Unexpected Inflation — How Often Does It Happen?

Discuss all general (i.e. non-personal) investing questions and issues, investing news, and theory.
Post Reply
User avatar
Topic Author
SimpleGift
Posts: 3983
Joined: Tue Feb 08, 2011 3:45 pm

Unexpected Inflation — How Often Does It Happen?

Post by SimpleGift »

Inflation seems to be top-of-mind for many investors, if the number of recent Forum posts is an indicator. One question that arose recently, for which I could not find a ready answer was: "How often has unexpected inflation occurred historically?" The context was an investor with a bond-heavy portfolio wondering how much to allocate to TIPS vs. nominal bonds.

TIPS were first auctioned in the U.S. in January of 1997, which give us almost 25 years (293 months) over which to compare monthly, annualized expected inflation rates with actual realized inflation (first chart below). The one-year expected inflation rates (in blue) are from models developed by the Cleveland Federal Reserve.
  • Image
    Data: Monthly one-year expected inflation from Fed models; CPI-U actual inflation from FRED.
The second chart below shows the difference between expected and actual inflation. Bars above zero are when actual inflation exceeded expected inflation (and TIPS would presumably outperform), while bars below zero are the opposite (when conventional nominal bonds would presumably outperform).
  • Image
    Data: Monthly one-year expected inflation from Fed models; CPI-U actual inflation from FRED.
A count of the months in the chart above indicates that inflation was greater-than-expected 54% of the time, and was less-than-expected 46% of the time. So the naive 50/50 allocation between TIPS and nominals so often recommended for bond-heavy portfolios appears to have historical support, if one's goal is to balance risk and return (not liability matching).

Thoughts?
NiceUnparticularMan
Posts: 2525
Joined: Sat Mar 11, 2017 7:51 am

Re: Unexpected Inflation — How Often Does It Happen?

Post by NiceUnparticularMan »

So taking a step back, if your IPS and nominal bonds are just going to cancel each other out in terms of their response to unexpectedly high or unexpectedly low inflation, you really don't have any net unexpected inflation protection either way from this combination of holdings.

Which may indeed be exactly what you want to achieve.

However, I do think it is worth keeping in mind that the study period in question (1997 to present) has been one of relatively low and somewhat stable (at least over longer time periods) inflation overall. Indeed, you can see all the periods of unexpectedly high or unexpectedly low inflation have been either modest or short-lived or both.

And that may well continue indefinitely, but it is also possible we will go through a period of more persistent unexpectedly high or low inflation. In that case, a bond mix which bets the wrong way could more persistently underperform, and even a bond mix which places no bet either way could persistently underperform a bond mix which bets the right way.

And I think the standard Bogleheads logic is to not place such bets, but I personally think there are reasons to be concerned the risks are not completely symmetric, particularly at the tails, in ways which may matter to personal investors but not necessarily as much to other sorts of bond investors.

To be sure, though, that sort of asymmetry has not shown up in a meaningful way from 1997 to present. But that doesn't mean it couldn't show up in the next 25 or so years.
seajay
Posts: 225
Joined: Sat May 01, 2021 3:26 pm

Re: Unexpected Inflation — How Often Does It Happen?

Post by seajay »

NiceUnparticularMan wrote: Sun Jun 13, 2021 4:54 am So taking a step back, if your IPS and nominal bonds are just going to cancel each other out in terms of their response to unexpectedly high or unexpectedly low inflation, you really don't have any net unexpected inflation protection either way from this combination of holdings.
If unexpected inflation sees long dated TIPS gain more than what long dated treasury lose, then like for stocks there may be a positive broad effect. A single stock in ten doubles, another halves, the rest all unchanged and the broad average is positive despite 90% of the stocks having underperformed the average (assuming equally weighted).

The longer the dated the TIPS the greater the price volatility, at the extreme gold might be considered a form of undated zero coupon inflation bond and inclusion of even a little of such within a portfolio has historically seen large/compounded gains that may have hedged the entire portfolio.
seajay
Posts: 225
Joined: Sat May 01, 2021 3:26 pm

Re: Unexpected Inflation — How Often Does It Happen?

Post by seajay »

SimpleGift wrote: Sun Jun 13, 2021 4:08 am A count of the months in the chart above indicates that inflation was greater-than-expected 54% of the time, and was less-than-expected 46% of the time. So the naive 50/50 allocation between TIPS and nominals so often recommended for bond-heavy portfolios appears to have historical support, if one's goal is to balance risk and return (not liability matching).

Thoughts?
That historic interest rate changes are perhaps in part a reflection of having under/over estimated inflation and that broadly counts of interest rate changes for risen/fallen would tend toward 50/50 occurrences.
NiceUnparticularMan
Posts: 2525
Joined: Sat Mar 11, 2017 7:51 am

Re: Unexpected Inflation — How Often Does It Happen?

Post by NiceUnparticularMan »

seajay wrote: Sun Jun 13, 2021 5:20 am
NiceUnparticularMan wrote: Sun Jun 13, 2021 4:54 am So taking a step back, if your IPS and nominal bonds are just going to cancel each other out in terms of their response to unexpectedly high or unexpectedly low inflation, you really don't have any net unexpected inflation protection either way from this combination of holdings.
If unexpected inflation sees long dated TIPS gain more than what long dated treasury lose, then like for stocks there may be a positive broad effect.
Sure, if your bond holdings net out that way. It also means your bond holdings will net out to negative responses to unexpectedly low inflation.

As always, this Vanguard white paper is very helpful on these concepts:

https://advisors.vanguard.com/iwe/pdf/ISGCTIPS.pdf

As an aside, I note short-term TIPS are going to be more responsive to unexpected inflation than long-term TIPS.
The longer the dated the TIPS the greater the price volatility, at the extreme gold might be considered a form of undated zero coupon inflation bond and inclusion of even a little of such within a portfolio has historically seen large/compounded gains that may have hedged the entire portfolio.
That white paper also points out how free-floating gold has not in fact behaved in the way that this poster is predicting it should have behaved.

However, if you are interested in portfolio level hedges against unexpected inflation, it points out you could consider a diversified basket of productive commodities instead.
User avatar
Sandtrap
Posts: 13639
Joined: Sat Nov 26, 2016 6:32 pm
Location: Hawaii No Ka Oi , N. Arizona

Re: Unexpected Inflation — How Often Does It Happen?

Post by Sandtrap »

OP:
As always: Great work and presentation.
Thanks!

I wonder how a classically constructed LMP (per Bernstein's: "Life Cycle Investing") would compare in your charts at various time frames?

Or, a 30/70 Allocation?

Or, the highest period of hyperinflationinflation: the 70's (1970's)?
(pre TIPS)

j :D

*edited to correct errors. :D
Last edited by Sandtrap on Sun Jun 13, 2021 7:04 am, edited 2 times in total.
Wiki Bogleheads Wiki: Everything You Need to Know
User avatar
nisiprius
Advisory Board
Posts: 44045
Joined: Thu Jul 26, 2007 9:33 am
Location: The terrestrial, globular, planetary hunk of matter, flattened at the poles, is my abode.--O. Henry

Re: Unexpected Inflation — How Often Does It Happen?

Post by nisiprius »

My answer is that there has been no significant unexpected inflation in the history of TIPS. Therefore, looking at short-term correlations between CPI and TIPS prices are in my opinion putting a microscope on background noise.

Image

My working belief is that there have been three periods of high inflation in the United States in the 108 years since the establishment of the consumer price index, an average of one every thirty years; early 1920s, 1940, 1970s.

In the absence of TIPS I don't know if there is a way to measure "expectedness" but I will venture a subjective opinion that all three were unexpected at the time, though of course people think they were predictable in hindsight.

The 1920s are particularly relevant because at that time the United States was on the gold standard, and anchoring the dollar to gold completely failed to provide price stability.
Annual income twenty pounds, annual expenditure nineteen nineteen and six, result happiness; Annual income twenty pounds, annual expenditure twenty pounds ought and six, result misery.
seajay
Posts: 225
Joined: Sat May 01, 2021 3:26 pm

Re: Unexpected Inflation — How Often Does It Happen?

Post by seajay »

NiceUnparticularMan wrote: Sun Jun 13, 2021 5:34 amAs always, this Vanguard white paper is very helpful on these concepts:

https://advisors.vanguard.com/iwe/pdf/ISGCTIPS.pdf

As an aside, I note short-term TIPS are going to be more responsive to unexpected inflation than long-term TIPS.
At just hedging the capital value invested in those ST TIPS. Not as a broader portfolio wide hedge.
That white paper also points out how free-floating gold has not in fact behaved in the way that this poster is predicting it should have behaved.

However, if you are interested in portfolio level hedges against unexpected inflation, it points out you could consider a diversified basket of productive commodities instead.
Gold is a commodity. Your disdain of gold has previously been well noted.
User avatar
nisiprius
Advisory Board
Posts: 44045
Joined: Thu Jul 26, 2007 9:33 am
Location: The terrestrial, globular, planetary hunk of matter, flattened at the poles, is my abode.--O. Henry

Re: Unexpected Inflation — How Often Does It Happen?

Post by nisiprius »

Sandtrap wrote: Sun Jun 13, 2021 6:24 am...the highest period of hyperinflation: the 70's (1970's)?
(pre TIPS)...
Apologies for being pedantic. First, the worst period of inflation in the US since the establishment of the consumer price index wasn't the 1970s, it was the early 1920s.

Second, people throw the word "hyperinflation" around far too casually. The 1970s were not a period of hyperinflation. They were simply a period of high inflation.

Wikipedia says the word hyperinflation was coined in 1956 by an economist named Philip Cagan, and that
Economists usually follow Cagan's description that hyperinflation occurs when the monthly inflation rate exceeds 50% (this is equivalent to a yearly rate of 12974.63%).
Actual hyperinflation is occurring now in Venezuela, recently in Zimbabwe, and in Germany in the 1920s and the Wikipedia article has a wideranging review.

There hasn't been any hyperinflation in the United States since the CPI was established. The same Wikipedia article says that hyperinflation did occur during the American Revolutionary War with the money authorized by the Continental Congress ("not worth a Continental,") during the Civil War in the Confederate States of America, and possibly in the Union states.
Annual income twenty pounds, annual expenditure nineteen nineteen and six, result happiness; Annual income twenty pounds, annual expenditure twenty pounds ought and six, result misery.
User avatar
Topic Author
SimpleGift
Posts: 3983
Joined: Tue Feb 08, 2011 3:45 pm

Re: Unexpected Inflation — How Often Does It Happen?

Post by SimpleGift »

nisiprius wrote: Sun Jun 13, 2021 6:50 am My answer is that there has been no significant unexpected inflation in the history of TIPS. Therefore, looking at short-term correlations between CPI and TIPS prices are in my opinion putting a microscope on background noise.
Actually, I tend to agree. Any meaningful analysis is definitely restricted by the lack of reliable data for expected inflation rates prior to the arrival of TIPS and inflation swap derivatives.

From a long-term historical perspective (which you've nicely provided, thanks), it is cosmically ironic that the history of TIPS in the U.S., from 1997 to present, has coincided with the longest period of moderate inflation, with a general disinflationary trend, in the country's history!

In short, the greatest potential benefit of having TIPS in a bond-heavy portfolio has never really been tested — and because of the destructive impact of runaway high inflation on most all financial assets, I honestly hope that it never is tested.
Last edited by SimpleGift on Sun Jun 13, 2021 8:36 am, edited 1 time in total.
Angst
Posts: 2711
Joined: Sat Jun 09, 2007 11:31 am

Re: Unexpected Inflation — How Often Does It Happen?

Post by Angst »

SimpleGift wrote: Sun Jun 13, 2021 4:08 am TIPS were first auctioned in the U.S. in January of 1997, which give us almost 25 years (293 months) over which to compare monthly, annualized expected inflation rates with actual realized inflation (first chart below). The one-year expected inflation rates (in blue) are from models developed by the Cleveland Federal Reserve.
  • Image
    Data: Monthly one-year expected inflation from Fed models; CPI-U actual inflation from FRED.
Thanks for this post Simplegift! But I want to be sure I understand what data are being presented. Your clevelandfed.org (models) link above shows a graph including the "Ten-Year Expected Inflation". Were you able to pull the one-year expected inflation data from within their website or is your blue line above actually ten-year expected data?

I'm interested in how accurate the forecasts have been over the years, particularly since TIPS market data became available in 1997, but I wouldn't have guessed that the one-year Fed forecasts were so far off from the actual inflation 1-year later as your graphs show. In addition, as much as I am definitely interested in the direct comparison between the Cleveland forecast and actual inflation, do you think it would help to add the 1-year BEI rate to your graph? The BEI rate seems to be the default "forecast" we're always going to here in the forum.
User avatar
Topic Author
SimpleGift
Posts: 3983
Joined: Tue Feb 08, 2011 3:45 pm

Re: Unexpected Inflation — How Often Does It Happen?

Post by SimpleGift »

Angst wrote: Sun Jun 13, 2021 7:31 am Were you able to pull the one-year expected inflation data from within their website or is your blue line above actually ten-year expected data?
Yes, it's actually the 1-year expected inflation rate. The Cleveland Fed models, in an Excel file on the website linked in the OP, have historical expected inflation rates for every maturity from 1 year to 30 years, going back to 1982. This is the direct link to their spreadsheet, if interested.
Angst wrote: Sun Jun 13, 2021 7:31 am ...do you think it would help to add the 1-year BEI rate to your graph? The BEI rate seems to be the default "forecast" we're always going to here in the forum.
Unless I'm missing something, there is no readily available source of 1-year breakeven inflation rates. There are published 5-year and 10-year breakeven rates, yes, but no 1-year as best I know. This one of the reasons the Cleveland Fed developed and maintains their expected inflation models, in order to fill these kind of data holes, I believe.
Angst
Posts: 2711
Joined: Sat Jun 09, 2007 11:31 am

Re: Unexpected Inflation — How Often Does It Happen?

Post by Angst »

SimpleGift wrote: Sun Jun 13, 2021 7:49 am
Angst wrote: Sun Jun 13, 2021 7:31 am Were you able to pull the one-year expected inflation data from within their website or is your blue line above actually ten-year expected data?
Yes, it's actually the 1-year expected inflation rate. The Cleveland Fed models, in an Excel file on the website linked in the OP, have historical expected inflation rates for every maturity from 1 year to 30 years, going back to 1982. This is the direct link to their spreadsheet, if interested.
Thanks for clarifying this.
SimpleGift wrote: Sun Jun 13, 2021 7:49 am
Angst wrote: Sun Jun 13, 2021 7:31 am ...do you think it would help to add the 1-year BEI rate to your graph? The BEI rate seems to be the default "forecast" we're always going to here in the forum.
Unless I'm missing something, there is no readily available source of 1-year breakeven inflation rates. There are published 5-year and 10-year breakeven rates, yes, but no 1-year as best I know. This one of the reasons the Cleveland Fed developed and maintains their expected inflation models, in order to fill these kind of data holes, I believe.
:oops:
Ah yes, of course. Surely the Cleveland Fed makes use of the real time/daily BEI data available when it makes its one-year forecasts. It's too bad the data don't appear to have been aggregated for long term use.
Angst
Posts: 2711
Joined: Sat Jun 09, 2007 11:31 am

Re: Unexpected Inflation — How Often Does It Happen?

Post by Angst »

nisiprius wrote: Sun Jun 13, 2021 7:01 am[Snip...]
...people throw the word "hyperinflation" around far too casually. The 1970s were not a period of hyperinflation. They were simply a period of high inflation.
Thanks for posting this reminder. My emphasis above.
NiceUnparticularMan
Posts: 2525
Joined: Sat Mar 11, 2017 7:51 am

Re: Unexpected Inflation — How Often Does It Happen?

Post by NiceUnparticularMan »

seajay wrote: Sun Jun 13, 2021 6:58 amGold is a commodity.
Again people can look at the white paper to see how free-floating gold behaved differently from a diversified basket of productive commodities. It is in fact true that free-floating gold is a commodity, and in fact that it has productive uses. But obviously on its own it is not diversified, and additionally because a lot of people buy gold speculatively and simply store it without using it productively, it does not act like it would if it was only a productive commodity.

But many diversified commodity indices and funds actually include precious metals generally, and then gold specifically, and I wouldn't suggest it is necessary to try to find a fund that excludes gold.
dcabler
Posts: 1932
Joined: Wed Feb 19, 2014 11:30 am

Re: Unexpected Inflation — How Often Does It Happen?

Post by dcabler »

Angst wrote: Sun Jun 13, 2021 8:14 am
SimpleGift wrote: Sun Jun 13, 2021 7:49 am
Angst wrote: Sun Jun 13, 2021 7:31 am Were you able to pull the one-year expected inflation data from within their website or is your blue line above actually ten-year expected data?
Yes, it's actually the 1-year expected inflation rate. The Cleveland Fed models, in an Excel file on the website linked in the OP, have historical expected inflation rates for every maturity from 1 year to 30 years, going back to 1982. This is the direct link to their spreadsheet, if interested.
Thanks for clarifying this.
SimpleGift wrote: Sun Jun 13, 2021 7:49 am
Angst wrote: Sun Jun 13, 2021 7:31 am ...do you think it would help to add the 1-year BEI rate to your graph? The BEI rate seems to be the default "forecast" we're always going to here in the forum.
Unless I'm missing something, there is no readily available source of 1-year breakeven inflation rates. There are published 5-year and 10-year breakeven rates, yes, but no 1-year as best I know. This one of the reasons the Cleveland Fed developed and maintains their expected inflation models, in order to fill these kind of data holes, I believe.
:oops:
Ah yes, of course. Surely the Cleveland Fed makes use of the real time/daily BEI data available when it makes its one-year forecasts. It's too bad the data don't appear to have been aggregated for long term use.
Not sure what you meant by the last sentence (underlined). Can you elaborate?

Cheers.
NiceUnparticularMan
Posts: 2525
Joined: Sat Mar 11, 2017 7:51 am

Re: Unexpected Inflation — How Often Does It Happen?

Post by NiceUnparticularMan »

nisiprius wrote: Sun Jun 13, 2021 6:50 am In the absence of TIPS I don't know if there is a way to measure "expectedness" but I will venture a subjective opinion that all three were unexpected at the time, though of course people think they were predictable in hindsight.
The Vanguard white paper contains one example of how people try to do that (see Appendix A), and discusses some other methods as well. In Figure 1 they have results back to 1960.

As you hypothesized, during the peaks of the 1970s inflation, CPI was generally running higher than their model of expectations. However, it is worth noting that in the mid-70s dip inflation was unexpectedly low, and again then in the very late 1970s and early 1980s, when actual inflation came way down, it was generally below expectations.

All of this is just pointing out what I think we should expect--expectations can fail to anticipate both the coming of unexpectedly high inflation, and then also its going.
User avatar
midareff
Posts: 7654
Joined: Mon Nov 29, 2010 10:43 am
Location: Biscayne Bay, South Florida

Re: Unexpected Inflation — How Often Does It Happen?

Post by midareff »

I would think the unexpected happens unpredictably.
User avatar
nisiprius
Advisory Board
Posts: 44045
Joined: Thu Jul 26, 2007 9:33 am
Location: The terrestrial, globular, planetary hunk of matter, flattened at the poles, is my abode.--O. Henry

Re: Unexpected Inflation — How Often Does It Happen?

Post by nisiprius »

NiceUnparticularMan wrote: Sun Jun 13, 2021 5:34 am...As always, this Vanguard white paper is very helpful on these concepts:

https://advisors.vanguard.com/iwe/pdf/ISGCTIPS.pdf
Thanks for continuing to nudge on this paper. Yes, it is useful, in itself and for pointing me toward Erten and Ocampo (2012), Super-cycles of commodity prices since the mid-nineteenth century
. I have been skeptical of "commodities" though, and both papers reinforce my skepticism (even though Vanguard's is arguing the opposite).

From the Erten and Ocampo paper, ignore the trend and supercycle data and just look at the total index (solid line).

Image

What we are seeing for a "basket of commodities" is very similar to what I saw when I looked at just one of them, copper, yesterday.

Possibly it is fair to say that "commodities" have held real value over the long term, although note that both the raw and the smoothed dashed trend line challenge this a little. The "index" seems to have fallen from about 5 to about 4.5, and since that is a log scale, it represents a factor of three loss in purchasing power over a period of 146 years = -0.8% real annualized. But not bad. Certainly far better than a non-interest-bearing bank account.

Within that period, though, we are seeing excursions of e.g 10^(4.9-4.1) = factors of 6X, in the case of 1915-1920 occurring within just a few years. And we are also seeing a fairly conspicuous failure to hold real value during the last important period of inflation, the 1970s.

That's about what I saw with copper prices, 1913-2017. No consistency. Copper has sometimes spiked when inflation has spiked. But sometimes it has spiked when there was no inflation. Sometimes (1920) it has failed to spike when inflation spiked. And sometimes (1930) it has dived when the dollar was increasing in value.

Another important observation to make is to compare a "basket of commodities" with Treasury bills. The Vanguard paper presents this chart:

Image

I find it hard to feel that given the volatility, that commodities offer any advantage over Treasury bills. The high volatility of commodities just makes them an easy sell. It is easy to get excited when they undergo a short-term spike, and the fluctuation also makes it easy to engage of overactive pattern recognition (or salesmanship) and say things like "Given that .. copper prices have a history of rising when inflation takes hold..."

Of course, the constant mixing up of "commodities" with "commodity futures" doesn't help.
Annual income twenty pounds, annual expenditure nineteen nineteen and six, result happiness; Annual income twenty pounds, annual expenditure twenty pounds ought and six, result misery.
User avatar
Topic Author
SimpleGift
Posts: 3983
Joined: Tue Feb 08, 2011 3:45 pm

Re: Unexpected Inflation — How Often Does It Happen?

Post by SimpleGift »

One hesitates to make a post suggesting "this time is different," but it's worth pointing that today nearly all of the world's major economies, emerging and developed, now have explicit inflation targets in place, managed by their central banks or governments (chart below). So in this sense, the world is very different today than it was during any of high, runaway inflation episodes of past historical periods.

A current list of all inflation-targeting countries and their target rates can be found here.
  • Image
Going forward, thinking about unexpected inflation in the future, I'm not sure the examples of the historical past back to the early 1900s are of much relevance. One can point to today's disinflationary trends of aging populations around the world and increased global competition in goods and services, but these are squishy factors, hard to quantify and subject to change. In contrast, formal inflation targeting by major countries worldwide is definitely here to stay.

Though a third of our bond portfolio is in TIPS, runaway unexpected inflation is not something we lose sleep over.
Angst
Posts: 2711
Joined: Sat Jun 09, 2007 11:31 am

Re: Unexpected Inflation — How Often Does It Happen?

Post by Angst »

dcabler wrote: Sun Jun 13, 2021 9:11 am
Angst wrote: Sun Jun 13, 2021 8:14 am
SimpleGift wrote: Sun Jun 13, 2021 7:49 am
Angst wrote: Sun Jun 13, 2021 7:31 am ...do you think it would help to add the 1-year BEI rate to your graph? The BEI rate seems to be the default "forecast" we're always going to here in the forum.
Unless I'm missing something, there is no readily available source of 1-year breakeven inflation rates. There are published 5-year and 10-year breakeven rates, yes, but no 1-year as best I know. This one of the reasons the Cleveland Fed developed and maintains their expected inflation models, in order to fill these kind of data holes, I believe.
:oops:
Ah yes, of course. Surely the Cleveland Fed makes use of the real time/daily BEI data available when it makes its one-year forecasts. It's too bad the data don't appear to have been aggregated for long term use.
Not sure what you meant by the last sentence (underlined). Can you elaborate?

Cheers.
Yes, I'm just trying to echo this bolded part of SimpleGift's response, but I could have worded it better. BEI rates can easily be calculated by subtracting the real Treasury rate (for a specific maturity) from the nominal Treasury rate (for the same maturity). However, the Treasury (NY Fed) only publishes daily 5, 7, 10, 20 and 30 year real rates:

Real (TIPS) rates. https://www.treasury.gov/resource-cente ... &year=2021
Nominal rates.... https://www.treasury.gov/resource-cente ... &year=2021

SimpleGift got the one-year expected inflation numbers from the Cleveland Fed, but I don't think they explicitly reflect 1-yr BEI rates historically.

As something of an aside, I find the graphic (link below) at the bottom of page 3 from the Vanguard white paper pdf that NiceUnparticularMan referenced to be helpful. Vanguard found ways to pull together together historical expected inflation rates although it's unclear how short-term they were.

https://advisors.vanguard.com/iwe/pdf/ISGCTIPS.pdf#page=3
NiceUnparticularMan
Posts: 2525
Joined: Sat Mar 11, 2017 7:51 am

Re: Unexpected Inflation — How Often Does It Happen?

Post by NiceUnparticularMan »

nisiprius wrote: Sun Jun 13, 2021 9:54 am
NiceUnparticularMan wrote: Sun Jun 13, 2021 5:34 am...As always, this Vanguard white paper is very helpful on these concepts:

https://advisors.vanguard.com/iwe/pdf/ISGCTIPS.pdf
Thanks for continuing to nudge on this paper. Yes, it is useful, in itself and for pointing me toward Erten and Ocampo (2012), Super-cycles of commodity prices since the mid-nineteenth century
. I have been skeptical of "commodities" though, and both papers reinforce my skepticism (even though Vanguard's is arguing the opposite).

From the Erten and Ocampo paper, ignore the trend and supercycle data and just look at the total index (solid line).

Image

What we are seeing for a "basket of commodities" is very similar to what I saw when I looked at just one of them, copper, yesterday.

Possibly it is fair to say that "commodities" have held real value over the long term, although note that both the raw and the smoothed dashed trend line challenge this a little. The "index" seems to have fallen from about 5 to about 4.5, and since that is a log scale, it represents a factor of three loss in purchasing power over a period of 146 years = -0.8% real annualized. But not bad. Certainly far better than a non-interest-bearing bank account.

Within that period, though, we are seeing excursions of e.g 10^(4.9-4.1) = factors of 6X, in the case of 1915-1920 occurring within just a few years. And we are also seeing a fairly conspicuous failure to hold real value during the last important period of inflation, the 1970s.

That's about what I saw with copper prices, 1913-2017. No consistency. Copper has sometimes spiked when inflation has spiked. But sometimes it has spiked when there was no inflation. Sometimes (1920) it has failed to spike when inflation spiked. And sometimes (1930) it has dived when the dollar was increasing in value.

Another important observation to make is to compare a "basket of commodities" with Treasury bills. The Vanguard paper presents this chart:

Image

I find it hard to feel that given the volatility, that commodities offer any advantage over Treasury bills. The high volatility of commodities just makes them an easy sell. It is easy to get excited when they undergo a short-term spike, and the fluctuation also makes it easy to engage of overactive pattern recognition (or salesmanship) and say things like "Given that .. copper prices have a history of rising when inflation takes hold..."

Of course, the constant mixing up of "commodities" with "commodity futures" doesn't help.
So that index does show a local spike in the 1970s.

The point of diversifying among productive commodities would be to smooth out some idiosyncratic factors.

But gnerally, it is absolutely the case diversified commodities should be expected to do much worse than something like short term TIPS or Tbills on a volatility-adjusted basis. That is essentially the insurance "premium" you are paying for the ability of diversified commodities to MAYBE have much higher Beta to unexpected inflation.

But you need that Beta if you want a smallish allocation to provide portfolio level insurance. But then it will cost you if you end up not needing it.

So it is a tradeoff, and recently investors in diversified commodities have paid the premiums but not needed the insurance. Which should and does look like a loss.
Last edited by NiceUnparticularMan on Sun Jun 13, 2021 11:14 am, edited 1 time in total.
User avatar
Oicuryy
Posts: 1613
Joined: Thu Feb 22, 2007 10:29 pm

Re: Unexpected Inflation — How Often Does It Happen?

Post by Oicuryy »

Measuringworth's U.S. consumer price index goes back to 1774.

Image

Under the gold standard, one could expect inflation during wars and deflation post war.

The high inflation of the 1970's is explained in this paper.
Arthur Burns and Inflation
Here is part of the conclusion.
The fundamental divide in monetary economics is whether the price level is a monetary or a nonmonetary phenomenon. If the price level is a monetary phenomenon, it varies to endow the nominal quantity of money with the real purchasing power desired by the public. The central bank is the cause of inflation.

If the price level is a nonmonetary or real phenomenon, its behavior possesses multiple, changing causes. Direct intervention by government in the price setting practices of the public can lower inflation. Such intervention permits a more expansionary monetary policy designed to lower unemployment and stimulate real growth.

Burns conducted monetary policy on the assumption that the price level is a nonmonetary phenomenon. The Congress and the administration, public opinion, and most of the economics profession supported that policy. The result was inflation. That inflation eventually led to the present consensus that the control of inflation is the paramount responsibility of the central bank.
These days one's expectation for inflation depends on one's expectation for the course of monetary policy. Discussions of monetary policy are banned on this site.

Ron
Last edited by Oicuryy on Sun Jun 13, 2021 11:14 am, edited 1 time in total.
Money is fungible | Abbreviations and Acronyms
NiceUnparticularMan
Posts: 2525
Joined: Sat Mar 11, 2017 7:51 am

Re: Unexpected Inflation — How Often Does It Happen?

Post by NiceUnparticularMan »

dcabler wrote: Sun Jun 13, 2021 9:11 am
Angst wrote: Sun Jun 13, 2021 8:14 am
SimpleGift wrote: Sun Jun 13, 2021 7:49 am
Angst wrote: Sun Jun 13, 2021 7:31 am Were you able to pull the one-year expected inflation data from within their website or is your blue line above actually ten-year expected data?
Yes, it's actually the 1-year expected inflation rate. The Cleveland Fed models, in an Excel file on the website linked in the OP, have historical expected inflation rates for every maturity from 1 year to 30 years, going back to 1982. This is the direct link to their spreadsheet, if interested.
Thanks for clarifying this.
SimpleGift wrote: Sun Jun 13, 2021 7:49 am
Angst wrote: Sun Jun 13, 2021 7:31 am ...do you think it would help to add the 1-year BEI rate to your graph? The BEI rate seems to be the default "forecast" we're always going to here in the forum.
Unless I'm missing something, there is no readily available source of 1-year breakeven inflation rates. There are published 5-year and 10-year breakeven rates, yes, but no 1-year as best I know. This one of the reasons the Cleveland Fed developed and maintains their expected inflation models, in order to fill these kind of data holes, I believe.
:oops:
Ah yes, of course. Surely the Cleveland Fed makes use of the real time/daily BEI data available when it makes its one-year forecasts. It's too bad the data don't appear to have been aggregated for long term use.
Not sure what you meant by the last sentence (underlined). Can you elaborate?

Cheers.
Speaking just for myself, as a I recall there were a couple isolated auctions of 30 year TBonds in 1963 that were bought at around 4%. Then over the next 30 years, inflation alone averaged around 5.3%.

Ouch.

However, due to the general lack of long bonds in this period, there is not much general information available on how much really long inflation expectations varied from actual inflation over those periods.
NiceUnparticularMan
Posts: 2525
Joined: Sat Mar 11, 2017 7:51 am

Re: Unexpected Inflation — How Often Does It Happen?

Post by NiceUnparticularMan »

Angst wrote: Sun Jun 13, 2021 10:36 am
dcabler wrote: Sun Jun 13, 2021 9:11 am
Angst wrote: Sun Jun 13, 2021 8:14 am
SimpleGift wrote: Sun Jun 13, 2021 7:49 am
Angst wrote: Sun Jun 13, 2021 7:31 am ...do you think it would help to add the 1-year BEI rate to your graph? The BEI rate seems to be the default "forecast" we're always going to here in the forum.
Unless I'm missing something, there is no readily available source of 1-year breakeven inflation rates. There are published 5-year and 10-year breakeven rates, yes, but no 1-year as best I know. This one of the reasons the Cleveland Fed developed and maintains their expected inflation models, in order to fill these kind of data holes, I believe.
:oops:
Ah yes, of course. Surely the Cleveland Fed makes use of the real time/daily BEI data available when it makes its one-year forecasts. It's too bad the data don't appear to have been aggregated for long term use.
Not sure what you meant by the last sentence (underlined). Can you elaborate?

Cheers.
Yes, I'm just trying to echo this bolded part of SimpleGift's response, but I could have worded it better. BEI rates can easily be calculated by subtracting the real Treasury rate (for a specific maturity) from the nominal Treasury rate (for the same maturity). However, the Treasury (NY Fed) only publishes daily 5, 7, 10, 20 and 30 year real rates:

Real (TIPS) rates. https://www.treasury.gov/resource-cente ... &year=2021
Nominal rates.... https://www.treasury.gov/resource-cente ... &year=2021

SimpleGift got the one-year expected inflation numbers from the Cleveland Fed, but I don't think they explicitly reflect 1-yr BEI rates historically.

As something of an aside, I find the graphic (link below) at the bottom of page 3 from the Vanguard white paper pdf that NiceUnparticularMan referenced to be helpful. Vanguard found ways to pull together together historical expected inflation rates although it's unclear how short-term they were.

https://advisors.vanguard.com/iwe/pdf/ISGCTIPS.pdf#page=3
If I am reading that paper right, I believe they are comparing actual 12 month inflation to 12 month inflation expectations.
User avatar
Topic Author
SimpleGift
Posts: 3983
Joined: Tue Feb 08, 2011 3:45 pm

Re: Unexpected Inflation — How Often Does It Happen?

Post by SimpleGift »

Oicuryy wrote: Sun Jun 13, 2021 11:11 am The high inflation of the 1970's is explained in this paper.
...snip...
These days one's expectation for inflation depends on one's expectation for the course of monetary policy. Discussions of monetary policy are banned on this site.
Discussions of current monetary policy are banned, but a review of historical monetary policy from the 1970s can be a great help to understanding the differences between the policy situation today compared with back then.

In the 1970s, Fed officials were not blind to the inflation that was occurring, but were struggling to fulfill their dual mandate that required both full employment and price stability. Once inflation started getting out of hand, they faced an unhappy dilemma: fighting high unemployment would almost certainly drive inflation higher still, while fighting inflation would just as certainly cause unemployment to spike.

In the end, they made a fateful choice, deciding to concentrate on full employment and to tolerate high levels of inflation, in the mistaken belief that this might help to stimulate economic growth. It wasn't until Paul Volker's appointment as Fed chair in 1979 that the Fed reversed course and actually began to take inflation seriously.

So the high, runaway inflation of the 1970s was as much due to a clear policy failure as it was to the oil supply shocks. Do believe though that central banks around the world have learned a valuable lesson from the 1970s, and they aren't ever likely to repeat this past mistake of failing to take inflation seriously. It's an entirely different policy world now.
User avatar
#Cruncher
Posts: 3245
Joined: Fri May 14, 2010 2:33 am
Location: New York City
Contact:

Re: Unexpected Inflation — How Often Does It Happen?

Post by #Cruncher »

SimpleGift wrote: Sun Jun 13, 2021 4:08 am... almost 25 years (293 months) over which to compare monthly, annualized expected inflation rates with actual realized inflation (first chart below). The one-year expected inflation rates (in blue) are from models developed by the Cleveland Federal Reserve.
I'm no statistics expert. But unless I've made a serious error, there appears to be little correlation between the Cleveland Fed's monthly expected one-year inflation and the actual year-over-year CPI change twelve months hence. For the 281 months January 1997 through May 2020 it was only 0.087 -- close to 0.0 which would indicate no correlation at all. [*]

In contrast the correlation between the Cleveland Fed's expected one-year inflation and the actual year-over-year CPI change for the same month was 0.548. In other words, the Cleveland Fed's estimate of one-year future inflation correlates much better with the actual inflation for the past year than it does with the actual inflation for the next year. [*]
SimpleGift in same post wrote:A count of the months in the chart above indicates that inflation was greater-than-expected 54% of the time, and was less-than-expected 46% of the time. So the naive 50/50 allocation between TIPS and nominals so often recommended for bond-heavy portfolios appears to have historical support, if one's goal is to balance risk and return (not liability matching).
I don't understand how the data provides "historical support" for your conclusion, SimpleGift. It's like saying that one should split one's stock portfolios 50-50 between stocks with ticker symbols starting with the letters "A" to "M" and "N" to "Z" because each group out-performs the other 50% of the time.

* I used the Excel CORREL function on the following data.
0.087 = CORREL(B22:B302, D34:D314) -- Year over year CPI change one year later
0.548 = CORREL(D22:D302, B22:B302) -- Year over year CPI change past year

Code: Select all

Row     Col A   Col B     Col C   Col D
             Expected  Seas Adj   12 Mo
               Inflat     CPI-U     Chg

Code: Select all

 10  Jan 1996           154.700
 11  Feb 1996           155.000
 12  Mar 1996           155.500
 13  Apr 1996           156.100
 14  May 1996           156.400
 15  Jun 1996           156.700
 16  Jul 1996           157.000
 17  Aug 1996           157.200
 18  Sep 1996           157.700
 19  Oct 1996           158.200
 20  Nov 1996           158.700
 21  Dec 1996           159.100
 
 22  Jan 1997   3.16%   159.400   3.04% 
 23  Feb 1997   3.13%   159.700   3.03%
 24  Mar 1997   3.23%   159.800   2.77% 
 25  Apr 1997   3.23%   159.900   2.43% 
 26  May 1997   3.13%   159.900   2.24% 
 27  Jun 1997   3.20%   160.200   2.23% 
 28  Jul 1997   3.09%   160.400   2.17% 
 29  Aug 1997   2.97%   160.800   2.29% 
 30  Sep 1997   3.01%   161.200   2.22% 
 31  Oct 1997   2.99%   161.500   2.09% 
 32  Nov 1997   2.93%   161.700   1.89% 
 33  Dec 1997   2.94%   161.800   1.70%
 
 34  Jan 1998   2.76%   162.000   1.63% 
 35  Feb 1998   2.62%   162.000   1.44% 
 36  Mar 1998   2.87%   162.000   1.38% 
 37  Apr 1998   2.73%   162.200   1.44% 
 38  May 1998   2.64%   162.600   1.69% 
 39  Jun 1998   2.47%   162.800   1.62% 
 40  Jul 1998   2.74%   163.200   1.75% 
 41  Aug 1998   2.71%   163.400   1.62% 
 42  Sep 1998   2.53%   163.500   1.43% 
 43  Oct 1998   2.36%   163.900   1.49% 
 44  Nov 1998   2.35%   164.100   1.48% 
 45  Dec 1998   2.27%   164.400   1.61% 
 46  Jan 1999   2.40%   164.700   1.67% 
 47  Feb 1999   2.42%   164.700   1.67% 
 48  Mar 1999   2.71%   164.800   1.73% 
 49  Apr 1999   2.70%   165.900   2.28% 
 50  May 1999   2.74%   166.000   2.09% 
 51  Jun 1999   2.35%   166.000   1.97% 
 52  Jul 1999   2.80%   166.700   2.14% 
 53  Aug 1999   2.81%   167.100   2.26% 
 54  Sep 1999   2.69%   167.800   2.63% 
 55  Oct 1999   2.88%   168.100   2.56% 
 56  Nov 1999   2.99%   168.400   2.62% 
 57  Dec 1999   3.16%   168.800   2.68% 
 58  Jan 2000   3.08%   169.300   2.79% 
 59  Feb 2000   3.16%   170.000   3.22% 
 60  Mar 2000   3.24%   171.000   3.76% 
 61  Apr 2000   3.16%   170.900   3.01% 
 62  May 2000   3.31%   171.200   3.13% 
 63  Jun 2000   3.76%   172.200   3.73% 
 64  Jul 2000   3.09%   172.700   3.60% 
 65  Aug 2000   3.12%   172.700   3.35% 
 66  Sep 2000   3.17%   173.600   3.46% 
 67  Oct 2000   3.11%   173.900   3.45% 
 68  Nov 2000   3.10%   174.200   3.44% 
 69  Dec 2000   3.17%   174.600   3.44% 
 70  Jan 2001   2.65%   175.600   3.72% 
 71  Feb 2001   2.63%   176.000   3.53% 
 72  Mar 2001   2.14%   176.100   2.98% 
 73  Apr 2001   2.56%   176.400   3.22% 
 74  May 2001   2.72%   177.300   3.56% 
 75  Jun 2001   2.75%   177.700   3.19% 
 76  Jul 2001   2.69%   177.400   2.72% 
 77  Aug 2001   2.59%   177.400   2.72% 
 78  Sep 2001   3.12%   178.100   2.59% 
 79  Oct 2001   2.29%   177.600   2.13% 
 80  Nov 2001   2.08%   177.500   1.89% 
 81  Dec 2001   2.73%   177.400   1.60% 
 82  Jan 2002   2.29%   177.700   1.20% 
 83  Feb 2002   2.15%   178.000   1.14% 
 84  Mar 2002   2.01%   178.500   1.36% 
 85  Apr 2002   2.59%   179.300   1.64% 
 86  May 2002   2.62%   179.500   1.24% 
 87  Jun 2002   2.34%   179.600   1.07% 
 88  Jul 2002   2.35%   180.000   1.47% 
 89  Aug 2002   2.13%   180.500   1.75% 
 90  Sep 2002   2.04%   180.800   1.52% 
 91  Oct 2002   2.01%   181.200   2.03% 
 92  Nov 2002   2.03%   181.500   2.25% 
 93  Dec 2002   1.95%   181.800   2.48% 
 94  Jan 2003   2.08%   182.600   2.76% 
 95  Feb 2003   2.11%   183.600   3.15% 
 96  Mar 2003   1.90%   183.900   3.03% 
 97  Apr 2003   1.99%   183.200   2.18% 
 98  May 2003   1.84%   182.900   1.89% 
 99  Jun 2003   2.28%   183.100   1.95% 
100  Jul 2003   1.63%   183.700   2.06% 
101  Aug 2003   1.85%   184.500   2.22% 
102  Sep 2003   1.98%   185.100   2.38% 
103  Oct 2003   1.79%   184.900   2.04% 
104  Nov 2003   1.93%   185.000   1.93% 
105  Dec 2003   2.22%   185.500   2.04% 
106  Jan 2004   1.91%   186.300   2.03% 
107  Feb 2004   1.99%   186.700   1.69% 
108  Mar 2004   1.70%   187.100   1.74% 
109  Apr 2004   1.85%   187.400   2.29% 
110  May 2004   2.21%   188.200   2.90% 
111  Jun 2004   2.57%   188.900   3.17% 
112  Jul 2004   2.26%   189.100   2.94% 
113  Aug 2004   2.34%   189.200   2.55% 
114  Sep 2004   2.68%   189.800   2.54% 
115  Oct 2004   2.15%   190.800   3.19% 
116  Nov 2004   2.21%   191.700   3.62% 
117  Dec 2004   1.75%   191.700   3.34% 
118  Jan 2005   2.14%   191.600   2.84% 
119  Feb 2005   2.14%   192.400   3.05% 
120  Mar 2005   2.33%   193.100   3.21% 
121  Apr 2005   2.46%   193.700   3.36% 
122  May 2005   2.58%   193.600   2.87% 
123  Jun 2005   2.36%   193.700   2.54% 
124  Jul 2005   2.33%   194.900   3.07% 
125  Aug 2005   2.46%   196.100   3.65% 
126  Sep 2005   2.35%   198.800   4.74% 
127  Oct 2005   2.85%   199.100   4.35% 
128  Nov 2005   3.05%   198.100   3.34% 
129  Dec 2005   3.23%   198.100   3.34% 
130  Jan 2006   2.40%   199.300   4.02% 
131  Feb 2006   2.51%   199.400   3.64% 
132  Mar 2006   1.77%   199.700   3.42% 
133  Apr 2006   2.68%   200.700   3.61% 
134  May 2006   2.88%   201.300   3.98% 
135  Jun 2006   2.74%   201.800   4.18% 
136  Jul 2006   2.80%   202.900   4.10% 
137  Aug 2006   2.87%   203.800   3.93% 
138  Sep 2006   2.65%   202.800   2.01% 
139  Oct 2006   2.40%   201.900   1.41% 
140  Nov 2006   1.89%   202.000   1.97% 
141  Dec 2006   1.99%   203.100   2.52% 
142  Jan 2007   2.68%   203.437   2.08% 
143  Feb 2007   2.64%   204.226   2.42% 
144  Mar 2007   2.80%   205.288   2.80% 
145  Apr 2007   2.72%   205.904   2.59% 
146  May 2007   2.92%   206.755   2.71% 
147  Jun 2007   3.21%   207.234   2.69% 
148  Jul 2007   2.61%   207.603   2.32% 
149  Aug 2007   2.65%   207.667   1.90% 
150  Sep 2007   2.57%   208.547   2.83% 
151  Oct 2007   2.29%   209.190   3.61% 
152  Nov 2007   2.63%   210.834   4.37% 
153  Dec 2007   2.54%   211.445   4.11% 
154  Jan 2008   2.47%   212.174   4.29% 
155  Feb 2008   2.30%   212.687   4.14% 
156  Mar 2008   1.93%   213.448   3.97% 
157  Apr 2008   2.02%   213.942   3.90% 
158  May 2008   2.29%   215.208   4.09% 
159  Jun 2008   2.73%   217.463   4.94% 
160  Jul 2008   2.95%   219.016   5.50% 
161  Aug 2008   3.05%   218.690   5.31% 
162  Sep 2008   2.73%   218.877   4.95% 
163  Oct 2008   1.61%   216.995   3.73% 
164  Nov 2008   0.78%   213.153   1.10% 
165  Dec 2008   0.28%   211.398  (0.02%)
166  Jan 2009   0.47%   211.933  (0.11%)
167  Feb 2009   0.38%   212.705   0.01% 
168  Mar 2009  (0.48%)  212.495  (0.45%)
169  Apr 2009   1.23%   212.709  (0.58%)
170  May 2009   1.30%   213.022  (1.02%)
171  Jun 2009   1.27%   214.790  (1.23%)
172  Jul 2009   1.55%   214.726  (1.96%)
173  Aug 2009   2.02%   215.445  (1.48%)
174  Sep 2009   2.36%   215.861  (1.38%)
175  Oct 2009   1.63%   216.509  (0.22%)
176  Nov 2009   1.81%   217.234   1.91% 
177  Dec 2009   1.79%   217.347   2.81% 
178  Jan 2010   1.70%   217.488   2.62% 
179  Feb 2010   1.73%   217.281   2.15% 
180  Mar 2010   1.97%   217.353   2.29% 
181  Apr 2010   1.59%   217.403   2.21% 
182  May 2010   1.56%   217.290   2.00% 
183  Jun 2010   1.35%   217.199   1.12% 
184  Jul 2010   0.96%   217.605   1.34% 
185  Aug 2010   1.34%   217.923   1.15% 
186  Sep 2010   0.91%   218.275   1.12% 
187  Oct 2010   1.37%   219.035   1.17% 
188  Nov 2010   1.40%   219.590   1.08% 
189  Dec 2010   1.49%   220.472   1.44% 
190  Jan 2011   1.79%   221.187   1.70% 
191  Feb 2011   1.95%   221.898   2.12% 
192  Mar 2011   2.11%   223.046   2.62% 
193  Apr 2011   2.08%   224.093   3.08% 
194  May 2011   2.29%   224.806   3.46% 
195  Jun 2011   2.12%   224.806   3.50% 
196  Jul 2011   2.43%   225.395   3.58% 
197  Aug 2011   1.51%   226.106   3.75% 
198  Sep 2011   1.01%   226.597   3.81% 
199  Oct 2011   1.47%   226.750   3.52% 
200  Nov 2011   1.51%   227.169   3.45% 
201  Dec 2011   1.71%   227.223   3.06% 
202  Jan 2012   1.47%   227.842   3.01% 
203  Feb 2012   1.50%   228.329   2.90% 
204  Mar 2012   1.36%   228.807   2.58% 
205  Apr 2012   1.69%   229.187   2.27% 
206  May 2012   1.59%   228.713   1.74% 
207  Jun 2012   0.79%   228.524   1.65% 
208  Jul 2012   1.30%   228.590   1.42% 
209  Aug 2012   1.37%   229.918   1.69% 
210  Sep 2012   1.93%   231.015   1.95% 
211  Oct 2012   1.61%   231.638   2.16% 
212  Nov 2012   1.66%   231.249   1.80% 
213  Dec 2012   2.10%   231.221   1.76% 
214  Jan 2013   1.32%   231.679   1.68% 
215  Feb 2013   1.36%   232.937   2.02% 
216  Mar 2013   1.00%   232.282   1.52% 
217  Apr 2013   1.46%   231.797   1.14% 
218  May 2013   1.18%   231.893   1.39% 
219  Jun 2013   1.35%   232.445   1.72% 
220  Jul 2013   1.00%   232.900   1.89% 
221  Aug 2013   1.80%   233.456   1.54% 
222  Sep 2013   2.20%   233.544   1.09% 
223  Oct 2013   1.56%   233.669   0.88% 
224  Nov 2013   1.45%   234.100   1.23% 
225  Dec 2013   1.92%   234.719   1.51% 
226  Jan 2014   1.64%   235.288   1.56% 
227  Feb 2014   1.59%   235.547   1.12% 
228  Mar 2014   1.22%   236.028   1.61% 
229  Apr 2014   1.65%   236.468   2.02% 
230  May 2014   1.77%   236.918   2.17% 
231  Jun 2014   1.53%   237.231   2.06% 
232  Jul 2014   1.86%   237.498   1.97% 
233  Aug 2014   1.81%   237.460   1.72% 
234  Sep 2014   2.12%   237.477   1.68% 
235  Oct 2014   1.60%   237.430   1.61% 
236  Nov 2014   1.32%   236.983   1.23% 
237  Dec 2014   2.19%   236.252   0.65% 
238  Jan 2015   0.92%   234.747  (0.23%)
239  Feb 2015   0.38%   235.342  (0.09%)
240  Mar 2015   0.80%   235.976  (0.02%)
241  Apr 2015   1.59%   236.222  (0.10%)
242  May 2015   1.70%   237.001   0.04% 
243  Jun 2015   1.78%   237.657   0.18% 
244  Jul 2015   1.80%   238.034   0.23% 
245  Aug 2015   1.72%   238.033   0.24% 
246  Sep 2015   1.93%   237.498   0.01% 
247  Oct 2015   1.31%   237.733   0.13% 
248  Nov 2015   1.31%   238.017   0.44% 
249  Dec 2015   1.47%   237.761   0.64% 
250  Jan 2016   1.47%   237.652   1.24% 
251  Feb 2016   1.12%   237.336   0.85% 
252  Mar 2016   0.78%   238.080   0.89% 
253  Apr 2016   1.69%   238.992   1.17% 
254  May 2016   1.66%   239.557   1.08% 
255  Jun 2016   1.28%   240.222   1.08% 
256  Jul 2016   1.65%   240.101   0.87% 
257  Aug 2016   1.59%   240.545   1.06% 
258  Sep 2016   2.06%   241.176   1.55% 
259  Oct 2016   1.74%   241.741   1.69% 
260  Nov 2016   1.90%   242.026   1.68% 
261  Dec 2016   2.29%   242.637   2.05% 
262  Jan 2017   2.10%   243.620   2.51% 
263  Feb 2017   1.94%   243.872   2.75% 
264  Mar 2017   1.50%   243.766   2.39% 
265  Apr 2017   1.72%   244.274   2.21% 
266  May 2017   1.56%   244.069   1.88% 
267  Jun 2017   0.96%   244.218   1.66% 
268  Jul 2017   1.71%   244.280   1.74% 
269  Aug 2017   1.65%   245.205   1.94% 
270  Sep 2017   2.06%   246.551   2.23% 
271  Oct 2017   1.95%   246.657   2.03% 
272  Nov 2017   2.07%   247.378   2.21% 
273  Dec 2017   2.82%   247.736   2.10% 
274  Jan 2018   1.90%   248.721   2.09% 
275  Feb 2018   2.14%   249.300   2.23% 
276  Mar 2018   1.69%   249.517   2.36% 
277  Apr 2018   1.85%   250.275   2.46% 
278  May 2018   1.97%   250.786   2.75% 
279  Jun 2018   1.59%   251.152   2.84% 
280  Jul 2018   2.12%   251.345   2.89% 
281  Aug 2018   2.13%   251.735   2.66% 
282  Sep 2018   2.41%   252.183   2.28% 
283  Oct 2018   2.21%   252.899   2.53% 
284  Nov 2018   2.17%   252.822   2.20% 
285  Dec 2018   1.98%   252.493   1.92% 
286  Jan 2019   1.80%   252.441   1.50% 
287  Feb 2019   1.79%   252.969   1.47% 
288  Mar 2019   1.38%   254.147   1.86% 
289  Apr 2019   2.04%   255.326   2.02% 
290  May 2019   2.19%   255.371   1.83% 
291  Jun 2019   1.61%   255.423   1.70% 
292  Jul 2019   1.75%   255.925   1.82% 
293  Aug 2019   1.71%   256.118   1.74% 
294  Sep 2019   1.64%   256.532   1.72% 
295  Oct 2019   1.72%   257.387   1.77% 
296  Nov 2019   1.70%   257.989   2.04% 
297  Dec 2019   1.70%   258.203   2.26% 
298  Jan 2020   1.80%   258.687   2.47% 
299  Feb 2020   1.65%   258.824   2.31% 
300  Mar 2020   0.99%   257.989   1.51% 
301  Apr 2020  (0.05%)  256.192   0.34% 
302  May 2020  (0.22%)  255.942   0.22%

303  Jun 2020   0.51%   257.282   0.73% 
304  Jul 2020   1.57%   258.604   1.05% 
305  Aug 2020   1.68%   259.511   1.32% 
306  Sep 2020   1.58%   260.149   1.41% 
307  Oct 2020   1.40%   260.462   1.19% 
308  Nov 2020   1.42%   260.927   1.14% 
309  Dec 2020   1.65%   261.560   1.30% 
310  Jan 2021   1.45%   262.231   1.37% 
311  Feb 2021   1.68%   263.161   1.68% 
312  Mar 2021   1.51%   264.793   2.64% 
313  Apr 2021   1.75%   266.832   4.15% 
314  May 2021   1.87%   268.551   4.93%
The expected inflation is from column B, "1 year Expected Inflation", from the "Expected Inflation" sheet of the Excel file I got by clicking Download our spreadsheet on the Cleveland Fed Expected Inflation web page referenced in the original post. I got the CPI-U by selecting All items in U.S. city average, all urban consumers, seasonally adjusted from this BLS web page.
Beardog
Posts: 102
Joined: Thu Mar 22, 2007 10:38 pm
Location: Arkansas

Re: Unexpected Inflation — How Often Does It Happen?

Post by Beardog »

midareff wrote: Sun Jun 13, 2021 9:24 am I would think the unexpected happens unpredictably.
Yep! Unexpectedly, the unexpected sometimes happens.
Beardog
User avatar
nisiprius
Advisory Board
Posts: 44045
Joined: Thu Jul 26, 2007 9:33 am
Location: The terrestrial, globular, planetary hunk of matter, flattened at the poles, is my abode.--O. Henry

Re: Unexpected Inflation — How Often Does It Happen?

Post by nisiprius »

Oicuryy wrote: Sun Jun 13, 2021 11:11 am Measuringworth's U.S. consumer price index goes back to 1774.
Thank you. I know of MeasuringWorth because I use it for gold prices but I keep forgetting all the other information it has.
...
Image
...
According to that chart, then, there has never yet been hyperinflation in the United States. Not even close.
Last edited by nisiprius on Sun Jun 13, 2021 2:24 pm, edited 1 time in total.
Annual income twenty pounds, annual expenditure nineteen nineteen and six, result happiness; Annual income twenty pounds, annual expenditure twenty pounds ought and six, result misery.
User avatar
nisiprius
Advisory Board
Posts: 44045
Joined: Thu Jul 26, 2007 9:33 am
Location: The terrestrial, globular, planetary hunk of matter, flattened at the poles, is my abode.--O. Henry

Re: Unexpected Inflation — How Often Does It Happen?

Post by nisiprius »

NiceUnparticularMan wrote: Sun Jun 13, 2021 11:07 am...so that index does show a local spike in the 1970s...
Well, IMHO the Vanguard chart does, and the Erten and Ocampo paper doesn't.

The Erten and Ocampo paper shows steady real loss from perhaps 4.5 to 4.2 on their smoothed trendline, 1970 to 1985 and since they are plotting the log that represents about a -50% real loss. Eyeballing the unsmoothed curve it seems like considerably more.

Any thoughts?
Annual income twenty pounds, annual expenditure nineteen nineteen and six, result happiness; Annual income twenty pounds, annual expenditure twenty pounds ought and six, result misery.
User avatar
telemark
Posts: 2918
Joined: Sat Aug 11, 2012 6:35 am

Re: Unexpected Inflation — How Often Does It Happen?

Post by telemark »

To me this suggests mainly that the Cleveland Fed uses conservative models that consistently underestimate swings in the inflation rate in both directions. But to what extent do market prices reflect those models? If we go by the financial media, someone is always predicting inflation, usually hyperinflation and just around the corner.
NiceUnparticularMan
Posts: 2525
Joined: Sat Mar 11, 2017 7:51 am

Re: Unexpected Inflation — How Often Does It Happen?

Post by NiceUnparticularMan »

nisiprius wrote: Sun Jun 13, 2021 2:23 pm
NiceUnparticularMan wrote: Sun Jun 13, 2021 11:07 am...so that index does show a local spike in the 1970s...
Well, IMHO the Vanguard chart does, and the Erten and Ocampo paper doesn't.

The Erten and Ocampo paper shows steady real loss from perhaps 4.5 to 4.2 on their smoothed trendline, 1970 to 1985 and since they are plotting the log that represents about a -50% real loss. Eyeballing the unsmoothed curve it seems like considerably more.

Any thoughts?
Well, that's what smoothing will do. If you look at Vanguard's chart of unexpected inflation (Figure 1 in Appendix A on Page 9), you have actual inflation largely being around or below expected inflation until two 1970s periods (with a dip in the middle), and then in 1979 into the early 1980s it is below expectations again.

Looking at your ex-oil commodities chart, there is in fact a local spike in real commodities prices during the relevant period from the early to late 1970s, with some ups and downs. By the early 1980s, though, that local spike dies off and it is back to its smoothed long-term trend line.

Now it looks to me like this is a different index from the one Vanguard is using for that period, BCOM, which among other things doesn't exclude oil. Based on glancing through the paper, it looks like that probably makes a bit of a difference. But even the non-oil commodities they focused on do seem to have departed meaningfully from their long-term trend for a while. It just didn't last beyond when inflation started falling below expectations again, which you would expect.
000
Posts: 4967
Joined: Thu Jul 23, 2020 12:04 am

Re: Unexpected Inflation — How Often Does It Happen?

Post by 000 »

SimpleGift wrote: Sun Jun 13, 2021 4:08 am Thoughts?
I think any data from the last forty years or so is irrelevant if we are entering a new era.

A pedant might argue that any past data is by definition irrelevant to unexpected occurrences.
User avatar
Topic Author
SimpleGift
Posts: 3983
Joined: Tue Feb 08, 2011 3:45 pm

Re: Unexpected Inflation — How Often Does It Happen?

Post by SimpleGift »

000 wrote: Sun Jun 13, 2021 4:33 pm I think any data from the last forty years or so is irrelevant if we are entering a new era.
I tend to agree, based on the widespread influence of formal inflation-targeting today by central banks around the world (which may or may not be what you had in mind by "new era").

The table below shows the current average inflation targets of developed and emerging market countries, in ascending order of their target levels. Also note there are now about 30 frontier market countries worldwide that have explicit inflation targets (in the 5%-8% range), with more countries adopting them every year.
  • Image
    Date source: CBN
Certainly, central banks will never be able to prevent temporary, event-driven inflation spikes or surges, but as far as the widespread increase in general inflation rates above target levels, it's hard to see how this might get established on the planet in this day and age — provided central banks are able to maintain their independence.
Last edited by SimpleGift on Sun Jun 13, 2021 6:14 pm, edited 1 time in total.
Tib
Posts: 104
Joined: Tue Oct 27, 2020 9:57 pm

Re: Unexpected Inflation — How Often Does It Happen?

Post by Tib »

As a retiree, not a worker with an inflatable salary and a remote investment horizon, I’m among those for whom inflation is indeed “top of mind” (as SimpleGift puts it). That’s because of “the destructive impact of runaway high inflation on most all financial assets” (SimpleGift again). Like many comfortable retirees I’m positioned to weather recessions of typical length and severity. By contrast, prolonged inflation like that from the late 1960s to the early 1980s could sink me. But where to look for insurance? Judging from the figure on page 7 of the Vanguard white paper, commodities outperformed stocks during the 1970s and the inflation of the early years following WW1, though evidently not during the inflation of the 1940s. Long-term, however, commodities have provided little or no real return, and market timing doesn’t seem sensible. Perhaps a fund holding the stocks of commodity producers would provide both a decent (if somewhat less reliable) inflation hedge and also a decent long-term return (even if one less than that of stocks generally). Unfortunately, I have no data on the returns of commodity producers during past periods of high inflation. MXI is an etf of global materials producers. XLB is comparable, but with a U.S. focus. DJP is an ETN designed to match the performance of a diversified basket of commodities. MXI, XLB, and DJP have all done very well during the current inflation scare, but Morningstar shows the 10-year returns of MXI and XLB to be 5.46% and 10.95%, as compared with -5.83% for DJP.
hudson
Posts: 4157
Joined: Fri Apr 06, 2007 9:15 am

Re: Unexpected Inflation — How Often Does It Happen?

Post by hudson »

Tib wrote: Sun Jun 13, 2021 5:42 pm But where to look for insurance?
I vote TIPS...duration matched.
Thesaints
Posts: 4163
Joined: Tue Jun 20, 2017 12:25 am

Re: Unexpected Inflation — How Often Does It Happen?

Post by Thesaints »

Shouldn't there be an offset between TIPS implied rate and inflation ?
I mean, when I bought a 5-year TIPS in January 1997 break-even inflation was X%. I have to wait until January 2002 to know if I actually broke even, or not.
I can maybe extract an annualized expected inflation rate from TIPS, by looking at the market price of running issues with a 1-year residual maturity and in that case there would be a 1-year lag.
What TIPS maturities were first issued in 1997 and following few years ?
NiceUnparticularMan
Posts: 2525
Joined: Sat Mar 11, 2017 7:51 am

Re: Unexpected Inflation — How Often Does It Happen?

Post by NiceUnparticularMan »

Tib wrote: Sun Jun 13, 2021 5:42 pm As a retiree, not a worker with an inflatable salary and a remote investment horizon, I’m among those for whom inflation is indeed “top of mind” (as SimpleGift puts it). That’s because of “the destructive impact of runaway high inflation on most all financial assets” (SimpleGift again). Like many comfortable retirees I’m positioned to weather recessions of typical length and severity. By contrast, prolonged inflation like that from the late 1960s to the early 1980s could sink me. But where to look for insurance? Judging from the figure on page 7 of the Vanguard white paper, commodities outperformed stocks during the 1970s and the inflation of the early years following WW1, though evidently not during the inflation of the 1940s. Long-term, however, commodities have provided little or no real return, and market timing doesn’t seem sensible.
So if you can meet your goals with negative real returns, as long as they are predictably SMALL negative real returns, you can just use a lot of TIPS.

Stable value funds (and the TSP G Fund) are another possible option, if you have such a thing available to you. There is some possibility if we moved into a higher long-term inflation environment with a steeper yield curve, these sorts of funds would make the transition without losing value and start returning better in real terms. Meaning you wouldn't be locking in real losses. On the other hand, absent those change in circumstances such instruments could also provide negative real returns going forward.

If that sort of solution doesn't work for you, at least not entirely, meaning you need to seek positive real returns as well in the long run . . . value-tilting your stocks is another idea. As always, I think you should assume there must be some risk to that, and I specifically think you should assume that in a combined economic crisis and high unexpected inflation scenario (stagflation or worse), this might not work very well. But, generally there is reason to believe value stocks will perform better in times of high inflation.
NiceUnparticularMan
Posts: 2525
Joined: Sat Mar 11, 2017 7:51 am

Re: Unexpected Inflation — How Often Does It Happen?

Post by NiceUnparticularMan »

Thesaints wrote: Sun Jun 13, 2021 5:51 pm Shouldn't there be an offset between TIPS implied rate and inflation ?
I mean, when I bought a 5-year TIPS in January 1997 break-even inflation was X%. I have to wait until January 2002 to know if I actually broke even, or not.
I can maybe extract an annualized expected inflation rate from TIPS, by looking at the market price of running issues with a 1-year residual maturity and in that case there would be a 1-year lag.
What TIPS maturities were first issued in 1997 and following few years ?
I'm not sure I understand the question. With your TIPS, the principal was adjusted for inflation as measured by the CPI. And so if you held it to maturity, you got the original expected return in real terms, as measured by the CPI, whatever that was.

These days, all the way out to 30 years, TIPS are promising real losses. But you won't get worse than that, as measured by CPI at least, if you hold them to maturity. You won't get better, either.
User avatar
#Cruncher
Posts: 3245
Joined: Fri May 14, 2010 2:33 am
Location: New York City
Contact:

Re: Unexpected Inflation — How Often Does It Happen?

Post by #Cruncher »

Thesaints wrote: Sun Jun 13, 2021 5:51 pm... when I bought a 5-year TIPS in January 1997 break-even inflation was X%. I have to wait until January 2002 to know if I actually broke even, or not.
The first 5-year TIPS was issued in July, not January, 1997. Its yield to maturity at the July 9th auction was 3.744%. It matured on July 15, 2002 with a nominal return of 6.103% as shown here. I estimate that one could have bought a regular 5-year Treasury note on the date of the auction with a 6.15% yield, or 0.05% points more than what the TIPS ended up returning on an inflation-adjusted basis. Another way to look at it:
  • When auctioned on 7/9/1997 the breakeven inflation rate was 2.41% (6.15% - 3.74%).
  • The actual inflation adjustment was only 2.36% (6.10% - 3.74%).
  • The nominal treasury therefore performed 0.05% points better (2.41% - 2.36%).
See the first row of the 5-year table in my post, Re: Treasury Notes Return vs Actual TIPS Return.
Thesaints in same post wrote:What TIPS maturities were first issued in 1997 and following few years?
See bottom table on this web page.
  • 1997: 5-year & 10-year
  • 1998: 10-year & 30-year
  • 1999: 10-year & 30-year
  • 2000: 10-year
  • ...
protagonist
Posts: 7171
Joined: Sun Dec 26, 2010 12:47 pm

Re: Unexpected Inflation — How Often Does It Happen?

Post by protagonist »

Nobody knows how often unexpected inflation will happen in the future in the USA.
It is statistically impossible to calculate a probability. It would be like calculating a probability as to whether there is extraterrestrial life (yes, I know there is the Drake equation, but that is equally ridiculous as trying to calculate the probability of unexpected inflation).
The economy is a highly complex, emergent system that is exquisitely sensitive to initial conditions. There are too many unknown variables to contend with.
All one can really say is that the probability will increase as the time frame increases, sort of like the probability of a nuclear attack.

If you can't calculate a number along with a margin of error using reliable techniques, you are just guessing.
Last edited by protagonist on Tue Jun 15, 2021 8:46 am, edited 2 times in total.
000
Posts: 4967
Joined: Thu Jul 23, 2020 12:04 am

Re: Unexpected Inflation — How Often Does It Happen?

Post by 000 »

SimpleGift wrote: Sun Jun 13, 2021 5:23 pm
000 wrote: Sun Jun 13, 2021 4:33 pm I think any data from the last forty years or so is irrelevant if we are entering a new era.
I tend to agree, based on the widespread influence of formal inflation-targeting today by central banks around the world (which may or may not be what you had in mind by "new era").

The table below shows the current average inflation targets of developed and emerging market countries, in ascending order of their target levels. Also note there are now about 30 frontier market countries worldwide that have explicit inflation targets (in the 5%-8% range), with more countries adopting them every year.
  • Image
    Date source: CBN
Certainly, central banks will never be able to prevent temporary, event-driven inflation spikes or surges, but as far as the widespread increase in general inflation rates above target levels, it's hard to see how this might get established on the planet in this day and age — provided central banks are able to maintain their independence.
I think you put too much faith in central banks. They are between a rock and a hard place. How can they fight inflation without raising rates? And how can they raise rates without bankrupting the world's debtor governments?
User avatar
Topic Author
SimpleGift
Posts: 3983
Joined: Tue Feb 08, 2011 3:45 pm

Re: Unexpected Inflation — How Often Does It Happen?

Post by SimpleGift »

000 wrote: Mon Jun 14, 2021 3:52 pm I think you put too much faith in central banks. They are between a rock and a hard place. How can they fight inflation without raising rates? And how can they raise rates without bankrupting the world's debtor governments?
You could be right — which is why the caveat at the end of the post, "provided central banks are able to maintain their independence." Studies have shown that the degree of a central bank's independence from political influence is negatively correlated with a country's inflation rate. In short, central bank independence leads to lower inflation.

But, as you suggest, in many countries there is nothing to stop governments from overruling their central banks to pursue policies other than price stability — which gets into the political realm of banned Forum discussion topics.
Tib
Posts: 104
Joined: Tue Oct 27, 2020 9:57 pm

Re: Unexpected Inflation — How Often Does It Happen?

Post by Tib »

Presumably, the Fed too, not just politicians, wants to avoid the damage that high interest rates would do to our heavily indebted country. Besides, controlling inflation is not the Fed's only mandate. Achieving and maintaining high levels of inclusive employment is likely to make it more difficult to keep inflation under good control.
User avatar
midareff
Posts: 7654
Joined: Mon Nov 29, 2010 10:43 am
Location: Biscayne Bay, South Florida

Re: Unexpected Inflation — How Often Does It Happen?

Post by midareff »

nisiprius wrote: Sun Jun 13, 2021 2:14 pm
Oicuryy wrote: Sun Jun 13, 2021 11:11 am Measuringworth's U.S. consumer price index goes back to 1774.
Thank you. I know of MeasuringWorth because I use it for gold prices but I keep forgetting all the other information it has.
...
Image
...
According to that chart, then, there has never yet been hyperinflation in the United States. Not even close.
Interesting plot Nisi.... I remember back about 1978 - 1979 my father, who had always been in the market since I can remember as a 5 or 6 year old listening to the stock market report of the tube radio, sold out of the stock market and loaded up on Savings & Loan 17% insured CDs. He was a good man, a good father and despite being mostly self educated, a very savvy investor.
namajones
Posts: 319
Joined: Sat Aug 08, 2020 12:41 pm

Re: Unexpected Inflation — How Often Does It Happen?

Post by namajones »

Producer prices climb 6.6% in May on annual basis, largest 12-month increase on record.

https://www.cnbc.com/2021/06/15/retail- ... -2021.html
staustin
Posts: 329
Joined: Mon Mar 14, 2011 9:36 am

Re: Unexpected Inflation — How Often Does It Happen?

Post by staustin »

Tib wrote: Mon Jun 14, 2021 5:14 pm Presumably, the Fed too, not just politicians, wants to avoid the damage that high interest rates would do to our heavily indebted country. Besides, controlling inflation is not the Fed's only mandate. Achieving and maintaining high levels of inclusive employment is likely to make it more difficult to keep inflation under good control.
It's the opposite actually.. policy seems to be moving towards inflating the debt away, at the expense of savers. As an earlier poster noted, presents a difficult situation for retirees. We've shortened bond duration and building cash short term.
namajones
Posts: 319
Joined: Sat Aug 08, 2020 12:41 pm

Re: Unexpected Inflation — How Often Does It Happen?

Post by namajones »

Professor Damodaran's blog is always interesting. He recently wrote a post on inflation:

http://aswathdamodaran.blogspot.com/202 ... rm-or.html

with these points toward the end:

"With a non-trivial chance of a breakout: If it is permanent, and we see inflation rise to levels not seen since the 1970s and 1980s (>5%), stocks and bonds will have to be repriced significantly. Not only will investors need to move money out of financial into real assets and collectibles, but companies and individuals that have chosen to borrow to capacity, based upon current low rates, will face a default risk reckoning.

And the Fed has to be ready: It behooves the Fed to get ahead of the inflation game. Since the probability of inflation rising to dangerous levels is non-trivial, in my view, the Fed should stop its happy talk about inflation being under control and interest rates staying low, no matter what. In fact, central bankers around the world would be well served reverting back to an old rule book of being seen very little and speaking even less, and letting their actions speak for themselves."
Post Reply