Update: 700 Years of Falling Interest Rates, 1311-2021

Discuss all general (i.e. non-personal) investing questions and issues, investing news, and theory.
Post Reply
User avatar
Topic Author
SimpleGift
Posts: 4477
Joined: Tue Feb 08, 2011 2:45 pm

Update: 700 Years of Falling Interest Rates, 1311-2021

Post by SimpleGift »

This post reports on a new follow-up paper to the groundbreaking 2019 study of the "supra-secular" decline in real interest rates over 700 years, since the 14th century (chart below). The original paper is discussed in this Forum thread.
  • Image
    Note: Trend line in black is GDP-weighted among countries. Chart source: The Economist
While the original paper was quite readable, this follow-up paper is more statistically oriented. Some key findings:
  • 1. Global real, long-term rates exhibit a gentle but firm downward trend (averaging about -1.6% every 100 years), and the evidence is strongly consistent with the hypothesis that they are stationary around this trend. In other words, rates tended to mean revert around this trend over long periods.

    2. The only two events to disrupt the trend were the Black Death and a wave of sovereign-debt defaults in the late 1550s. All other deviations — including after the founding of America’s Federal Reserve in 1913 and the advent of inflation-targeting central banks — could not be distinguished statistically from random chance.

    3. The 700-year pattern suggests that current real rates should be about 0.7%. The authors speculate that the trend may be more likely to flatten out as rates approach negative territory than to continue further downward.
There's much more in this follow-up paper that may be of particular interest to those with statistics and econometrics backgrounds — but it may also be of more general interest just for its financial history. Comments?
sean.mcgrath
Posts: 786
Joined: Thu Dec 29, 2016 5:15 am
Location: US in NL

Re: Update: 700 Years of Falling Interest Rates, 1311-2021

Post by sean.mcgrath »

I was just about to post on this!

"The authors speculate that the trend may be more likely to flatten out as rates approach negative territory than to continue further downward." However, they admit that they have not yet seen any evidence for this.

The interesting question to me is the mechanism. Admittedly I only skimmed the papers, but the only stab at this I was able to find was "though capital accumulation trends may go some way in explaining the phenomenon"

Do you have any guesses about what the mechanism might be?
Last edited by sean.mcgrath on Fri Oct 07, 2022 5:28 am, edited 1 time in total.
Frank2012
Posts: 194
Joined: Fri Jan 25, 2013 6:16 pm

Re: Update: 700 Years of Falling Interest Rates, 1311-2021

Post by Frank2012 »

SimpleGift wrote: Fri Oct 07, 2022 2:04 am This post reports on a new follow-up paper to the groundbreaking 2019 study of the "supra-secular" decline in real interest rates over 700 years, since the 14th century (chart below). The original paper is discussed in this Forum thread.
  • Image
    Note: Trend line in black is GDP-weighted among countries. Chart source: The Economist
While the original paper was quite readable, this follow-up paper is more statistically oriented. Some key findings:
  • 1. Global real, long-term rates exhibit a gentle but firm downward trend (averaging about -1.6% every 100 years), and the evidence is strongly consistent with the hypothesis that they are stationary around this trend. In other words, rates tended to mean revert around this trend over long periods.

    2. The only two events to disrupt the trend were the Black Death and a wave of sovereign-debt defaults in the late 1550s. All other deviations — including after the founding of America’s Federal Reserve in 1913 and the advent of inflation-targeting central banks — could not be distinguished statistically from random chance.

    3. The 700-year pattern suggests that current real rates should be about 0.7%. The authors speculate that the trend may be more likely to flatten out as rates approach negative territory than to continue further downward.
There's much more in this follow-up paper that may be of particular interest to those with statistics and econometrics backgrounds — but it may also be of more general interest just for its financial history. Comments?

Well this seems to run counter to the current narrative that we're seeing a paradigm shift with inflation being more or less the new regime. If this research is true, is it actionable? For example, with 10-year Treasuries yielding around 3.8%, do bonds represent a buying opportunity? VGIT anyone?
sean.mcgrath
Posts: 786
Joined: Thu Dec 29, 2016 5:15 am
Location: US in NL

Re: Update: 700 Years of Falling Interest Rates, 1311-2021

Post by sean.mcgrath »

Frank2012 wrote: Fri Oct 07, 2022 5:28 am
Well this seems to run counter to the current narrative that we're seeing a paradigm shift with inflation being more or less the new regime. If this research is true, is it actionable? For example, with 10-year Treasuries yielding around 3.8%, do bonds represent a buying opportunity? VGIT anyone?
He's talking about real interest rates, so not clear to me whether it says anything about inflation. I think it could be actionable in terms of setting expectations for "expected" real risk-free interest going forward. But even there, seven hundred year trend lines will never eliminate short term random deviations, of course. I'd see it as a very general "expectation."
ScubaHogg
Posts: 3386
Joined: Sun Nov 06, 2011 2:02 pm

Re: Update: 700 Years of Falling Interest Rates, 1311-2021

Post by ScubaHogg »

I haven’t read this paper yet. But I have read A History of Interest Rates by Homer and Sylla. Without looking up the exact dates and numbers, the interesting part is that European interest rates had the opposite trend for the millennium or so prior to this paper above. Indeed, from somewhere around the 1st or 2nd century CE in Rome, interest rates rose from a number we would be familiar with (low single digits) to the large numbers you see at the start of this graph above.
There are more things in Heaven and Earth, Horatio, than are dreamt of in your Expected Returns
sean.mcgrath
Posts: 786
Joined: Thu Dec 29, 2016 5:15 am
Location: US in NL

Re: Update: 700 Years of Falling Interest Rates, 1311-2021

Post by sean.mcgrath »

ScubaHogg wrote: Fri Oct 07, 2022 5:54 am I haven’t read this paper yet. But I have read A History of Interest Rates by Homer and Sylla. Without looking up the exact dates and numbers, the interesting part is that European interest rates had the opposite trend for the millennium or so prior to this paper above. Indeed, from somewhere around the 1st or 2nd century CE in Rome, interest rates rose from a number we would be familiar with (low single digits) to the large numbers you see at the start of this graph above.
And did that one have a theory about the mechanisms? I assume the collapse of Rome.
toddthebod
Posts: 5218
Joined: Wed May 18, 2022 12:42 pm

Re: Update: 700 Years of Falling Interest Rates, 1311-2021

Post by toddthebod »

sean.mcgrath wrote: Fri Oct 07, 2022 5:27 am I was just about to post on this!

"The authors speculate that the trend may be more likely to flatten out as rates approach negative territory than to continue further downward." However, they admit that they have not yet seen any evidence for this.

The interesting question to me is the mechanism. Admittedly I only skimmed the papers, but the only stab at this I was able to find was "though capital accumulation trends may go some way in explaining the phenomenon"

Do you have any guesses about what the mechanism might be?
That sounds like a supply-and-demand argument. Capital accumulation means people with lots of money sitting in the bank losing real value. It also implies the minimum theoretical interest rate would be 0% nominal.
Backtests without cash flows are meaningless. Returns without dividends are lies.
homebuyer6426
Posts: 1807
Joined: Tue Feb 07, 2017 8:08 am

Re: Update: 700 Years of Falling Interest Rates, 1311-2021

Post by homebuyer6426 »

Interesting for sure. How did one go about buying a bond in 1300? I assume this type of activity would be limited to noblemen? Can the data apply broadly in an age where investing is accessible to many more people?
41% Total Stock Market | 54% Consumer Staples | 5% Short Term Reserves
Tamalak
Posts: 1956
Joined: Fri May 06, 2016 2:29 pm

Re: Update: 700 Years of Falling Interest Rates, 1311-2021

Post by Tamalak »

IMO the mechanism is

1. People are living longer
2. Catastrophic scenarios (destruction of the government etc) are less likely

Both those things lower the time value of money. People are willing to make money more slowly vs spending it now.

Once technology advances to the point where we can reverse aging, and our lifespans are indefinite, I'm pretty sure the real interest rate will turn negative.
User avatar
Topic Author
SimpleGift
Posts: 4477
Joined: Tue Feb 08, 2011 2:45 pm

Re: Update: 700 Years of Falling Interest Rates, 1311-2021

Post by SimpleGift »

Just to note that the original study's entire database — including 700 years of interest rates, inflation data, weighted GDP data and other time series — is freely available as an Excel file, downloaded here.

This database is a compilation of painstakingly gathered data from original sources going back centuries in Italy, the United Kingdom, Holland, Germany, France, Spain, Japan and the United States.
User avatar
Topic Author
SimpleGift
Posts: 4477
Joined: Tue Feb 08, 2011 2:45 pm

Re: Update: 700 Years of Falling Interest Rates, 1311-2021

Post by SimpleGift »

sean.mcgrath wrote: Fri Oct 07, 2022 5:27 am Do you have any guesses about what the mechanism might be?
To begin to get at the question of WHY interest rates have been falling for 700 years, the authors in this paper find a negative relationship between falling real rates and:
  • • Population growth trends
    • GDP output trends
Since the authors use the phrase "suprasecular decline" in rates (i.e., "existing or continuing through ages or centuries"), we perhaps should be looking at fundamental trends that transcend our normal sense of financial history, such as:
  • • The world has gotten steadily wealthier since the 1300s (GDP-per-capita has increased), so capital has become less scarce and thus the cost to borrow it has steadily fallen.

    • The wealth destroying forces of war, famine and epidemic have gradually lessened over the centuries, leading to a safer world where lending money has become a less risky proposition.

    • The overall "societal risk premium" has steadily declined, as global societies have become more stable, prosperous, and interconnected — and better at more accurately pricing risk.
The "WHY" question behind 700 years of falling real rates is bound to be a fertile field of financial research and inquiry for some years to come.
Last edited by SimpleGift on Fri Oct 07, 2022 9:08 am, edited 1 time in total.
AlohaJoe
Posts: 6608
Joined: Mon Nov 26, 2007 1:00 pm
Location: Saigon, Vietnam

Re: Update: 700 Years of Falling Interest Rates, 1311-2021

Post by AlohaJoe »

Tamalak wrote: Fri Oct 07, 2022 8:12 am IMO the mechanism is

1. People are living longer
I'm not sure this one matches the pattern we see in the rate declines.

"from 1200 to 1745, 21-year-olds would reach an average age of anywhere between 62 and 70 years – except for the 14th Century, when the bubonic plague cut life expectancy to a paltry 45".

Nowadays, for a 21 year old, it is up to 77 years (for a man). So anywhere from 7 to 15 years improvement. I suppose that might be enough to trigger a change in real rates but I'm a bit dubious. Virtually all of the increase in life expectancy has come from reduced infant mortality. Since infants aren't buying bonds, it seems unlikely that mattered much.

Also, the shift didn't really happen until after 1750 but the research shows that bond yields kept dropping even before that. So changing life expectancy couldn't have been a driving force before 1750. Plus, we'd expect to see a step-change after 1750, which we don't really see, if life expectancy was a large contributing factor.
User avatar
Topic Author
SimpleGift
Posts: 4477
Joined: Tue Feb 08, 2011 2:45 pm

Re: Update: 700 Years of Falling Interest Rates, 1311-2021

Post by SimpleGift »

Frank2012 wrote: Fri Oct 07, 2022 5:28 am If this research is true, is it actionable?
What is new in this research is the statistical identification of a stationary, declining trend in real interest rates (-0.016% per year), a falling trend line that rates have tended to revert back towards over long periods.

What's actionable is that investors in the 21st century should not have expectations that real interest rates will revert back to their historical averages of the 20th century (say 3% or 4%) — but rather should expect that real rates will remain low, perhaps becoming negative, for extended periods in the decades ahead.

Shorter periods of higher real rates can and will occur (like today), but the persistent trend is toward low real rates.
ScubaHogg
Posts: 3386
Joined: Sun Nov 06, 2011 2:02 pm

Re: Update: 700 Years of Falling Interest Rates, 1311-2021

Post by ScubaHogg »

sean.mcgrath wrote: Fri Oct 07, 2022 6:32 am
ScubaHogg wrote: Fri Oct 07, 2022 5:54 am I haven’t read this paper yet. But I have read A History of Interest Rates by Homer and Sylla. Without looking up the exact dates and numbers, the interesting part is that European interest rates had the opposite trend for the millennium or so prior to this paper above. Indeed, from somewhere around the 1st or 2nd century CE in Rome, interest rates rose from a number we would be familiar with (low single digits) to the large numbers you see at the start of this graph above.
And did that one have a theory about the mechanisms? I assume the collapse of Rome.
I think yes, that was the general story. As governments and trade collapsed (and everything that goes with it) the value of promises by governments went down and the risk of loaning out capital went up accordingly.

Though to be fair, in my memory, the authors were mostly concerned with what happened, as opposed to theorizing why it happened. I could be mistaken.
There are more things in Heaven and Earth, Horatio, than are dreamt of in your Expected Returns
ScubaHogg
Posts: 3386
Joined: Sun Nov 06, 2011 2:02 pm

Re: Update: 700 Years of Falling Interest Rates, 1311-2021

Post by ScubaHogg »

SimpleGift wrote: Fri Oct 07, 2022 9:48 am
Frank2012 wrote: Fri Oct 07, 2022 5:28 am If this research is true, is it actionable?
What is new in this research is the statistical identification of a stationary, declining trend in real interest rates (-0.016% per year), a falling trend line that rates have tended to revert back towards over long periods.

What's actionable is that investors in the 21st century should not have expectations that real interest rates will revert back to their historical averages of the 20th century (say 3% or 4%) — but rather should expect that real rates will remain low, perhaps becoming negative, for extended periods in the decades ahead.

Shorter periods of higher real rates can and will occur (like today), but the persistent trend is toward low real rates.
I can’t see that in the graph. The individuals experience looks to be overwhelmed with the noise of a single lifetime.
There are more things in Heaven and Earth, Horatio, than are dreamt of in your Expected Returns
User avatar
Topic Author
SimpleGift
Posts: 4477
Joined: Tue Feb 08, 2011 2:45 pm

Re: Update: 700 Years of Falling Interest Rates, 1311-2021

Post by SimpleGift »

ScubaHogg wrote: Fri Oct 07, 2022 10:36 am The individuals experience looks to be overwhelmed with the noise of a single lifetime.
Right, the actual experience of a 30-, or 40-, or 50-year investing career will have a lot of variance in the century ahead — but, based on this research, the expectation for low real rates should probably be what informs financial planning.
ScubaHogg
Posts: 3386
Joined: Sun Nov 06, 2011 2:02 pm

Re: Update: 700 Years of Falling Interest Rates, 1311-2021

Post by ScubaHogg »

SimpleGift wrote: Fri Oct 07, 2022 10:46 am
ScubaHogg wrote: Fri Oct 07, 2022 10:36 am The individuals experience looks to be overwhelmed with the noise of a single lifetime.
Right, the actual experience of a 30-, or 40-, or 50-year investing career will have a lot of variance in the century ahead — but, based on this research, the expectation for low real rates should probably be what informs financial planning.
I disagree. The decade to decade variations overwhelm by several multiples the minute trend line. It’s a bit like using earths average temperature to decide what I’m going to wear today.

Personally I’d stick with the “I don’t know” idea and go from there (ie, just don’t time interest rates)
There are more things in Heaven and Earth, Horatio, than are dreamt of in your Expected Returns
dertere
Posts: 83
Joined: Fri Sep 03, 2021 12:23 pm

Re: Update: 700 Years of Falling Interest Rates, 1311-2021

Post by dertere »

One thing I find interesting is the noticeable decrease in dispersion between different countries in the post-war era. Even during the relatively peaceful period for Europe between the end of the 30-Years War and the beginning of the French Revolution and Napoleonic Wars interest rates for each state were much less correlated.
homebuyer6426 wrote: Fri Oct 07, 2022 7:41 am Interesting for sure. How did one go about buying a bond in 1300? I assume this type of activity would be limited to noblemen? Can the data apply broadly in an age where investing is accessible to many more people?
The Church, Italian and French banking families, and Jewish merchant communities were all fairly common and significant non-noble sources of lending for most European monarchs in 1300. According to the authors by 1450's debt markets are fairly liquid, though I don't have a strong enough understanding of the material to grasp the implications of what they cite:
And the evidence suggests that these markets were remarkably liquid: for instance, by the mid-15th century, the face value of annually-traded
long-maturity public debt in hubs such as Venice reaches up to 29% of GDP.
User avatar
Topic Author
SimpleGift
Posts: 4477
Joined: Tue Feb 08, 2011 2:45 pm

Re: Update: 700 Years of Falling Interest Rates, 1311-2021

Post by SimpleGift »

dertere wrote: Fri Oct 07, 2022 2:16 pm One thing I find interesting is the noticeable decrease in dispersion between different countries in the post-war era. Even during the relatively peaceful period for Europe between the end of the 30-Years War and the beginning of the French Revolution and Napoleonic Wars interest rates for each state were much less correlated.
Good observation. This was yet another finding in Schmelzing's original paper. Not only have real interest rates themselves been declining, but there's been a gradual, but persistent trend toward lower volatility and higher correlation of real rates in the developed world (chart below).
  • Image
    NOTE: Standard deviations are 30-year rolling averages. Data source: Bank of England
On the scale of centuries, global real interest rates have not just been trending lower, but have become "stickier" with time. One more reason one might expect low or even negative real rates to become persistent in the 21st century.

PS. This also puts the turbulence of the early 20th century in historical perspective. With the Great Depression and two world wars, it was a hugely volatile time for financial assets in the developed world — even compared with previous centuries.
Last edited by SimpleGift on Fri Oct 07, 2022 4:20 pm, edited 1 time in total.
User avatar
Kevin M
Posts: 15665
Joined: Mon Jun 29, 2009 3:24 pm
Contact:

Re: Update: 700 Years of Falling Interest Rates, 1311-2021

Post by Kevin M »

If we believe this, then long-term TIPS at 2% today are a screaming buy. I think I'll buy some when the seasonally-adjusted yield hits 2.03%, which is about 2.0% ask. Now the 2045 TIPS is at 2.007% ask for min qty 10.
If I make a calculation error, #Cruncher probably will let me know.
User avatar
packer16
Posts: 1488
Joined: Sat Jan 04, 2014 1:28 pm

Re: Update: 700 Years of Falling Interest Rates, 1311-2021

Post by packer16 »

SimpleGift wrote: Fri Oct 07, 2022 3:39 pm
dertere wrote: Fri Oct 07, 2022 2:16 pm One thing I find interesting is the noticeable decrease in dispersion between different countries in the post-war era. Even during the relatively peaceful period for Europe between the end of the 30-Years War and the beginning of the French Revolution and Napoleonic Wars interest rates for each state were much less correlated.
Good observation. This was yet another finding in Schmelzing's original paper. Not only have real interest rates themselves been declining, but there's been a gradual, but persistent trend toward lower volatility and higher correlation of real rates in the developed world (chart below).
  • Image
    NOTE: Standard deviations are 30-year rolling averages. Data source: Bank of England
On the scale of centuries, global real interest rates have not just been trending lower, but have become "stickier" with time. One more reason one might expect low or even negative real rates to become persistent in the 21st century.

PS. This also puts the turbulence of the early 20th century in historical perspective. With the Great Depression and two world wars, it was a hugely volatile time for financial assets in the developed world — even compared with previous centuries.
This makes sense in terms of creditworthiness. Before the end of WWII, you primarily had autocracies whose creditworthiness is dependent upon the leader in charge vs. democracies where the credit was based upon the taxing ability of the freely elected government.

Packer
Buy cheap and something good might happen
McQ
Posts: 1391
Joined: Fri Jun 18, 2021 12:21 am
Location: California

Re: Update: 700 Years of Falling Interest Rates, 1311-2021

Post by McQ »

SimpleGift wrote: Fri Oct 07, 2022 2:04 am This post reports on a new follow-up paper to the groundbreaking 2019 study of the "supra-secular" decline in real interest rates over 700 years, since the 14th century (chart below). The original paper is discussed in this Forum thread.
  • Image
    Note: Trend line in black is GDP-weighted among countries. Chart source: The Economist
While the original paper was quite readable, this follow-up paper is more statistically oriented. Some key findings:
  • 1. Global real, long-term rates exhibit a gentle but firm downward trend (averaging about -1.6% every 100 years), and the evidence is strongly consistent with the hypothesis that they are stationary around this trend. In other words, rates tended to mean revert around this trend over long periods.

    2. The only two events to disrupt the trend were the Black Death and a wave of sovereign-debt defaults in the late 1550s. All other deviations — including after the founding of America’s Federal Reserve in 1913 and the advent of inflation-targeting central banks — could not be distinguished statistically from random chance.

    3. The 700-year pattern suggests that current real rates should be about 0.7%. The authors speculate that the trend may be more likely to flatten out as rates approach negative territory than to continue further downward.
There's much more in this follow-up paper that may be of particular interest to those with statistics and econometrics backgrounds — but it may also be of more general interest just for its financial history. Comments?
Much obliged, SimpleGift, I had not seen the update.

Those interested in the action implications, per Kevin M, may find this thread of use: viewtopic.php?t=386240
You can take the academic out of the classroom by retirement, but you can't ever take the classroom out of his tone, style, and manner of approach.
User avatar
Topic Author
SimpleGift
Posts: 4477
Joined: Tue Feb 08, 2011 2:45 pm

Re: Update: 700 Years of Falling Interest Rates, 1311-2021

Post by SimpleGift »

Reading further into this paper, the most surprising result so far has been the authors' assertion that:
Rogoff et al. wrote:...long-term real interest rates series appear to be trend stationary; and regardless of the underlying cause, the analysis does suggest that when real interest rates fall far below the (declining) long-term trend, there is a high chance of mean reversion — unless of course there has been a rare structural break.
Question: Are there any other data series in investing or finance that have been statistically proven to be trend stationary, with predictable reversion to the mean — as opposed to a random walk? I'm aware the researchers here had 700 years of data to work with, which is of course not available in many other cases. Perhaps readers with a deeper background in statistics or econometrics will weigh in on this question.
User avatar
whodidntante
Posts: 13013
Joined: Thu Jan 21, 2016 10:11 pm
Location: outside the echo chamber

Re: Update: 700 Years of Falling Interest Rates, 1311-2021

Post by whodidntante »

SimpleGift wrote: Fri Oct 07, 2022 5:45 pm Reading further into this paper, the most surprising result so far has been the authors' assertion that:
Rogoff et al. wrote:...long-term real interest rates series appear to be trend stationary; and regardless of the underlying cause, the analysis does suggest that when real interest rates fall far below the (declining) long-term trend, there is a high chance of mean reversion — unless of course there has been a rare structural break.
Question: Are there any other data series in investing or finance that have been statistically proven to be trend stationary, with predictable reversion to the mean — as opposed to a random walk? I'm aware the researchers here had 700 years of data to work with, which is of course not available in many other cases. Perhaps readers with a deeper background in statistics or econometrics will weigh in on this question.
Bogle argued that growth and value mean revert. He had a chart and everything. :P
User avatar
Topic Author
SimpleGift
Posts: 4477
Joined: Tue Feb 08, 2011 2:45 pm

Re: Update: 700 Years of Falling Interest Rates, 1311-2021

Post by SimpleGift »

whodidntante wrote: Fri Oct 07, 2022 5:53 pm Bogle argued that growth and value mean revert. He had a chart and everything.
Yes, but looking for evidence of trend stationarity that's been proven with a similar degree of rigor as the OP paper:
Rogoff et al. wrote:...we find unambiguous evidence in favor of trend stationarity for global real rates when tested via the ADF-GLS test. The null hypothesis of a unit root can be rejected at the 5 percent significance level for the headline real series, for all headline real rate country series, and for an arithmetically-weighted global series. Specifically, we also note that the evidence in favor of trend stationarity is confirmed for U.S. and U.K. country-level data.
Not familiar with "Augmented Dickey-Fuller" tests (ADF-GLS), or the "unit root" — but it sounds state-of-the-art. :wink:
simplextableau
Posts: 435
Joined: Wed Oct 05, 2016 9:20 pm

Re: Update: 700 Years of Falling Interest Rates, 1311-2021

Post by simplextableau »

SimpleGift wrote: Fri Oct 07, 2022 7:55 pm
whodidntante wrote: Fri Oct 07, 2022 5:53 pm Bogle argued that growth and value mean revert. He had a chart and everything.
Yes, but looking for evidence of trend stationarity that's been proven with a similar degree of rigor as the OP paper:
Rogoff et al. wrote:...we find unambiguous evidence in favor of trend stationarity for global real rates when tested via the ADF-GLS test. The null hypothesis of a unit root can be rejected at the 5 percent significance level for the headline real series, for all headline real rate country series, and for an arithmetically-weighted global series. Specifically, we also note that the evidence in favor of trend stationarity is confirmed for U.S. and U.K. country-level data.
Not familiar with "Augmented Dickey-Fuller" tests (ADF-GLS), or the "unit root" — but it sounds state-of-the-art. :wink:
I remember doing those in college. Always good for a snicker the first time the method was introduced :D
finite_difference
Posts: 3601
Joined: Thu Jul 09, 2015 7:00 pm

Re: Update: 700 Years of Falling Interest Rates, 1311-2021

Post by finite_difference »

SimpleGift wrote: Fri Oct 07, 2022 9:48 am
Frank2012 wrote: Fri Oct 07, 2022 5:28 am If this research is true, is it actionable?
What is new in this research is the statistical identification of a stationary, declining trend in real interest rates (-0.016% per year), a falling trend line that rates have tended to revert back towards over long periods.

What's actionable is that investors in the 21st century should not have expectations that real interest rates will revert back to their historical averages of the 20th century (say 3% or 4%) — but rather should expect that real rates will remain low, perhaps becoming negative, for extended periods in the decades ahead.

Shorter periods of higher real rates can and will occur (like today), but the persistent trend is toward low real rates.
Shouldn’t that be “real bond yields” and not “real interest rates”?

The term “real interest rate” is rather confusing to me, but I guess it’s like the interest rate a bank would pay you, except in real terms and not nominal terms.

By “real bond yields” do they mean “real total return of bonds”?

Don’t we always assume 0% real return for bonds and CDs and treasuries?
The most precious gift we can offer anyone is our attention. - Thich Nhat Hanh
User avatar
Topic Author
SimpleGift
Posts: 4477
Joined: Tue Feb 08, 2011 2:45 pm

Re: Update: 700 Years of Falling Interest Rates, 1311-2021

Post by SimpleGift »

finite_difference wrote: Fri Oct 07, 2022 9:34 pm Shouldn’t that be “real bond yields” and not “real interest rates”?
The authors use the term "real interest rates." Their data set is primarily the historical nominal rates on sovereign long-term bonds issued by eight developed countries (bonds that approximate the modern 10-year Treasury bond in maturity) — which are then corrected for the prevailing inflation rate.

Since the yields on new investments in sovereign debt reflect the interest rates at the time they are issued, bond yields and interest rates would seem to be the same in this context. The paper doesn't address total bond returns.
Frank2012
Posts: 194
Joined: Fri Jan 25, 2013 6:16 pm

Re: Update: 700 Years of Falling Interest Rates, 1311-2021

Post by Frank2012 »

SimpleGift wrote: Fri Oct 07, 2022 9:48 am
Frank2012 wrote: Fri Oct 07, 2022 5:28 am If this research is true, is it actionable?
What is new in this research is the statistical identification of a stationary, declining trend in real interest rates (-0.016% per year), a falling trend line that rates have tended to revert back towards over long periods.

What's actionable is that investors in the 21st century should not have expectations that real interest rates will revert back to their historical averages of the 20th century (say 3% or 4%) — but rather should expect that real rates will remain low, perhaps becoming negative, for extended periods in the decades ahead.

Shorter periods of higher real rates can and will occur (like today), but the persistent trend is toward low real rates.


Excellent post and thread, thanks!
User avatar
Topic Author
SimpleGift
Posts: 4477
Joined: Tue Feb 08, 2011 2:45 pm

Re: Update: 700 Years of Falling Interest Rates, 1311-2021

Post by SimpleGift »

A friend has made me aware that, Paul Schmelzing, the main author of the OP study, appeared on a podcast earlier this year, talking about this research and several of the issues raised in this thread:
The podcast is 54 minutes long, and the financial history may be of interest. A written transcript is included on the podcast site, but it must have been done by machine, untouched by human hands, as it's a bit of a mess.

Also a recent written interview with Schmelzing: Lessons from Real Interest Rates Since the 14th Century
Last edited by SimpleGift on Mon Oct 10, 2022 10:52 pm, edited 1 time in total.
CletusCaddy
Posts: 2678
Joined: Sun Sep 12, 2021 4:23 am

Re: Update: 700 Years of Falling Interest Rates, 1311-2021

Post by CletusCaddy »

finite_difference wrote: Fri Oct 07, 2022 9:34 pmDon’t we always assume 0% real return for bonds and CDs and treasuries?
Why would anyone assume this? Why would anyone buy those if they were zero real return?
finite_difference
Posts: 3601
Joined: Thu Jul 09, 2015 7:00 pm

Re: Update: 700 Years of Falling Interest Rates, 1311-2021

Post by finite_difference »

CletusCaddy wrote: Mon Oct 10, 2022 10:51 pm
finite_difference wrote: Fri Oct 07, 2022 9:34 pmDon’t we always assume 0% real return for bonds and CDs and treasuries?
Why would anyone assume this? Why would anyone buy those if they were zero real return?
1. There have been many threads on this topic, and best I can tell, looking forward, the majority of Bogleheads expect 4% real return from stocks and 0% real return from bonds.

What do you think Bogleheads expect? I think we are a pretty conservative bunch.

2. Because zero real return is much better than negative real return (keeping as cash)?

Zero real return is awesome! Anything better than that is icing on cake.
The most precious gift we can offer anyone is our attention. - Thich Nhat Hanh
finite_difference
Posts: 3601
Joined: Thu Jul 09, 2015 7:00 pm

Re: Update: 700 Years of Falling Interest Rates, 1311-2021

Post by finite_difference »

SimpleGift wrote: Fri Oct 07, 2022 10:56 pm
finite_difference wrote: Fri Oct 07, 2022 9:34 pm Shouldn’t that be “real bond yields” and not “real interest rates”?
The authors use the term "real interest rates." Their data set is primarily the historical nominal rates on sovereign long-term bonds issued by eight developed countries (bonds that approximate the modern 10-year Treasury bond in maturity) — which are then corrected for the prevailing inflation rate.

Since the yields on new investments in sovereign debt reflect the interest rates at the time they are issued, bond yields and interest rates would seem to be the same in this context. The paper doesn't address total bond returns.
Thanks for the clarification! I’m getting used to it. Interest rates are so low these days and always in nominal terms, so that the term “real interest rate”, which is here a positive value, threw me into a loop.

The y-axis label of the chart was also different which confused me, but your explanation helps a lot.

I suppose over long periods and if you hold to maturity the real bond yields will equal real interest rates? So the total return is approximated by the yield if you hold to maturity?
The most precious gift we can offer anyone is our attention. - Thich Nhat Hanh
Valuethinker
Posts: 48773
Joined: Fri May 11, 2007 11:07 am

Re: Update: 700 Years of Falling Interest Rates, 1311-2021

Post by Valuethinker »

dertere wrote: Fri Oct 07, 2022 2:16 pm One thing I find interesting is the noticeable decrease in dispersion between different countries in the post-war era. Even during the relatively peaceful period for Europe between the end of the 30-Years War and the beginning of the French Revolution and Napoleonic Wars interest rates for each state were much less correlated.
homebuyer6426 wrote: Fri Oct 07, 2022 7:41 am Interesting for sure. How did one go about buying a bond in 1300? I assume this type of activity would be limited to noblemen? Can the data apply broadly in an age where investing is accessible to many more people?
The Church, Italian and French banking families, and Jewish merchant communities were all fairly common and significant non-noble sources of lending for most European monarchs in 1300. According to the authors by 1450's debt markets are fairly liquid, though I don't have a strong enough understanding of the material to grasp the implications of what they cite:
And the evidence suggests that these markets were remarkably liquid: for instance, by the mid-15th century, the face value of annually-traded
long-maturity public debt in hubs such as Venice reaches up to 29% of GDP.
Finally a topic I know/ knew something about.

The Bardi and the Frescobaldi financed the wars of the King Edwards I, II & III of England - the conquest of Wales and the abortive conquest of Scotland. You can see their chapels in the Church of St Croce in Firenze (Florence). They were Italian wool merchants (from Lucca) and at least one of the Edwards defaulted on his debts (?from memory).

The origins of many anti-Semitic myths originated in the Jewish role in finance. You couldn't be a warrior or priest, but you could go into trade. Trade means finance. International trade means international finance.

If you didn't want to pay your debts, starting a pogrom against your lenders was not a bad strategy.

Philip the something of France pulled something similar on the Knights Templar - accused them of satanic possession and devil worship. Dissolved the order and seized all its extensive properties. By and large, this was easier than actually leaving the Roman Catholic church to get hold of all monastic lands (Henry VIII of England's Dissolution of the Monasteries who owned something like 1/3rd of all the land in the Kingdom - he also badly needed a divorce and the Pope, under the sway of the ?Spanish? king, refused to grant it).

A big distinction in the early modern period (say up to 1700) and indeed up to the Napoleonic period (to 1815) was the ability of certain states (Britain, and Netherlands in particular) to raise finance more regularly and cheaply than their opponents. Armies and fleets were expensive to both assemble and to keep in being - they had a bad habit of mutineering if unpaid. The formation of the Bank of England in 1696 was a key strategic advantage in the long century of wars with France that followed - whereas the French King was always going broke.

A strong and stable system of government finance was a big part of what allowed the "nation of shopkeepers" (Napoleon's derisive description) to conquer & control more of the planet's surface than any Empire before or since. (The American mechanism of Empire (or arguably hegemony) has been different). Perfidious Albion.

We had "perpetual consols" in circulation. A refinancing of the 1854 (?) Crimean War Loan at (from memory) 2 3/4% coupon rate, no redemption date. Eventually the Chancellor post 2010 (George Osborne) had them compulsory purchased (there were only a few million £ still in circulation) which I think was a great pity from an Economic Historical perspective.
User avatar
Topic Author
SimpleGift
Posts: 4477
Joined: Tue Feb 08, 2011 2:45 pm

Re: Update: 700 Years of Falling Interest Rates, 1311-2021

Post by SimpleGift »

The greatest contribution of this research on interest rates over seven centuries, to my mind, is dispelling the idea that there is a steady, intrinsic "normal rate of interest." This notion must be deeply embedded somewhere in the collective unconscious, as evidenced by the many Forum posts to the tune of "When will interest rates get back to normal?"

In one of his interviews, Paul Schmelzing, the study's primary author, discusses this contradiction:
Paul Schmelzing wrote:The big underlying assumption in all of these contributions and all of these discussions was that we had a “normal” level of interest rates during the 1960s, and that level got shocked by the oil crisis in the ’70s and ’80s. [The assumption was that] we just have to find the right policy tool, the right fiscal tool, to get back to that normal level of interest, which has always existed and which we want to go back to, and which is the ideal steady state interest rate level.

That is also a big idea that is formalized in a lot of the finance models and a lot of the underlying economic models/growth models. It has a really foundational role in a lot of the literature—the idea that there is a natural level of interest rates. I would say you can at least trace it back to Irving Fisher and his 1908 book, The Rate of Interest, where he says something very similar. He doesn’t call it the steady state interest rate, but he says there is a natural level of interest rates which should be related to the consumption growth rate.
Perhaps the key takeaway from this research is that there is no steady, intrinsic, normal rate of interest — and that the long-term historical trend should continue to put downward pressure on interest rates through the 21st century.
CletusCaddy
Posts: 2678
Joined: Sun Sep 12, 2021 4:23 am

Re: Update: 700 Years of Falling Interest Rates, 1311-2021

Post by CletusCaddy »

The key takeaway for me is we should all be loading up on long term TIPS yielding 2% real right now
CletusCaddy
Posts: 2678
Joined: Sun Sep 12, 2021 4:23 am

Re: Update: 700 Years of Falling Interest Rates, 1311-2021

Post by CletusCaddy »

finite_difference wrote: Tue Oct 11, 2022 9:24 am
CletusCaddy wrote: Mon Oct 10, 2022 10:51 pm
finite_difference wrote: Fri Oct 07, 2022 9:34 pmDon’t we always assume 0% real return for bonds and CDs and treasuries?
Why would anyone assume this? Why would anyone buy those if they were zero real return?
1. There have been many threads on this topic, and best I can tell, looking forward, the majority of Bogleheads expect 4% real return from stocks and 0% real return from bonds.

What do you think Bogleheads expect? I think we are a pretty conservative bunch.

2. Because zero real return is much better than negative real return (keeping as cash)?

Zero real return is awesome! Anything better than that is icing on cake.
Academic practitioners like Illamen expect 1-2% real returns out of bonds going forward. I will trust that over Bogleheads any day.

Not to mention TIPS right now are yielding 2% real which you can lock in for 20 years.
ryman554
Posts: 1627
Joined: Sun Jan 12, 2014 8:44 pm

Re: Update: 700 Years of Falling Interest Rates, 1311-2021

Post by ryman554 »

So how does one exactly lock in 2% real TIPS? Are these 20year bonds? Or are there ladders of 1-19 years to choose from?
User avatar
Kevin M
Posts: 15665
Joined: Mon Jun 29, 2009 3:24 pm
Contact:

Re: Update: 700 Years of Falling Interest Rates, 1311-2021

Post by Kevin M »

ryman554 wrote: Wed Oct 12, 2022 4:07 pm So how does one exactly lock in 2% real TIPS? Are these 20year bonds? Or are there ladders of 1-19 years to choose from?
There are TIPS maturing in Feb of every year between 2/15/40 and 2/15/52. Here are ask yields for all 49 TIPS at Fidelity today:

Image

You can pretty much ignore the seasonally-adjusted (SA) values for longer-term TIPS.

The 2043 through 2047 had yields of about 2% today.

Kevin
If I make a calculation error, #Cruncher probably will let me know.
Valuethinker
Posts: 48773
Joined: Fri May 11, 2007 11:07 am

Re: Update: 700 Years of Falling Interest Rates, 1311-2021

Post by Valuethinker »

ryman554 wrote: Wed Oct 12, 2022 4:07 pm So how does one exactly lock in 2% real TIPS? Are these 20year bonds? Or are there ladders of 1-19 years to choose from?
You cannot. At least as a retail investor.

What you would need would be a "0 coupon TIPS" ie which sold at a discount to par (=100) value and then would only pay a return = 100 + inflation accretion, and only at maturity.

That solves the problem of the reinvestment of bond coupons - the Yield to Maturity is an approximation (for all bonds that are not zero coupon) for this reason.

However with TIPS, a big part of the return is the inflation adjustment (accretion) of the principal at redemption (it redeems at the higher of 100 or the accretion + 100). Less is in the coupon than with a normal bond.

So the "reinvestment risk" is relatively low. Not nothing, but low. The Yield To Maturity is still the best estimate of likely returns.

If you buy & hold 20 year TIPS, and the real yield (the ask yield - remember you get the higher of the 2 prices quoted) is 2.0% or above then you have a pretty good chance of locking in that return (pre tax).
User avatar
Topic Author
SimpleGift
Posts: 4477
Joined: Tue Feb 08, 2011 2:45 pm

Re: Update: 700 Years of Falling Interest Rates, 1311-2021

Post by SimpleGift »

JoMoney wrote: Thu Oct 13, 2022 8:30 am Even on sovereign debts there are many historical defaults, and what I'm interested in is one looking at the hundreds of years the declining interest rates charts are showing. Maybe the financial world is more stable now than it was when the compared interest rates were higher.
The interest rates series used in this study is a GDP-weighted average of all available advanced economy, long-term sovereign debt yields. The study aims to derive a historical interest rate series for the global "safe asset rate," to sidestep the issue of default risk over time (see below). The methodology is detailed in the original Schmelzing paper.
Paul Schmelzing wrote:[...my overall long-term real series is for the historical “safe asset provider”.... Its central feature consists in the fact that it remained entirely default-free over its 707-year span, and that it can with recourse to historical documents be shown to have represented the preferred “safe” investment asset over time.
There are studies of historical sovereign defaults over centuries, I believe, but this one doesn't address them.
User avatar
Kevin M
Posts: 15665
Joined: Mon Jun 29, 2009 3:24 pm
Contact:

Re: Update: 700 Years of Falling Interest Rates, 1311-2021

Post by Kevin M »

Valuethinker wrote: Thu Oct 13, 2022 7:58 am
ryman554 wrote: Wed Oct 12, 2022 4:07 pm So how does one exactly lock in 2% real TIPS? Are these 20year bonds? Or are there ladders of 1-19 years to choose from?
You cannot. At least as a retail investor.

What you would need would be a "0 coupon TIPS" ie which sold at a discount to par (=100) value and then would only pay a return = 100 + inflation accretion, and only at maturity.

That solves the problem of the reinvestment of bond coupons - the Yield to Maturity is an approximation (for all bonds that are not zero coupon) for this reason.
This is false. It's true that your investment return may not equal your original YTM because of different reinvestment rates, but it is not true that YTM is just an estimate.

YTM is the constant discount rate that equalizes price and present value of future cash flows. Period.

Your internal rate of return (IRR) will be almost the same as your initial YTM, with a slight difference because of the way YTM is calculated for bonds. YTM and IRR do not require reinvestment of cash flows--they are simply time value of money results. A cash flow received earlier has a higher present value than one received later if the yield (discount rate) is positive.

Image

Source: https://www.economics-finance.org/jefe/ ... lpaper.pdf

If you want more, see this also:

http://economics-finance.org/jefe/issue ... ypaper.pdf

Kevin
If I make a calculation error, #Cruncher probably will let me know.
CletusCaddy
Posts: 2678
Joined: Sun Sep 12, 2021 4:23 am

Re: Update: 700 Years of Falling Interest Rates, 1311-2021

Post by CletusCaddy »

Kevin M wrote: Thu Oct 13, 2022 1:19 pm
Valuethinker wrote: Thu Oct 13, 2022 7:58 am
ryman554 wrote: Wed Oct 12, 2022 4:07 pm So how does one exactly lock in 2% real TIPS? Are these 20year bonds? Or are there ladders of 1-19 years to choose from?
You cannot. At least as a retail investor.

What you would need would be a "0 coupon TIPS" ie which sold at a discount to par (=100) value and then would only pay a return = 100 + inflation accretion, and only at maturity.

That solves the problem of the reinvestment of bond coupons - the Yield to Maturity is an approximation (for all bonds that are not zero coupon) for this reason.
This is false. It's true that your investment return may not equal your original YTM because of different reinvestment rates, but it is not true that YTM is just an estimate.

YTM is the constant discount rate that equalizes price and present value of future cash flows. Period.

Your internal rate of return (IRR) will be almost the same as your initial YTM, with a slight difference because of the way YTM is calculated for bonds. YTM and IRR do not require reinvestment of cash flows--they are simply time value of money results. A cash flow received earlier has a higher present value than one received later if the yield (discount rate) is positive.

Image

Source: https://www.economics-finance.org/jefe/ ... lpaper.pdf

If you want more, see this also:

http://economics-finance.org/jefe/issue ... ypaper.pdf

Kevin
Is it accurate to say that the TIPS real yield assumes that the coupons earn zero nominal return?
User avatar
Kevin M
Posts: 15665
Joined: Mon Jun 29, 2009 3:24 pm
Contact:

Re: Update: 700 Years of Falling Interest Rates, 1311-2021

Post by Kevin M »

CletusCaddy wrote: Fri Oct 14, 2022 8:19 am
Kevin M wrote: Thu Oct 13, 2022 1:19 pm
Valuethinker wrote: Thu Oct 13, 2022 7:58 am
ryman554 wrote: Wed Oct 12, 2022 4:07 pm So how does one exactly lock in 2% real TIPS? Are these 20year bonds? Or are there ladders of 1-19 years to choose from?
You cannot. At least as a retail investor.

What you would need would be a "0 coupon TIPS" ie which sold at a discount to par (=100) value and then would only pay a return = 100 + inflation accretion, and only at maturity.

That solves the problem of the reinvestment of bond coupons - the Yield to Maturity is an approximation (for all bonds that are not zero coupon) for this reason.
This is false. It's true that your investment return may not equal your original YTM because of different reinvestment rates, but it is not true that YTM is just an estimate.

YTM is the constant discount rate that equalizes price and present value of future cash flows. Period.

Your internal rate of return (IRR) will be almost the same as your initial YTM, with a slight difference because of the way YTM is calculated for bonds. YTM and IRR do not require reinvestment of cash flows--they are simply time value of money results. A cash flow received earlier has a higher present value than one received later if the yield (discount rate) is positive.

Image

Source: https://www.economics-finance.org/jefe/ ... lpaper.pdf

If you want more, see this also:

http://economics-finance.org/jefe/issue ... ypaper.pdf

Kevin
Is it accurate to say that the TIPS real yield assumes that the coupons earn zero nominal return?
No. The TIPS yield calculation is done on the unadjusted price and coupon rate (before inflation adjustments). One of the inputs to YIELD is the coupon rate, so yield factors in all the real cashflows, including coupon payments. The nominal cash flows are unknown, so can't be used for calculating TIPS yields.

Kevin
If I make a calculation error, #Cruncher probably will let me know.
qwertyjazz
Posts: 2000
Joined: Tue Feb 23, 2016 3:24 am

Re: Update: 700 Years of Falling Interest Rates, 1311-2021

Post by qwertyjazz »

https://open.substack.com/pub/braddelon ... paign=post

Brad Delong’s interesting take on the question. I like how he breaks down 9 elements that drive it
G.E. Box "All models are wrong, but some are useful."
Post Reply