I have been updating graphs for my money and banking course, and here is the graph I generated to illustrate the tremendous impact of government borrowing on interest rates via portfolio crowding out (as argued in this post).
Figure 1: Ten year constant maturity Treasury yields (blue), Ten year constant maturity inflation indexed yields (red), and real ten year rate implied by expected ten year inflation (purple +), in percentage points. Observations for September are 9/22. NBER defined recession dates shaded gray. Source: St. Louis Fed FRED II, Treasury constant maturity rates, expected inflation from Philadelphia Fed Survey of Professional Forecastters, NBER, and author’s calculations.
Oops. Well, the tremendous impact isn’t showing up now. It might in the future. Although, it’s useful to recall that the ten year interest rate is, under the expectations hypothesis of the term structure [0], the average of the expected short term interest rates over the next ten years. Holding the term premium constant, this outcome implies that expected short rates are falling. One can localize the low interest rates by inspecting the interest rates of different maturities. For the next five years, we can look at the 5 year constant maturity rates.
Figure 2: Five year constant maturity Treasury yields (blue), and five year constant maturity inflation indexed yields (red). Observations for September are 9/22. NBER defined recession dates shaded gray. Source: St. Louis Fed FRED II, Treasury constant maturity rates, NBER, and author’s calculations.
On 9/22, the 5 year TIPS yield was 0.01%.
Other commentary: [Free Exchange/RA] [Krugman] [CG&G/Collender] [CG&G/Collender] [DeLong].
Excuse me? Are we ignoring the impact of the Fed printing money to buy Treasury debt and manipulate yields lower?
And are we cherry-picking our data points? Love to talk about what TIPS are saying but no comment on what gold is saying?
Of course expected short rates are falling. The Fed has signaled ZIRP for the foreseeable future. That does not imply that structural deficits of 10% of GDP will play out well.
W.C. Varones: Do you really understand the concept of “structural deficits”? I sincerely doubt it. Where exactly are you getting your structural deficit estimates of 10% going forward? Definitely not from the CBO (see here), and not even under their alternate fiscal scenario.
The recent study from J D Hamilton and C Wu (the effectiveness of alternative monetary policy tools in a zero lower bound environment).
This study is showing that the portfolio crowding out theory can be easily circumvented when the CB is swapping long term securities for the purchase of short term TB The same study quantifies the amount required based on 2006 debt stocks that is 400 USD billions swaps of long term debts for the equivalent purchase of short term maturities in ZLB environment would have decreased the 10years yields by 14 basis point.
Few questions remain where is the wealth creation?Every time the stock of TB is increasing the LM curve is shifted downwards and so does i? Would an opposite action from the market (market selling 400 billion usd TB) and not from the Fed produce an uplift of only 14 BP?
The study is discarding the shadow banking instruments (Interest rates swaps,bonds futures,options).The sames are assuming that cash and only cash can move an asset price, when cash and contracts default swaps CDS may have as well a significant contribution on i.
The private investor may well be confused when seeing the cash component of an investment only.
Could it be that private investors are underwriting the risk and treasury when the shadow banking is creaming the volatility?
Hi Menzie,
“This outcome implies that expected short rates are falling”. Following Krugman’s posts on the subject, you’ll find in the following post a more detailed analysis on the expected future short rates implied by the current 10Y yield (plus alternative scenarios).
http://www.market-melange.com/2010/09/13/fed-policy-and-the-yield-curve/
As for TIPS, liquidity and issuance size still remain a key issue…
The CBO sees long-term revenue growth of 7%, outlay growth of just 4.5%, and a CPI of just 2.1%. That’s somewhere between rose-colored and lunatic asylum.
But who am I to question them? After all, these guys predicted the housing crash and economic collapse so well, right?
“…no comment on what gold is saying?”
Gold is saying “SUCKERS!!!!” Pretty much like it was saying in 1980. What does that have to do with the discussion at hand?
ckb,
We didn’t have Ben Bernanke, Janet Yellen, 10% deficits, or global competitive devaluation in 1980.
We had Paul Volcker.
I suspect the crowding out will be on the fiscal, not monetary, side. I don’t believe that the pillar-of-the-community crowd believes current and forecast deficits can be sustained. Consequently, as the administration has signaled its intent to turn primarily to this group for tax revenues, those with good incomes are bracing for five digit tax increases.
Let me give you a specific example, again from my home town of Princeton, of what this might mean.
The typical home–not bad, not great–in the township or boro costs about $750k. Do the math, and assuming 35% of net income goes to a mortgage and real estate taxes, the required income is about $270,000. So, as in Lake Wobegon, the average resident of Princeton is above average, in fact, in the top 3% of earners on a household basis. (Probably true for the top third, actually.)
Now, this cohort will be hit with about a $10,000 tax bill if the Bush cuts expire. Again, do the math, and ceteris paribus, that after-tax income now supports a house around $700,000–a decrease of about 7% or so.
Thus, the increased taxes to support the deficit might be anticipated to ‘crowd out’ house values by about 7%, at least in the better parts of Princeton.
And consider the effect on prospective buyers: Should they buy now or wait for the tax increases to take effect?
Menzie,
Thanks for a post showing the MMT guys are right yet again.
wcv: I’d like to get your prediction on (virtual) paper for reference in the future: if I have it right, your forecast is a structural budget deficit of 10% of GDP for the foreseeable future, including FY2011 and FY2012, and so on, onto FY2020.
Gregory Gadzinski: Yes, I agree there are liquidity issues exist in the TIPS markets. That’s why Figure 1 includes the 10 year nominal Treasury yield minus expected ten year inflation from SPF.
I expected Fed buying of Treasuries was the reason for the lack of visible “crowding out” in real interest rates, we would expect to see the following:
-A weakening dollar index
-Record gold prices
-Record Raw Materials CRB Index prices
-Positive 5yr TIPS inflation expectations and the real TIPS yield at zero bound
In other words, the set of observations is consistent with the view that the price of Treasuries is set, on the margin, by expected Fed purchases; and that these purchases are increasing future inflation uncertainty.
The alternative explanation is that the above inflation indicators are all due to high emerging markets real growth.
Both explanations are valid. I’m not sure which one you subscribe to, but it seems you are quite certain that Treasury price signals are “market determined”. Perhaps such certainty is unwarranted.
“under the expectations hypothesis of the term structure”
Well under the WishesWereTrue Hypothesis all we need to do is wish to be rich and we will be.
The expectations hypothesis is a bad theory. But now its not a bad theory, ITS A F****** [edited for profanity — mdc] HORRENDOUS THEORY. Investors are afraid and when they are afraid they look for safety. They certainly don’t sit around calculating expected interest rates and arbitraging correctly.
The current term structure tells us nothing about future interest rates.
“Oops. Well, the tremendous impact isn’t showing up now.”
What this tells me is the market does not think the Fed will create inflation. With excess capacity, I don’t think the Fed can create steady, controllable inflation – its choice is between mild deflation and inflation completely relying on expectations being self-fulfilling, which could become unmoored. This is uncharted territory, and I don’t think Ben would knowingly tread to far from shore. However, if he continues trying to lever the U.S. (and pegged Asia) on the backs of the euro economies, he may create pressures that will result in sovereign defaults among the marginal euro bloc members, which could lead to some very ugly choices going forward.
Could someone please summarize the evidence that suggests that forward expectations in the bond market are realistic, accurate, and worth considering as reflecting anything other than contemporary investor sentiment and preferences for asset allocation?
The bond market was wrong in 1970, wrong in 1980, wrong in 2003, wrong in early 2008… the future did not conform to the market’s expectations at any of those times. Why should it be right now?
The argument that the FedGov’s borrowing is “safe” because “rates are still low” is meaningless, because rates aren’t a rational indicator in the relevant sense.
As for those saying that the FedGov’s borrowing hasn’t been crowding out other investments, because rates are still low: try to explain that to the Greeks, Irish, Portuguese and Spanish! Dubai? Argentina? Russia in the 1990s? Their rates were nice and low, too — until they weren’t. The same is true for most prior sovereign default situations as well.
There will be a lot of crowding out when those financing the bond market no longer permit the high deficits. It will be either through higher taxes, lower spending (ex-interest payments), partial default (money-printing or devaluation), or outright default.
What I see is that there is still too much credit in the system, leading to overvaluation and/or oversupply of stocks, bonds, housing and commercial real estate. Corporate profits are near record historical highs as a share of GDP; there is not much room for growth there unless GDP itself grows. But there is little anyone is willing to invest in, to grow GDP, in part because existing capacity is more than adequate to meet current and near-term demand.
Menzie,
I believe 10% deficits are unsustainable, therefore they will not continue indefinitely. Something will break first.
With so many variables (Fed monetization, China vendor financing, U.S. Government backstop of entire U.S. mortgage market, etc.), I don’t have the hubris to try to predict GDP and deficits 10 years out. But I am fairly confident that the CBO’s prediction of sustained, robust economic growth in conjunction with low inflation will be very, very wrong.
wcv: OK, you seem to be iterating, but then your original comment seems to indicate you believe there will be structural deficit of 10% at least in FY 2011 (starting in a few days) and perhaps FY2012 as well. Can I put that down to (virtual) paper as your forecast, for the record?
David Pearson: Joe Gagnon’s estimate of the effect on ten year rates (duration adjusted) was 58 bp. Adding that in still leaves a pretty low interest rate.
Menzie Chinn’s post said: “W.C. Varones: Do you really understand the concept of “structural deficits”? I sincerely doubt it. Where exactly are you getting your structural deficit estimates of 10% going forward? Definitely not from the CBO (see here), and not even under their alternate fiscal scenario.”
For Menzie, W.C., and others, are you assuming this economic situation is an aggregate demand shock or an aggregate supply shock?
Menzie,
I don’t devote a lot of time to federal budget forecasting, so I’ll trust the Administration’s estimates for next year. The New York Times reports $1.4 trillion, or 9.2% of GDP. That’s the forecast for two years after the recession ended. If it’s not structural, why are we still running it so long into the recovery?
GetRidoftheFed,
I’d say the Dirty Fed created both excess consumption and excess investment by blowing bubbles and encouraging leverage with easy money. Now we have excess capacity and reduced demand.
Menzie,
Did Joe Cagnon include the impact of the Fed buying $1.45tr in Agencies and RMBS, which the banks, in aggregate, sold to purchase Treasuries? In other words, obviously, the Fed’s purchases in one obligation of the government might have an impact on substitute assets. And did he also incorporate expected future Fed purchases and their likely impact on bond prices? Certainly expected future demand from the Fed has an impact on today’s price.
If he did both, then I would want to take his work as a guide to the Fed’s impact on Treasury prices. If he did neither, I cannot imagine how the analysis could useful.
wcv said: “I’d say the Dirty Fed created both excess consumption and excess investment by blowing bubbles and encouraging leverage with easy money. Now we have excess capacity and reduced demand.”
If there was excess consumption with debt and excess investment with debt, why wasn’t there price inflation? Is there some other “imbalance” that allowed more and more debt to be produced without price inflation?
@wvc
I agree about excess capacity and I’d like to add that most of this excess capacity is located outside of the US.
The USA has actually too little capacity in many industries and therefore needs to rely on imports despite the negative trade balance.
I also find the argument unconvincing that the current low interest rates indicate no crowding out. The current low long-term interest rates are to some part caused by the Fed buying bonds and thus driving down rates. Who knows where rates would be without this intervention. Maybe they would be just a little higher and the point of the original post would hold even then. But as the rates are rigged, I don’t follow the argument.
One avenue I would be curious to explore is the relationship between corporate credit spreads and treasury issuances. You would obviously have to control for recessions and exogenous financial shocks otherwise the results would become distorted.
On a purely theoretical basis, one would expect that the more marginal corporate issues would see their spreads compared to treasuries widen as overall debt market issues increase due to government borrowing. The expected result would be that Aaa corporate credit spreads would be slightly wider and Baa and below spreads would become progressively wider. I ran a very primitive regression on this front and discovered a statistically significant, though small result that supported this hypothesis. However, with better controls for exogenous financial crises, what results there were might have disappeared.
Ketzerisch: “I agree about excess capacity and I’d like to add that most of this excess capacity is located outside of the US.
The USA has actually too little capacity in many industries and therefore needs to rely on imports despite the negative trade balance.”
I agree, and I think a good part of this resource misallocation was caused by currency interventions.
Ketzerisch: “I also find the argument unconvincing that the current low interest rates indicate no crowding out. The current low long-term interest rates are to some part caused by the Fed buying bonds and thus driving down rates. Who knows where rates would be without this intervention.”
Your argument here seems to imply that private investment is so interest sensitive that even at such low rates there may be significant displacement. This seems unlikely – business investment is notoriously insensitive to interest rates.
Ketzerisch, I agree.
Get Rid, I think technological productivity enhancements and cheap overseas outsourcing provided a countervailing force that kept inflation in check. We had deflation in electronics but inflation in other areas like real estate, food, and energy.
Menzie,
From your lecture,
Portfolio crowding out arises because higher government spending is associated with higher
bond sales, hence higher wealth, and hence higher demand for money which, given the fixed
money supply, results in higher interest rates for all income levels.
How does the dramatic reduction in transactions we have experienced effect your LM curve? How do you introduce quantitative easing that matches every government bond sale with a FED purchase into the model. When the FED purchases debt from the Treasury, does that increase wealth? How does the slope change, given a large amount of excess capacity so that firms are unwilling to invest at any real interest rate?
I think the model predictions match reality when you introduce realistic assumptions.
No one can trust what they see on there screens because of the poorly thought nonsense of the last few years. I completely agree with wcv that just because bonds are trading off the fed doesnt mean they signal anything or are priced correctly. We know we have a leverage problem that has been swept under the wrong by the idiots we vote for. The country has cancer.
“If there was excess consumption with debt and excess investment with debt, why wasn’t there price inflation?”–Get Rid of Fed
Price inflation occurred 1st in production goods, houses, durables–things for which people borrow money. The typical pattern with credit expansions. Then money gradually moves to consumer goods, & indeed the CPI heated up to 4.1% in 2007 before the collapse.
David Pearson: Yes. The study by Joe Gagnon et al. has been cited several times on Econbrowser, but here is the link again. You might find it useful to read.
tj: The point of imposing the ceteris paribus assumptions, and tracing out the model implications is, then, you can impose your own assumptions regarding for instance the other components of autonomous spending. A specific example: with the other components (of investment) shrinking, the increase in G and reduction in lump sum taxes only shifts out the IS curve a small distance. Then the combined effects of transactions and portfolio crowding out should be small.
In other words, I didn’t mean for the assumptions in the handout to be realistic; rather it was pedagogical in nature. I leave the more realistic implementations for problem sets and exams.
By the way, when the Fed buys Treasuries, and pays with high powered money, net wealth of the private sector (i.e., outside assets) is unchanged.
ketzerisch: Please take a look at the Gagnon et al. paper for estimates of the impact of the LSAP.
wcv: Clearly you don’t understand what a cyclically adjusted budget balance is (or are using a funky time series filter to get your measure of the output gap). See this post for some pictures, and a link to the CBO document that provides forward looking projections of the cyclically adjusted budget balance.
So, I’m going to put you down as predicting a cyclically adjusted (aka structural) budget deficit at near 10% for FY 2011 and FY 2012. I look forward to revisiting your forecast in a year.
“In other words, I didn’t mean for the assumptions in the handout to be realistic; rather it was pedagogical in nature.”–Menzie
How can something which is pointedly misleading be pedagogical? To ignore the monetization of Federal debt by the Fed, the unique situation of the free money made available to the banks [with which they in turn buy gov’t debt], & foreign central banks [esp. Japan presently] buying US gov’t debt by the billions…well, to consider this problem without considering very relevant data is not to teach but confuse.
Menzie,
No, I’m not making 2012 deficit forecasts.
What I’m saying is the CBO forecast of a permanent Goldilocks scenario of rapid growth and low inflation is not going to happen.
Here’s what you should write down and hold me to: the U.S. will not experience a decade of at least 7% annual government revenue growth concurrent with at most 2.1% inflation and at most 4.5% expenditure growth, which is what the idiots in the CBO are forecasting.
wcv: Then what of FY 2011? You did say 10% deficits for some time period…Or do you retract that forecast now?
Bryce: We sometimes provide supply and demand examples for widgets where price elasticity of demand is assumed to be high. Doesn’t mean we believe it — just that we want to illustrate something. Similarly, I can show the mechanics of crowding out holding constant money supply. Then modifying (as I do in lecture or problem set question) to allow the Fed to monetize is trivial — once the mechanics are clear. If you teach economics, then you might find your method superior, but this has worked for me for the past 24 years.
Menzie,
An announcement study? Was each announcement a complete surprise to the market, or did traders perhaps front-run the likely Fed actions? And did expectations for QE end, neatly, with the last announcement, or did the market already have the Fed’s possible “exit plan” in mind?
The study’s author could continue the study to incorporate a possible QE2 announcement on or around Nov. 3. How would the study incorporate the fact that traders spent months forming expectations about the Nov. 3 meeting?
Menzie,
That’s a White House OMB forecast, not mine. If you’re looking for someone to retract it, ask Obama.
W.C. Varones: The linked to article pertains to the deficit. In your original 9/27, 7:48PM post, you wrote:
The operative adjective is structural. Do you now retract that forecast?
Time to abandon your theory? Seriously, this is a complicated issue, it would be shocking if something as simple as your graph worked.
Dave Backus: I apologize: I didn’t mean to convey the messagge that one should drop the theory of either portfolio crowding out, or the expectations hypothesis of the term structure (especially since I use both of them in different papers). Rather, my point is in making predictions, one wants to condition on other factors that typically people drop out of the analysis. In the portfolio crowding out theory, with low demand for credit should result in reduced interest rates; with respect to the EHTS, low expected future short interest rates are consistent with what we see in long rates. Hence, in neither case are these interpretations a refutation (note the correct spelling!) of the models.
wcv said: “Get Rid, I think technological productivity enhancements and cheap overseas outsourcing provided a countervailing force that kept inflation in check.”
Add in cheap legal and cheap illegal labor and make that outsourcing in tradable goods to keep price inflation in check.
And, “We had deflation in electronics but inflation in other areas like real estate, food, and energy.”
There are probably too many to list, but the point stands, and I agree.
Bryce said: “If there was excess consumption with debt and excess investment with debt, why wasn’t there price inflation?”–Get Rid of Fed
Price inflation occurred 1st in production goods, houses, durables–things for which people borrow money. The typical pattern with credit expansions. Then money gradually moves to consumer goods, & indeed the CPI heated up to 4.1% in 2007 before the collapse.”
I agree there was price inflation in housing, energy, and somewhat in food. I don’t believe there was price inflation in durables/production goods that were tradable (especially tradable goods from china).
See this link:
http://www.nakedcapitalism.com/2010/01/why-bernankes-defense-of-super-low-interest-rates-does-not-hold-up.html
Menzie Chinn, how about putting up some numbers for the federal budget between 2007 and 2013?
Then break those numbers down into what you believe are structural vs. cyclical?
don said: “This seems unlikely – business investment is notoriously insensitive to interest rates.”
Bingo! Businesses pay a lot more attention to their budgets than consumers.
Is it a lot easier to try and trick people into going further and further into debt so they end up retiring later than tricking businesses into going further and further into debt (businesses don’t need to retire)?
@Menzie
Thank you for pointing me to Gagnon et al. (2010). But after reading it, I find my point supported in it. It says in the abstract
“We present evidence that the purchases led to
economically meaningful and long-lasting reductions in longer-term interest rates on a
range of securities, including securities that were not included in the purchase programs.”
Which was just my point (Sorry for having my trader’s view on things): Rates are set by supply and demand for longterm bonds. Fed is providing extra demand and thus drives down the rates.
Your argument was: Given constant term premium, the expectation hypothesis implies that expected short rates are falling. I didn’t buy this argument, because the rates are rigged by the artificial demand from the Fed. As I understand it, Gagnon et al. (2010) supports my point as they write:
“These reductions in interest rates primarily reflect lower risk premiums, including term
premiums, rather than lower expectations of future short-term interest rates.”
ketzerisch: I think we have been and still are in agreement. I didn’t mean to say LSAP had no impact on the term premia. What I was trying to show in the graph was that given the current low demand for credit in the economy, and current monetary policy, one is not seeing any upward pressure on interest rates.
As I indicated in previous posts, LSAP seems to have had an impact on interest rates, independent of total amount of bonds. That impact is consistent with a preferred habitat/segmented markets view. I’m happy with that interpretation. I’m further willing to say in the absence of LSAP, interest rates would be higher. At the ten month maturity, the impact is about 58 bps.
I hope that clarifies that we are in agreement on at least these points.
That being said, even with the 58 bps, interest rates — both real and nomninal — are pretty low.
Menzie,
From your portfolio crowding out notes:
“Portfolio crowding out arises because higher government spending is associated with higher bond sales, hence higher wealth, and hence higher demand for money which, given the fixed money supply, results in higher interest rates for all income levels.”
It is not clear how higher volume of bonds is associated with higher wealth. Surely there can be substitution effects across different asset classes but wealth increasing?
JimK: You have to understand that the model is highly stylized, and there are only two assets — money and bonds. In the simple exposition, wealth (outside assets) only changes due to changes in government liabilities, in this case government bonds. When there are more than two assets (say equities), then one can get crowding in as well as crowding out (see this post).
Not being an economist, I did notice that deflation expectations are not part of this thread. Maybe deflation is irrelevant to this conversation.
Although the price of gold has soared, probably due to two factors, one being safety, the other being fear of inflation sometime down the road, at least one possible reason for low interest rates is the fear of deflation going forward, plus the fear of safety and the resulting demand for AAA stuff.
Anyway, just wondered if deflation fears could play any role in this discussion.
W.C. Varones: Still waiting to see if you retract your forecast of “structural deficits of 10% of GDP” going forward.
I’ll keep on asking as long as you don’t answer… as I do want to get you “on the record”.