Current pricing for caps and floors shows that the market sees a 30% probability that inflation will be above 3% for the next five years, and a 5% probability that inflation will be below 1%, see chart below. A similar worry about high inflation can be seen in 5-year breakevens, currently trading at 2.5%, the highest level since 2008.
A perpetual inflation worrier, I habitually confront the fact that bond prices don't signal inflation. I am forced to point out that they never do -- interest rates did not forecast the inflations of the 1970s, nor the disinflation of the 1980s. And I say inflation is unforecastable, a risk like a California Earthquake.
But for once there does seem some inflation risk in asset prices.
These are option prices. The main forecast remains subdued inflation. But these option prices are pointing to a larger chance that inflation does break out. More risk, not so much a sure thing. Also, it's not really screaming -- after all, we're about at the prices of July 2018.
In Torsten's view, despite these prices,
Five years of CPI inflation above 2.5% or 3% is in my view extremely unlikely.
Velocity is quite low at the moment:
ReplyDeletehttps://fred.stlouisfed.org/series/M2V
When velocity picks up, combined with the fiscal expansion, inflation seems unavoidable.
What kind of asset prices are being used here? I'm asking because I'm wondering what the best way is for someone to bet on high inflation that starts five years from now. --best that is, in the sense of low transactions cost and high leverage, like out of the money options.
ReplyDeleteGiven: Nominal Debt/Price Level = PV expected primary surplus, The Biden administration is preoccupied with the right side of the equation. Recently, Goldman Sachs predicted 8% GDP growth fiscal 2021. If r < g guides the administration's fiscal policy and the administration accepts that prediction as gospel, they may well be emboldened to increase borrowing. Private citizens are more constrained when borrowing , unlike governments, who can tax, print fiat money and borrow. The left side of the equation is where they are either blind or accept as an acceptable cost in general equilibrium. If the price level ramps up, rates rise, bond holders, (lenders) are harmed. Will inflation increase as people dump dollars? It seems a fair question. If I use John's anticipated inflation of 2.5% per year for the next five years,
ReplyDelete(1-.025)^5 = .8810. If I anticipate losing nearly 12% of my buying power over the following five years, will I dump more dollars today? I might.
Under which probability measure?
ReplyDeleteAlthough it is not indicated, but the context suggests it's the physical measure. But, then again, it's a fair question with options -- e.g., Breeden-Litzenberger would give risk-neutral probabilities.
DeleteI'd like to broaden your question. At the risk of asking for something stupid, what is the methodology involved here?
I ask again hoping for an answer.
ReplyDeletehow does this occur for an extended time? what is the transmission mechanism(s)?
I can remember back in the 1980s when the Wall Street Journal opined that any rate of inflation under 5% was good enough, and the Fed should not over-tighten.the Reagan days.
ReplyDeleteIn the early 1990s, Milton Friedman chastised the Fed for being too tight when inflation was about 3%.
Would a few years of inflation in the 3% to 4% range really be so bad?
Every major central bank is facing an economy well below their inflation target, from Japan to China to the EU to the US.
So, what have fiscal and monetary authorities done to reach such low rates of inflation? And globally?
Would putting actual housing prices (instead of owners equivalent rent) back in the Consumer Price Index (CPI) really be so bad?
DeleteYes.
DeleteHere is a FRED chart from Japan. In 1997, the CPI in Japan was at 100. In January of 2021, it was was 101.6.
ReplyDeletehttps://fred.stlouisfed.org/series/JPNCPIALLMINMEI
So, call it 33 years at zero inflation.
If no inflation is the goal, then what are the lessons of Japan?
That no inflation, zero interest rates, low unemployment are a steady state?
DeleteWell...also the 1997-2021 period was also defined by low real economic growth, and a modestly declining, and aging population.
DeleteAdditionally, property zoning is light in Tokyo (the only real growing area) and so housing costs are held in check.
But Japan did run up large national budget deficits, that were then bought by the Bank of Japan (their central bank).
Does this central-bank buying back of debt give the public confidence that remaining outstanding debt will be easier to service? And so no inflation (from your perspective)?
Also, Japan runs chronic trade surpluses. A factor in confidence in the yen?
Benjamin Cole is making a great deal of sense. Government seems to be doing things that have caused inflation, but there are cases like Japan which seems immune to inflation.
DeleteThere are more reasons for inflation to stay low. See Ed Yardeni’s reasons at http://blog.yardeni.com/2020/12/inflation-was-sooo-1970s-will-it-roar.html?m=1
Printing money will only cause inflation if the money gets spent. Fiscal stimulus means that more will be spent than was in QE when banks couldn’t find any solvent borrowers to loan it to.
Income inequality means that rich people have money they save rather than spend, while large numbers of poor consumers can’t afford to buy what the economy could produce, if only someone would order the products. When interest rates hit zero, government needs to spend because no one else can or will.
Ray Dalio says we are at the end of a long term debt cycle. Private debt can no longer increase, so lower interest rates do not cause people to borrow and spend.
I see big forces on both sides of the inflation question. I think Ed Yardeni has presented the best reasoning.
John Cochrane: "A perpetual inflation worrier..."
ReplyDeleteMe too. That's why I keep coming back to your blog hoping for some reassurance. I'm not finding it.
Torsten Slok's description of the implied probability distribution, i.e., estimate A: prob(i>3%)=30% and prob(i<1%)=5%, and estimate B: E(t,t+1)(i)=2.5% are one the same. Taking the probability density of A ~ N(m,s) and plugging in the data [prob(i>3%)=30% and prob(i<1%)=5%] gives the following values for the mean (m) and standard distribution (s): m=0.0252 and s=0.0092. E(t,t+1)=0.025 for population B is consistent with distribution A ~ N(0.0252,0.0092) and implies B ~ A.
ReplyDeleteExpected inflation of 2.5% is in the range that Chm. Powell and others at the Federal Reserve have recently indicated to be within their anticipated central tendency in the next year to two year horizon.
What can we conclude from this? We can conclude, on the basis of Mr. Slok's evidence, that the market is conforming to the Fed Reserve's forward guidance for the rate of inflation out to circa 2022. Which is to say that the market has no special information or insight on the future rate of inflation (i.e., just as Mr. Slok's default working assumption would have implied).
Should we categorize this as further evidence of the weak-form of market efficiency, or evidence of the semi-strong form?
Whether there is to be inflation, deflation, or stagflation, I cannot say.
ReplyDeleteI am however, hedging both sides of my Gaussian curve with about 15% of bonds & cash firewalling the left tail of deflation risk and another 15% of commodities & tax free Midwest timberland on the right tail of inflation. My objective in this market of no nothing Millennial short squeezers, overconfident Bitcoin speculators, $30 trillion national debts, and inexplicably enthusiastic Tesla aficionados is to not get my own portfolio blown up.
Will the two hedges hold? Who really knows at this point.
>interest rates did not forecast the inflations of the 1970s, nor the disinflation of the 1980s. And I say inflation is unforecastable, a risk like a California Earthquake.
ReplyDeleteCausality is not coincidence (statistics). Mises says statistics are valid for economic history, not economics. Maybe an inflation hidden by other factors did predict the above events. Mises predicted the 1929 Depression because prices DIDNT decrease despite a big increase in tech. He also says inflation's effect cant be concretely predicted because each inflationary situation is concretely different.
You have to make this statement precise. Unforecastable usually means you're looking at a random walk or a white noise process -- something like that. This can be tested and you would find that the statement makes more sense nowadays -- trivial forecasting models are very hard to beat in the past 10 years --, but that ultimately you can do better.
DeleteYou don't just pull out if thin air "things change" to dismiss the possibility of doing something. You formalize the hypothesis and you test it out.
The future course of inflation will be determined in the Senate this year.
ReplyDeleteJohn - could you provide a link to the full presentation?
ReplyDeleteThat's all I have, sorry. Email from Torsten Slok at Apollo. I am also curious about just how to construct it.
DeleteI think these numbers are coming from the Minneapolis Fed. See https://www.minneapolisfed.org/banking/current-and-historical-market--based-probabilities
DeleteThis is a fantastic link, thank you! Others, enjoy the link. Many blog posts will come from it!
Delete@michael " Most importantly, the market-based probability accounts for how valuable resources will be in the future relative to today."
DeleteIt appears to be under the risk-neutral measure.
The pandemia implies a supply shock. Now the fiscal stimulus should be inflationary. I think prices will be increased depending on the duration of the pandemia. That is, if the supply shock continued to exert its negative effects, then inflation would be a real problem. Simply because the growth in demand would not be answered by means of an increasing production.
ReplyDeleteIt seems to me like new deflationary tech innovations (like amazon) and positive labour shocks overseas are stalling out creating the perfect storm for inflation especially with monetary and fiscal stimulus but I guess time will tell.
ReplyDelete"interest rates did not forecast the inflations of the 1970s"
ReplyDeleteHow far out? Eventually, interest rates rose so much that they did reflect high inflation expectations in the 70s, right?
To be fair, even if one believes that inflation will be whatever the Fed wants it to be (except for perhaps some short-lived supply shock inflation), inflation is still somewhat unpredictable because what the Fed will do or want in the far future is somewhat unpredictable. In the 1960s, it was probably unpredictable that the Fed would pursue an easy money policy in the 1970s.
So, the question of whether the Fed can control inflation, whether the Fed is currently credible on inflation, and whether the Fed will remain credible on inflation in the future are really three separate questions even though they all are relevant to the question of whether inflation is predictable.
I would say that the greatest long-term inflation threat right now is that there is an emerging contingent of (mostly young) economists that seem to believe that, if one just ignores the 70s, then one will be able to permanently "run the economy hot" to achieve permanently low unemployment and high output. (They also seem to ignore that, to the extent inflation reduces unemployment, it does so by lowering real wages.) I don't view that contingent as having captured the Fed, which is probably why the market doesn't currently expect inflation, but who knows in 5-10 years? Ten years ago I would not have expected so many today to view equality purely in terms of outcomes rather than equal rights and neutral processes. So even fundamental ideology can change quickly.
John, avid reader here, as you may know. What is your central/baseline inflation forecast in 2021 (let's say end-2021 yoy). What is your inflation forecast in 2022? Beyond the "inflation may be around the corner", are you concerned about the level of inflation in your baseline scenario? If you add the mechanism through which inflation rises in your central forecast, that would be a bonus.
ReplyDeleteI respect the art and science of forecasting too much to offer offhand numbers. See the link at Michael's comment above for market based forecasts!
Delete"In Torsten's view, despite these prices, 'Five years of CPI inflation above 2.5% or 3% is in my view extremely unlikely.'"
ReplyDeleteOn the contrary, given Chmn Powell's statements to the House and Senate committees, average inflation rates will have to be on the order of at least 2.5% acgr in order to meet the FOMC's 2% acgr for CPI inflation from, say, February 1, 2010 through February 1, 2026 (i.e., five years hence). Using data from FRED (https://fred.stlouisfed.org/graph/?g=Crsk) the average annual compound growth rate of inflation from 2/1/2010 to 2/1/2021 is 1.767% (s.d., 2.555%). The data is monthly observations converted to annual compound growth rates of the CPI. As can be seen from the value of the standard deviation for the eleven years worth of observations, the forward values of acgr of the CPI are likely to be noisy. Assuming a constant acgr of CPI for the next five years indicates that an annual compound growth rate of 2.494% (certainty) is required to meet the FOMC's five-year forward CPI target. The variability of actual inflation observations requires that the expected value for inflation must exceed the certainty value of 2.494% acgr. Ergo, inflation is likely to exceed 2.5% if the FOMC target for CPI is to be reached. A longer time horizon for reaching target will lower the required average acgr, but increase the uncertainty of reaching the target. No one is as yet certain what the FOMC has in mind. This enhances the uncertainty in the treasury note prices going forward. A difference of opinion makes horse races, and bond markets.
Something will happen. Given our debt and spending trends, we'll have inflation of much higher taxes at some point. Not sure what I should do with my long term savings.
ReplyDeleteWaiting for Inflation: https://mmt-inbulletpoints.blogspot.com/2017/10/blog-post.html
ReplyDeleteAll this hand-wringing over inflation is a total waste of time. Inflation is easily managed. For example, the Job Gty/Green New Deal law should include AUTOMATIC across-the-board tax increases that kick in when certain monthly wage inflation target are hit-say 5% annual rate for 6 months in a row. These can include:
ReplyDeletea) Income Taxes,
b) Sales/VAT Taxes
c) Asset Value (or Wealth) Taxes
That'll cool things off pronto.
Easy Peasy.
Why not automatic spending cuts instead? That way, we can minimize deadweight losses instead of maximizing them. Also, we won't have to worry about running out of Other People's Money to tax.
DeleteThe sovereign, like the US does not need to borrow or tax in order to spend. So right off the bat your comment exhibits a profound ignorance of how our economic/monetary system works. Automatic spending cuts are an option but if the objective is to maintain a Job Gty in place where unemployment is always at ZERO, automatic cuts would be counter productive. Moreover if the economy begins overheating, private sector bids for Job Gty workers and the program automatically shrinks. And if you think public sector is "Deadweight", answer me this: are the following "Deadweight"?: pulling your children out of a burning building, forming the thin blue line between your wife and getting raped, building and repaving the roads and bridges you drive over, cleaning the toxic waster dump left by some failed job creator, making sure most folks on the road know the rules of the road and can see straight, winning WW2, invent the internet, fund and execute our basic research, making sure your dullard kids are educated and can be functioning adults....(infrastructure; alternative energy; high speed rail; rehire every teacher, fire fighter, cop laid off in last 8 years; quintuple trade school and community college staff - free tuition, free elder and child care
DeleteI believe that it is you that is displaying ignorance.
Delete"The sovereign, like the US does not need to borrow or tax in order to spend."
The sovereign taxes in order to get people to accept government money when it tries to spend. So yes, the government does "need" to tax in order to spend.
If I am a book salesman and I don't have to pay taxes then why should I accept government money for one of my books?
"Automatic spending cuts are an option but if the objective is to maintain a Job Gty in place where unemployment is always at ZERO, automatic cuts would be counter productive."
That's disingenuous. Even with "Job Gty", wages on those "guaranteed jobs" can be cut automatically (Automatic spending cuts as BC described).
I tentatively assert that you are correct that the government does not need to tax and borrow in order to spend. Anybody can make an assertion. In a normal year when not fighting a recession as in 2009 or pandemic as in 2020, government spending is about 38% of GDP. If we cannot increase production by 38% each year, then it seems that prices will increase by some amount when the private sector tries to spend the money it did not have to pay in tax and the government in the same year also tries to spend the usual 38% of GDP. The increase might be near 38%. People can become accustomed to prices increasing. If the rate of increase is reliable and stable people can make good plans. How likely is the expectation to be off by half? How likely is inflation to be only 19% or off by half in the other direction, so 57%. Savers would feel regret that they over-saved if inflation turned out to be 57%. If 19% then they regret under-saving. Buyers of homes, cars, machinery, businesses would face regret if the inflation rate were unstable. Perhaps in this high inflation environment, realized inflation is less likely to be off by half than it is for "managed" inflation of 2% to be off by half, resulting in realized values of 1% or 3%. Despite convenient access to computers, people might feel uncomfortable with large rates of change and may be it will be hard to adapt.
Delete"And if you think public sector is Deadweight, answer me this: are the following Deadweight?: pulling your children out of a burning building; forming the thin blue line between your wife and getting raped, building and repaving the roads and bridges you drive over, etc."
ReplyDeleteWhen those public sector employees are guaranteed a job irregardless of their own abilities and past performance, I would say in some cases they are worse than dead weight - they are a dangerous liability.
Would you want your life to depend on a fireman or policeman who was guaranteed a job regardless of actual skill or training?