming-the-fed-for-the-great-rec...
But it's also a bit philosophical, because any causal claim implicitly invokes a counterfactual. Like, if a bus drives off a cliff, was it the fault of the cliff or the bus driver who didn't respond quickly enough in trying to steer away? To me, common sense says that sometimes it's the cliff's fault and sometimes it's the driver's fault - the difference depends on how early the driver 'should' have known.
Yeah if by caused you mean could have acted more aggressively and quickly which would have reduced the depth and breadth of the recession than I've heard and agree with that.
It's just a weird use of the word caused that gives people the wrong impression a lot of the time. If I say "Fireman Rick caused that house to burn down" most people will think Fireman Rick is an arsonist and not necessarily that he didn't act swiftly and quickly to put it out. And then if you say most houses burning down were caused by firemen many people might get the wrong impression that if we got rid of the fire department less houses would burn down instead of more.
P.s. I know Milton Friedman is to blame for this terribly tortured use of the word.
Yeah, I totally agree. But on the other hand, if Fireman Rick's job is to put out the occasional small wildfire, and one day a small controllable fire starts but Fireman Rick says not "not my problem, I'm gonna stay home and play video games" and then that fire burns down half the state, there's a sense in which the fireman is to blame by choosing to abandon his job.
Or to give it more of a HN spin: if Rick's job was to run a website, but hackers broke in and stole all the plaintext passwords, it could be construed as partially Rick's fault for negligently storing the passwords in plaintext, even though the hackers were the ones directly responsible for the hack.
Yeah the Fed is screwing up massively when it comes to crises. They're keeping inflation too low(so they spend too much time trapped at the zero lower bound), and they're not as willing to engage in unconventional policy when the risks are clearly asymetric(a great recession is way more damaging than accidentally creating a little too much inflation temporarily)
FAS157 was the trigger for the recession. Look it up. Hindight is 2020 looking back. It's pretty evident. But yes interest rates higher and the risky investments almost certainly blew things up even more. FAS157 was the spotlight
We can debate if FAS 157 was “the” trigger but it certainly was not the cause.
Inflation has been way too high. Not uniformly, but chiefly housing where new money is introduced into the consumer realm.
In an inflation-neutral environment, one would expect the prices of manufactured goods to be trending slowly downwards. After all, producing something less expensively is exactly what economic competition leads to.
Housing, startups, art, cars, watches...pretty much everything collectible or rare has seen its value skyrocket since the great recession.
The big pool of money has been sloshing between asset classes for a while now, but it doesn’t get reflected in inflation metrics.
I think this is a combination of low interest rates making the opportunity cost of owning collectibles much lower and rising inequality driving up the purchasing power of collectors.
Well houses have gone up, but the price of housing (rent) have only gone up a bit.(except for certain markets driven by regulation and fundamentals)
If the government increased interest rates until inflation was zero unemployment would be twice what it is right now. And that seems like a high cost to pay for a neutral monetary base.
Inflation is an average. Sure, you can disagree on how to compute the average (what gets more weight) but then it gets pretty subjective and then the question is: too high compared to what?
I don't have the quote, but Alan Greenspan actually said the rapid lowering and raising of rates was part of the trigger. The financial environment was also not good, but they set it off:
https://en.wikipedia.org/wiki/Federal_funds_rate#/media/File...
Note the massive drop (stimulus) in the early 2000s. That pushed debt levels up and encouraged the crappy loans and all that. Then the rapid increase set it off. Home prices vary inversely with interest rates, so raising them rapidly cooled the market and put all that cheap debt under water.
Because we have 9 rate cuts before 0% interest, the ECB is at negative yield and so the BoJ?
The fed funds rate was around 4.5% in July of 2007 before the shit hit the fan. We need room to cut interest rates to spur growth when the next recession hits. 2.25% isn't a very big margin.
Because of the aging of the population and the consolidation of wealth the natural interest rate have fallen quite a bit. And you're right I think that's a big risk for the next recession.
But the only way to a higher nominal interest rate while keeping the same real rate is too raise the inflation rate to 3 or 4%.
> sometimes they're even called risk free
They may be called that, but they aren't. There are at least two risks associated with U.S. government bonds:
1. Inflation can exceed the interest rate on the bond and cause it to lose value when measured in actual purchasing power
2. Political dysfunction could cause the U.S. government to default notwithstanding its ability to print money
(To be fair, #1 is a risk associated with any bond. #2 is peculiar to U.S. government bonds.)
Risk-free is in relation to the risk of getting your money back. Not the risk that the return is lower than inflation.
#2 is true, but extremely unlikely.
Risk-free basically means the lowest possible risk. Not true zero risk.
> #2 is peculiar to U.S. government bonds.)
not really, with private bonds it's replaced with the risk of the company defaulting or declaring bankruptcy.
>2. Political dysfunction could cause the U.S. government to default notwithstanding its ability to print money
This risk is likely also going to be true for the Dollar, so bonds don't carry any more risk than USD in this regard.
> >2. Political dysfunction could cause the U.S. government to default notwithstanding its ability to print money
> This risk is likely also going to be true for the Dollar
It can't be, because it doesn't apply to the currency, which while it is structurally created by creation of debt is not a debt the government is obligated to pay and on which it can fail and default.
> so bonds don't carry any more risk than USD in this regard
Even if there was a default risk in the dollar, bonds necessarily have more risk since they are fully exposed to every risk to the currency they are denominated in and also their own default risk, which even if the currency had such a risk would sit on top of it.
Except that if/when political dysfunction does cause a default on the bonds, the currency gets its legs chopped out from under it.
Which makes it exceptionally unlikely to happen..
Although Brexit happened, so never say never I guess
At least we've not returned to the era of "return-free risk"
Also exchange rate risk, if the sovereign bond in question is not in your home currency.
Every investment carries risk in the form of opportunity cost.
I-bonds gain interest based on inflation
>> Here's a simple example for explaining what this means
This was not simple, nor did you explain what this means
I found the example to be very clear. What exactly are you confused about?
Here is the summary from the GP.
>[An inverted yield curve] often indicates that investors think a recession is around the corner. The reason is that investors are betting that the prevailing interest rates will drop. If interest rates drop, then future bond purchases may have much lower interest rates than those bought today. It turns out you might like the idea of getting 2% interest for 10 years rather than 2.13% for 5 years and then 0.5% for the next 5 years.
A simple example is 3 or 4 sentences, not 5 paragraphs.
He didn't address what it means for the economy. What happened last time the yield curve inverted? The last 10 times? The headline implies recession by referencing 2007, is there a valid statistical correlation?
> A simple example is 3 or 4 sentences, not 5 paragraphs.
Where is that rule written? Where did those particular numbers come from?
> He didn't address what it means for the economy. What happened last time the yield curve inverted? The last 10 times? The headline implies recession by referencing 2007, is there a valid statistical correlation?
You want the answer trimmed down to 3 or 4 sentences, but you also want answers to multiple new questions. Which one would you rather have?
I don't want anything. I'm just holding the top comment to its author's own standards that he voluntarily stated in his first sentence.
Can you get it into three sentences? That do not assume the reader knows any financial jargon?
Yes.
Are you going to or are you just gonna be a smartass?
>Because I'm the government, and I can print money to repay any debt, these bonds are considered very low risk - sometimes they're even called risk free.
This is absolutely not the reason they're considered low risk. If the USG tried to repay its debts by printing money, debtholders uninsured against inflation would lose tons of money. The expectation that they're _not_ going to print money to repay debt is part of why it's so safe.
> If the USG tried to repay its debts by printing money, debtholders uninsured against inflation would lose tons of money.
No, they'd lose anticipated value of the debt repayment compared to other goods and services if the gov did that without other compensatory policy, but get all the money they expected.
> The expectation that they're _not_ going to print money to repay debt is part of why it's so safe.
You are confusing “why dollars are safe” with “why US government debt is safe”.
The expectation that the money supply will be managed, in total, for stability and reasonably low inflation is why dollars are safe. (The expectation of not monetizing the debt, to the extent it is relevant at all, is relevant here.)
The expectation that the US government will not in any case default is why, among dollar-denominated instruments, US government debt is considered almost as safe as actual dollars.
You're acting as if default risk is the only relevant risk to creditors. Inflation risk is just as real. You can't just wish it away by saying it's "why dollars are safe". And then there are other risks on top of that - liquidity, reinvestment, etc.
> You're acting as if default risk is the only relevant risk to creditors.
No, I'm correctly characterizing inflation risk, which—by definition—does not involve losing money you had or anticipated.
> You can't just wish it away by saying it's "why dollars are safe".
No, the perceived lack of significant actual inflation risk is why dollars are considered safe, not the risk category itself.
> While the 3-month to 10-year spread “has a relatively decent track record of predicting recessions, it suffers from a timing problem,” said TD Securities U.S. rates strategist Gennadiy Goldberg. “Its inversion can suggest a recession occurred six months ago or will occur two years from now.”
Not just "relatively decent," but perfect over the last seven recessions at least. When the 3-month treasury yields more than the 10-year treasury, the economy has been in recession within two years 7/7 times.
https://seekingalpha.com/article/4250287-fully-inverted-yiel...
There are a lot of leading economic indicators, but this one beats them all in terms of reliability.
There's one complicating factor that I don't see getting the attention it deserves. Unlike previous cycles, the Fed has been actively manipulating long-term yields for over a decade. That used to be taboo, but not anymore. The Fed holds trillions of dollars of long-term treasuries, accumulated under its QE/X programs.
The ownership of that debt by the Fed puts downward pressure on long-term yields. Recently, the Fed started allowing its long bonds to mature through its "quantitative tightening" (QT) program. That pressure (which really ramped up a few months ago) may have temporarily pumped up yields, and in so doing delayed a yield curve inversion signal. The status going forward of QT and the Fed's gargantuan balance sheet is anybody's guess. My bet is the Fed is done with QT until at least after the 2020 election. If things really start going south, watch for the Fed to begin buying stocks directly like the Bank of Japan has been doing for years. That and negative-yielding treasuries.
Either way, this is unprecedented territory. Signals could be quite confounded for the foreseeable future - which means policy makers are winging it. They almost certainly will make the wrong decisions.
As an investor, I find this rather a good thing. If I have to constantly afraid of recession then I would tend to just wait and sit on cash. If Fed is doing all these interventions including directly buying stock with their trillion dollar war chest + printing press then may be recession can be avoided or at least softened out. I also worry that recession may happen just because people are expecting it to happen even though economy is pretty sound. In that case, Fed intervention can avoid things falling fast just for short time and collective fear can be subsided back to sanity.
> There's one complicating factor that I don't see getting the attention it deserves. Unlike previous cycles, the Fed has been actively manipulating long-term yields for over a decade.
Yeah, very few are talking about what has been going on in the back end by them and how it distorts the curve gymnastics, but even the fed warned last year that it (long end inversion) might not be a reliable signal as it once was (because of their dealings) implying that things may come faster than people may expect if that was what they were only looking at.
The big problem is that people are now starting to realize is that even though QE can provide liquidity, it can't provide solvency: more debt just makes it worse when banks try to lend out their fresh reserves they were able to get from their trade in for liquid assets from the fed, especially if the top line of companies stays the same or starts to deteriorate, even setting aside companies that have been skirting along on negative income for so long because they were able to keep rolling over debt at ZLB.
>The status going forward of QT and the Fed's gargantuan balance sheet is anybody's guess. My bet is the Fed is done with QT until at least after the 2020 election.
QT is still on, just tweaked now:
" - Begin tapering the “runoff” of Treasury securities in May.
- End the runoff of Treasury securities on September 30.
- Continue shedding mortgage-backed securities (MBS) at the current maximum of $20 billion a month, essentially until they’re gone.
- After September, reinvest MBS principal payments into Treasury securities. Chair Jerome Powell said during the press conference that the balance sheet will by then be “a bit above $3.5 trillion.”
- The balance sheet will remain at this level even as the economy grows, thus slowly shrinking in relationship to GDP.
- The Fed may sell MBS outright to speed up the process of getting rid of them.
- No decision has been made on the delicate issue of the maturity composition of the balance sheet – which would require buying short-term bills for the first time in years to replace longer-term notes and bonds." [0][1]
>They almost certainly will make the wrong decisions.
Yes they will, and plenty of traders are betting on it. It's amazing that some people think CB's have control of this ship… but hey, someone needs to be on the other side of the trade.
[0] https://wolfstreet.com/2019/03/20/feds-new-balance-sheet-pla...
[1] https://www.federalreserve.gov/newsevents/pressreleases/mone...
Most of the commentary on the signaling of an inverted yield curve is all parroted by everyone else as “ominous”. Catherine Wood is the only person I’ve seen to go against the grain on this one [0].
Essentially, she notes that the yield curve has been inverted roughly 50% of the time in the late 19th century and early 20th century. Furthermore, during the times it was inverted, it marked periods of strong growth!
It’s an interesting take as everyone else conveniently (more like lazy) are not doing the research and going back as far back as late 18th century. Her point of view is contrarian and refreshingly unique!
[0] https://www.cnbc.com/2018/12/04/the-economy-will-surprise-to...
> Essentially, she notes that the yield curve has been inverted roughly 50% of the time in the late 19th century and early 20th century.
Actually, she says specifically 1800-1929, which is an interesting choice because AFAICT the “3-month" security (the 13-week T-bill) referenced when discussing the 3m/10y inversion (the one discussed here as a key indicator) was first regularly issued in 1929. [0]
The behavior of the markets when either or both of the instruments in the analysis were not regularly issued is obviously not comparable. (Further, when they aren't both regularly issued, assessing how much of the time an inversion was in effect is, at best, creative extrapolation, and more likely outright fiction.)
> It’s an interesting take as everyone else conveniently (more like lazy) are not doing the research and going back as far back as late 18th century.
So lazy of everyone else to only study a measure as far back as the basic premises of using the measure are valid and the components that make it up actually exist.
[0] https://www.treasurydirect.gov/indiv/research/history/histmk...
wouldn't the motion of the yield curve in that era be very different given that the dollar was long-term noninflationary?
if you expect the dollar to deflate in the long term, you might be OK accepting a lower long term yield.
A lot of the people in Finance are surprisingly superstitious. I had a boss once who would not conduct trades if he saw a totally unrelated bad omen.
The treasury yield curve is a measure of expectations. It's a market marker that is the result of traders and reflects the overall view of the economy.
Having said that, it's also taken on a mythical status where now that it has inverted, people will start planning for a recession which may in turn lead to happening much sooner/if at all. It's a self fulfilling prophecy in a way where the expectation of a recession leads you act in a way that leads to that outcome.
Is it that surprising though? For a group of people that believes they can predict the future?
similar story with charting.
notably, the 2 & 10 year has not yet inverted
https://fred.stlouisfed.org/series/T10Y2Y
In December it was the 3 and 5 year which inverted, and today it was the 3 month and 10 year. Typically I’ve understood the biggest one to watch is the 2 and 10 year though.
I'm not sure that the yield curve isn't us being fooled by randomness. Everyone is looking at every metric for predictive ability. The odds that one of them went 7/7 have to approach 1.
Yield curve inversion is something like like 15/7, actually.
I am surprised this is not happening more frequently. An inversion between the 1 and 5 year bonds just means that the expected average yield for each of the next 5 years is less than the yield for next year. Statistically that should be the case 50% of the time. Treasure bonds are liquid, so buying a 5 year one does not mean that you have to hold it to maturity.
Somebody tell me where I'm wrong.
There are a few factors that cause one to expect a monotonically increasing, concave yield curve [0].
Firstly, one must consider the term premium [1]. In short, you expect to be compensated more for lending money for a longer period of time. Consider lending money for two years. You could lend your money in two single year-long increments, in which case you would compound your gains from the first year when you lend out the second year. If you lend in a single, two year-long commitment, you rationally expect, all else being equal, to earn as much as you would have by lending as described previously. This drives the general upwards slope (or contango) of the yield curve.
Next, consider a credit component. When you lend someone money, how likely is it that you are paid back? If you expect there is default risk, you factor that into the interest rate you demand. As the tenor of the bond increases, the risk of default increases. However, the extra interest you demand for credit risk also decreases as time goes out. Why? What are the chances that someone defaults on a loan between 5 and 10 years? Probably greater than the risk that the party defaults between 10 and 15 years. This drives the concavity of the yield curve.
Finally, the movement in the curve is driven by expectations of future interest rates, as determined by Federal Reserve policy. Fed typically acts at the short end of the curve. In its tool chest, fed can manipulate rates like the fed funds rate, IOER, and ON REPO. As the economy improves, fed will raise rates. When economic outlook declines, fed will lower rates. If you expect that rates will be higher in the future, you will demand higher yields for longer dated bonds, since your invested money will earn less of a premium to interest rates in the future relative to where you invested today. On the flip side, if you expect interest rates to decline, you will want to lock in your money now, so you purchase longer dated bonds, since you do not expect to be able to get as high a yield in the future. This action at the long end of the curve, coupled with fed policy at the short end, ultimately drives yield curve fluctuations.
[0] https://obliviousinvestor.com/wp-content/uploads/2012/07/yie...
[1] https://www.bloomberg.com/news/articles/2017-10-30/what-s-a-...
>Statistically that should be the case 50% of the time.
Only if the risk premium is the same.
>Treasure bonds are liquid, so buying a 5 year one does not mean that you have to hold it to maturity.
Here you claim that the risk premium is the same because it's liquid. However the term is not the same so you'll be hit with a bigger effect when short term interest rates change on a longer time horizon instrument.
Where are you getting this 50% from? Also, you're ignoring that there is more price risk in long bonds even if you can sell them, so there should anyway be a term premium (upward-sloping curve) even if the expected interest rate over five years is constant.
Hmm for some reason I thought this was already the case actually.
I was looking at Ally CD rates a few weeks ago, and I saw an inversion there and I just chalked that up to the CD rates being dependent on the yield curve.
Does it?
There was a less extreme yield curve inversion a few months ago, which is likely what you're referring to. But it wasn't so far reaching as the current inversion, which puts the 1 year Treasury over the 10 year.
What would it cost to make this happen?
Can someone explain what kind of impact this has?
To my understanding, this is a metric, so it's not quite right to say that it has an "impact". But, from the article:
> Inversion is considered a reliable harbinger of recession in the U.S., within roughly the next 18 months.
Apparently it has predicted the last seven recessions. People are using the expression "self-fulfilling prophecy."
It may have predicted ten out of the last seven recessions. That is, sometimes the yield curve inverts and the recession doesn't happen. So it's not quite a self-fulfilling prophesy - though it still may be a sometimes-self-fulfilling prophesy...
correct. we can't determine the usefulness of this statistic without a false positive rate.
Exactly. Every one looks at the last 10 years of a bull market and says “recession is coming!”.
Many of these people are bidding on these same bonds. So does the inversion signal a coming recession or just the belief of a coming recession?
Lending money for longer term is disadvantaged vs. shorter term so borrowing is impacted. This leads to less stability and in the past has been a fairly good indicator of an upcoming recession.
Past performance is not an indicator of future returns.
When investors buy bonds, specifically US Treasury securities, the relationship between the yield and the time to maturity (2,3,5,7,10,30 years) is generally upward sloping because, as the logic goes, the longer dated securities should have a higher yield because the investor wants to get paid more since they are parting with their money for longer.
https://en.wikipedia.org/wiki/Yield_curve#/media/File:Yield_...
something like that is usually "what's expected".
When we see yield curve inversions, usually beginning with just a small 'kink' on some point in the curve, that means that the yield at a point further out is lower than a shorter dated security. When the yield curve inverts like this, it means that demand for shorter-dated bonds is higher than longer-dated bonds, and because the duration of longer-dated bonds is higher than shorter-dated bonds, when demand falls for longer-dated bonds, those investors will instead buy shorter-dated bonds. Can also mean that the market believes that a recession is coming, so there's a flight to safety out of equities and into shorter-dated treasuries.
That's the gist of it
I think you have the relationship between yield and demand backwards. The yield of a bond and its price are inversely related, so an inverted yield curve (lower yields at longer durations) implies the opposite of what you said: there is higher demand for longer-dated bonds relative to short-dated ones. One way to interpret this is that investors expect poorer asset returns in the short term, so they are willing to lock in a lower yield for a long period of time instead of, for example, taking significant losses in the stock market.
> When the yield curve inverts like this, it means that demand for shorter-dated bonds is higher than longer-dated bonds
That is true for a booming economy, but not for the inverted yield case.
It all comes down to where investors think the Fed is headed when it comes to interest rates.
For a booming economy the Fed puts up rates in an effort to slow growth.
For a lagging economy the Fed cuts the rate in an effort to stimulate growth.
Now consider the investor wanting to invest in bonds hoping for the best possible return.
In a booming economy you would only invest in short term 5 year bonds because chances are future interest rates will be higher than what you are currently getting.
But for a slowing economy it might be better to invest in the 10 year bond, to lock in a long term interest rate only because in the future interest rates may well be lower.
Now consider the supply and demand forces that also come into play.
When the demand is high the price goes up and the yield goes down and when the demand is low the price goes down and the yield goes up.
So when you have large numbers of investors wanting to get out of 5 year bonds (i.e. high selling) and into 10 year bonds (i.e. high buying) you get this inverted yield situation.
Perfect in layman’s terms. The Federal reserve keeps spewing the economy is so strong and nobody in the bond world is buying it, thus locking in the long term rates now (they’ll never admit a recession is coming, or blindly don’t know).
It’s interesting to note the now 60% of bonds are now inverted, it was just 50% a day ago!!!
Even if one was to say this is just one indicator look at Canada -50% inverted and housing bubble currently