The Atlanta Fed's macroblog provides commentary and analysis on economic topics including monetary policy, macroeconomic developments, inflation, labor economics, and financial issues.

Authors for macroblog are Dave Altig, John Robertson, and other Atlanta Fed economists and researchers.

« July 2012 | Main | September 2012 »

August 29, 2012

Rising House Prices: The Good Fortune Spreads

On the heels of a rash of pretty good news related to residential real estate—including yesterday's pending home sales report—the June S&P/Case-Shiller report on housing prices checks in with positive monthly gains across all markets in its 20-city composite for the second month in a row. What's more, the index posted its first year-over-year gain since last summer.

The early reviews found little to dislike, from Calculated Risk...

This was better than the consensus forecast and the change to a year-over-year increase is significant.

...to Carpe Diem...

More evidence that the U.S. housing market has passed the bottom and is now in a period of sustainable recovery.

...to TimeBusiness...

[T]he housing market is steadily improving and is poised to contribute to economic growth this year. Modest economic growth and job gains are encouraging more Americans to buy homes.

The widespread nature of price firming evident in the Case-Shiller index is strikingly confirmed by looking at even more disaggregated data. The following chart shows June year-over-year price growth by zip code, before the crisis hit and since, based on data available from CoreLogic:

The sample represented by the chart covers about 21 percent of all of the zip codes in the nation, and is based (like Case-Shiller) on a repeat-sales methodology.

The striking aspect, of course, is that there haven't been price increases in the majority of the sample's zip codes since before 2007 (although there was improvement evident in 2010, followed by the re-emergence of broader weakness in 2011). Furthermore, the uniformity of the picture becomes even more apparent when you look market by market (across which the experience is not so uniform). Two of the big comeback stories—Miami and Phoenix—were uniform in the breadth of the suffering across their metro areas during the worst of the slump and are now just as uniform in recovery:

Folks in Atlanta, on the other hand—which remains the big negative outlier in the year-over-year Case-Shiller statistics—are just as uniform as Miami and Phoenix, but in the pain rather gain department:

Even so, the Atlanta market has had two consecutive months of Case-Shiller housing price appreciation and experienced the largest monthly percentage gains in the June report. It does appear that the rising residential real estate tide is raising most boats.

David AltigBy Dave Altig, executive vice president and research director;

Myriam Quispe-AgnoliMyriam Quispe-Agnoli, research economist and assistant policy adviser; and

Jessica DillJessica Dill, senior economic research analyst, and all with the Atlanta Fed

August 29, 2012 in Housing, Real Estate | Permalink


TrackBack URL for this entry:

Listed below are links to blogs that reference Rising House Prices: The Good Fortune Spreads:


the same analysts who panned the homebuyer tax credits as temporary market distortions are ignoring the temporary distortion in home prices caused by operation twist induced record low interest rates, which effectively reduces the amount a buyer pays for a home even as the list price rises..

the average interest rate on fixed rate 30 year mortgages in July was 3.55%, a full percentage point lower than Freddie Mac's had the 30 year mortgage rate at a year ago...a simple mortgage calculation shows that the monthly cost per $100,000 on a 30 year mortgage in july of 2012 was $451.84, compared to the $509.66 per $100K one would have paid monthly on a 30 year mortgage last July; that means to buy the same house a year ago would have cost a potential homeowner 12.8% more in payments monthly than it would cost under current interest rate regimes...so even should July's home price indexes show a 2.8% year over year gain in the principal price of the house, it would mean that potential home buyers are still commiting 10% less to homeownership than they were a year ago...the so-called housing recovery is merely a fiction of low interest rates, which will not stay this low forever...

Posted by: rjs | August 29, 2012 at 07:32 PM

Just like the unemployment figures, which don't reflect the numbers of those unemployed who have exited the job market, the housing data do not reflect those houses that have been taken off the market nor do they reflect those houses that have not been put on the market due to depressed prices. Once a steady trend of home sales begins, those who have been waiting on the sidelines will put their homes on the market driving prices down once again. The solution is not short term. The reality is, we are in the midst of (and have been for the past three years) the greatest redistribution of wealth in our Nation's history.

Posted by: Kirk Wiles | September 04, 2012 at 01:39 PM

Post a comment

Comments are moderated and will not appear until the moderator has approved them.

If you have a TypeKey or TypePad account, please Sign in

August 24, 2012

The Cost-Benefit Challenge

In its latest Room for Debate feature, The New York Times poses the question "Should the Fed Risk Inflation to Spur Growth?" Befitting a balanced panel of blogging experts, Mark Thoma (Economist's View) says "yes," John Cochrane (The Grumpy Economist) says "no," and Edward Harrison (Credit Writedowns) says something like "irrelevant question, it's going to do neither."

The whole discussion, naturally, is about differing assessments of the costs and benefits of additional monetary stimulus. Not surprisingly, this was also a theme disclosed in the just released minutes of the July 31–August 1 meeting of the Federal Open Market Committee:

Participants also exchanged views on the likely benefits and costs of a new large-scale asset purchase program. Many participants expected that such a program could provide additional support for the economic recovery both by putting downward pressure on longer-term interest rates and by contributing to easier financial conditions more broadly. In addition, some participants noted that a new program might boost business merits of purchases of Treasury securities relative to agency MBS. However, others questioned the possible efficacy of such a program under present circumstances, and a couple suggested that the effects on economic activity might be transitory. In reviewing the costs that such a program might entail, some participants expressed concerns about the effects of additional asset purchases on trading conditions in markets related to Treasury securities and agency MBS, but others agreed with the staff's analysis showing substantial capacity for additional purchases without disrupting market functioning. Several worried that additional purchases might alter the process of normalizing the Federal Reserve's balance sheet when the time came to begin removing accommodation. A few participants were concerned that an extended period of accommodation or an additional large-scale asset purchase program could increase the risks to financial stability or lead to a rise in longer-term inflation expectations...

The questions about the costs and benefits of any particular policy intervention are abundant, and for virtually every potential pro there is a potential con. Here is my personal, certainly incomplete list of pros/cons or benefits/costs associated with another round of large-scale asset purchases:

Pro:  Lower interest rates (and perhaps a lower dollar) will on balance spur spending.
Con:  The expectation of low interest rates for a longer period of time will reduce the urgency to borrow and spend.
Pro:  Expanded asset purchases and lower rates will preserve needed liquidity in financial markets.
Con:  Expanded asset purchases and lower rates will create or exacerbate financial market distortions.
Pro:  More monetary stimulus reduces the probability of an undesirable disinflation in the near term.
Con:  More monetary stimulus increases the probability of undesirable inflationary pressures in the longer term.
Pro:  Lower Treasury and MBS rates will induce an appetite for risk taking that is needed to get productive resources "off the sidelines."
Con:  Lower Treasury and MBS rates will induce an appetite for risk taking that sets us up for the next bubble.
Pro:  Monetary policy is the only channel of support for the economy, absent new fiscal policies.
Con:  Monetary policy support is relieving the pressure to make needed fiscal reforms that would be much more effective than monetary stimulus.
Pro:  With additional monetary stimulus, GDP growth will be higher and unemployment lower than they would otherwise be, and outcomes may be more consistent with the FOMC's mandate to promote maximum employment.
Con:  With additional monetary stimulus, the exit from monetary stimulus once the economy improves will be more difficult than it would otherwise be, and outcomes may be inconsistent with the FOMC's mandate to achieve price stability.
Pro:  The performance of the economy has not been consistent with the FOMC's mandated objectives.
Con:  The economy is slowly moving in the direction of the FOMC's mandated objectives, and the Fed should "keep its powder dry" in case of further deterioration of the economy.

Many, if not all, of these benefits and costs are familiar, and there has been no shortage of opinions advanced. To some extent, it is inevitable that the weighting of these costs and benefits will be to a large degree judgmental. How one weights the risks associated with continued high rates of unemployment versus the risks of imbalances that may arise from low interest rates, for example, is a subjective thing.

Nonetheless, it would be helpful to frame these subjective judgments with a background of some hard evidence. For example, how much employment can we gain for a given quantity of asset purchases (or any other monetary policy option on the table)? What are the likely—or existing—distortions created by low interest rates, and what do we think are the tangible costs associated with those distortions? And so on.

So, here is my challenge question: No matter what your opinion about what should be done, and whether you arrive at a conclusion by casual observation, econometric studies, or historical evidence, what do you think is the best evidence concerning any or all of these costs and benefits?

OK, then. Bloggers blog, commentators comment.

Update: Jon Hilsenrath sizes up some of the costs of a new bond-buying program.

David AltigBy Dave Altig, executive vice president and research director at the Atlanta Fed


August 24, 2012 in Federal Reserve and Monetary Policy, Monetary Policy | Permalink


TrackBack URL for this entry:

Listed below are links to blogs that reference The Cost-Benefit Challenge:


I think the Fed has destroyed its credibility on the full employment front and hence the aggregate demand front. Expectations are coming completely unmoored.

For example, the looming 'fiscal cliff' would be considered a non-event if the Fed could be counted on to maintain nominal demand. However, years of opportunistic disinflation confuse any signal and make recession much more likely next year, thanks to political dysfunction and Fed fecklessness.

Without credibility on the AD front you are all making your own jobs much harder than it should be.

Posted by: OGT | August 25, 2012 at 09:45 AM

Good, hard question. But we might not have to answer it if fiscal policymakers would do a much easier cost-benefit analysis. The U.S. govrnment can borrow money in the TIPS market for 30 years at a real rate of 0.42%. Surely, SURELY, there are some long-lived infrastructure investment projects that aren't underway right now that will generate an internal rate of return higher than tat over their useful lives.

Posted by: ASG | August 27, 2012 at 12:20 PM

unfortunately you did not number the arguments.

con #1,3,4,6,7 are arguments that apply to any period of Fed easing. If you accept these then the FOMC should disband, do nothing, and never try to ease, because cutting rates deeper means makes it takes more time to raise them back to neutral. They are also in direct contradiction to Bernanke's testimony that the Fed has suitable room to unwind the balance sheet. Implicit in most is the booms-cause-busts theory of the business cycle.

If the Fed is steering the bus, then it can control the speed.

Con #2 has been rebutted by the Fed's internal research, and we have not seen any evidence for it.

Con #5 accepts the flawed theory that the FOMC should be defacto Senators or elected representatives that put pressure on Congress. It accepts the flawed, disproven, German idea of expansionary fiscal contraction.

Fundamentally, most of these arguments against action are tantamount to the business cycle solves itself and the Fed cannot control the economy. If that's the case, then we should disband the FOMC and save the taxpayers money. Really, FOMC members should thus be justifying their jobs.

Posted by: dwb | August 27, 2012 at 02:10 PM

can i posit a different cost/benefit analysis? with the fiscal mechanism broken? I fail to see who is benefiting from zero rates?!?! so at this point i can name a whole host of cons and less and less pros for the zero bound. For the sake of keeping it short I'll take the first shot across the bow on the zero rate regime, INTEREST INCOME 2007 was $492bln now its $64bln a DROP of $428bln.. WOW, meanwhile (roughly) DEBT Servicing has dropped $195bln.. MAYBE INCOMES WILL GO UP WITH RATES?? thats just the beginning of a very long argument. The purpose of low rates was clear before, but becomes increasingly unclear to me over time.

Posted by: cidiel | August 29, 2012 at 09:43 AM

Thanks - love the pros/cons list. Keep them coming!

Posted by: RebBowDur | September 04, 2012 at 07:05 PM

When we analysis cost and benefit analysis on any business so, it changes business to business. As, it depend o the business which we are handling. Before making any policies, we should first analysis it's pros and cons, as its meter a lot in any business.

Posted by: joshef | December 14, 2012 at 01:31 AM

Post a comment

Comments are moderated and will not appear until the moderator has approved them.

If you have a TypeKey or TypePad account, please Sign in

August 23, 2012

Do Firms Have Pricing Power?

Here's a question every policymaker would like to know the answer to: How much slack is there in the economy? You know the drill: If there's lots of slack, firms have little power to pass cost increases on to customers. But as the economy approaches full employment, the pricing power of firms strengthens, and cost pressures get more readily passed along. The problem is, there doesn't seem to be any practical consensus about the amount of slack in our economy. To paraphrase a well-known adage widely attributed to Nobel Laureate Ronald Coase, we've been torturing the Phillips Curve, but it isn't talking.

We wondered if businesses might be in a better position to say what kind of pricing power they have. So this month we asked our Business Inflation Panel to weigh in on how they might adjust their prices in the face of a hypothetical, unanticipated increase in costs. Not knowing what kind of cost pressure a firm might respond to, we randomly sorted our panel into two groups—one given a 2 percent cost-increase scenario, and one given a 6 percent cost-increase scenario. Here's what we learned:

On average, firms faced with the 2 percent cost increase were likely to pass about 1.3 percentage points (or 66 percent) to their customers. In the portion of our panel considering the 6 percent cost increase, the average impact on customer prices was 3.8 percentage points (or about 63 percent of the cost increase). So in the aggregate, firms think they can pass about two-thirds of any cost increase through to prices, and that belief holds roughly true whether the cost increase is relatively modest or somewhat large.

We asked a similar question of our panel last October, but the responses we got then indicated that firms held a more conservative view of their pricing power.

What effect would an unticipated, 2 percent rise in unit costs have on your prices

Of the firms facing the 2 percent cost-hike scenario last October, only 31 percent said they would pass along most of the cost on to their customers, compared with the 53 percent that said they would do so today (see the chart above).

What effect would an unticipated, 6 percent rise in unit costs have on your prices

Likewise, of the panelists asked to consider a 6 percent cost increase, only 37 percent said they would pass most of the increase on to customers last October, compared with 60 percent who said they would do so today (see chart above).

What's behind the reported increase in firms' pricing power since October? Well, we'd be speculating on that point, but one thing that's changed since October is that a smaller proportion of firms are reporting sales levels “less than normal” (54 percent compared with 64 percent in October). Said another way, firms, in the aggregate are reporting less slack today than they were ten months ago—a finding that aligns with recent results from the NFIB small business survey showing a decrease in the percentage of firms that consider poor sales to be the primary issue they face.

Percentage of firms reporting sales less than normal

Comparisons across industry groups also suggest that slack may be influencing firms' reported pricing power. Retailers, for example, say they would pass through about 75 percent of a cost increase to their customers, compared with only 64 percent for manufacturers and 60 percent for other firms. At the same time, retailers are reporting that sales and price margins are closer to normal—a sign that their perception of “slack” is not as great as other firms.

Industry breakdown of the pass-through from the hypothetical...


Does the analysis above indicate that firms are running out of slack? Sorry, but we can't quite go that far on the basis of these two surveys. But one thing seems clear—while our panel of businesses says sales levels are still below normal, sales and pricing power are better today than they were last October.

Mike BryanBy Mike Bryan, vice president and senior economist;

Laurel GraefeLaurel Graefe, economic policy analysis specialist;

Nicholas ParkerNicholas Parker, economic research analyst; and

Kate ReesKate Rees, economic research analyst, all with the Atlanta Fed

August 23, 2012 in Business Inflation Expectations, Inflation | Permalink


TrackBack URL for this entry:

Listed below are links to blogs that reference Do Firms Have Pricing Power?:


Post a comment

Comments are moderated and will not appear until the moderator has approved them.

If you have a TypeKey or TypePad account, please Sign in

August 22, 2012

The (Unfortunately?) Consistent Record of the Recovery: Duy and Thoma Respond

Mark Thoma, always generous in linking and reposting our musings here at macroblog, took a look at my last post and read a sense of helpless resignation:

David Altig of the Atlanta Fed argues that "the majority of FOMC participants don't seem to think that the unemployment rate will improve that quickly, but "it is not at all obvious that the pace of the recovery is inconsistent with the FOMC's view of achieving its dual mandate." It sounds as though the Fed has given up -- we've done all that we can, there's nothing more we can do, so we won't even try -- and we're not about to risk even the tiniest bit of inflation to find out if we are wrong (and this is despite assurances from Bernanke and others that the Fed is not out of bullets)...

Tim Duy, whose observations in fact motivated my post, had his own response to my comments:

Altig's calculations make the important assumption that the labor force participation rate holds at 63.7%. This effectively assumes that none of the decline in the labor force participation is cyclical. Instead, it is all structural...

There are really two separate thoughts in these comments. So let me take them in turn, but first recap what I said in the previous post (or at least what I meant to say):

  • Stepping back and looking at the data, I am drawn to the conclusion that the U.S. economy looks like it has settled into a pattern of something like 2 percent GDP growth with net job creation somewhere around 150,000 payroll jobs per month.

  • The unemployment projections published in the FOMC participants.' June Summary of Economic Projections (SEP) would, under certain assumptions, be consistent with annual job growth averaging the 150,000 per month pace we have seen over the past year and a half.

The under certain assumptions caveat is obviously important. To Duy's point, my mapping of the apparent employment trend to the SEP unemployment forecasts does assume a constant labor force participation rate. Multiple macroblog posts this year have offered skepticism about exactly that assumption (here and here, for example), and any rise in the labor force participation rate will require faster job growth to get the same unemployment rate outcomes. But as far as I know—I think as far as anyone knows—participation will rise only if we get that faster job growth in the first place. My only point was that the SEP unemployment rate submissions in June are not obviously out of line with what appears to be the current trend in job creation.

In fact, I cannot tell you what assumptions underlie the unemployment rate (and growth) projections in the SEP. I can tell you only that these are the outcomes the individual participants view as consistent with "appropriate monetary policy." And that brings us to Mark Thoma's concern that the very slow progress toward higher growth and lower unemployment in the SEP implies that the Fed has "given up, "done all that we can," and "won't even try."

I definitively do not want to leave an impression that this view is implied by the SEP. As I noted in my original post (and duly noted in both the Thoma and Duy responses), the definitions of appropriate monetary policy that condition the individual SEP contributions are not spelled out. Lacking that information, one should not infer that monetary policy is assumed by any one individual as fixed or without influence.

Could it possibly be that an unemployment rate at 7.5 percent and GDP growth of 2.8 percent in 2013 (the more optimistic forecasts in the majority, or the "central tendency" range in the June SEP) are consistent with monetary policy having a nontrivial positive impact on the economy?

Of course monetary policy does not operate in a vacuum. As our boss, Atlanta Fed President Dennis Lockhart, said in a speech yesterday:

Monetary policy can exert a powerful positive influence on an economy, but as Chairman Bernanke has pointed out, monetary policy is not a panacea.

I'm not really aware of any models matched to real-world data that suggest monetary policy actions can (at acceptable cost) quickly and completely overcome all of the shocks and headwinds that may present themselves.

You may believe otherwise—that is, you may believe that, for current circumstances, monetary policy is a panacea. Or, less dramatically, you may believe that more monetary stimulus would surely yield something better than what was implied in the June SEP. Fair enough. But you should not believe that lackluster numbers in the SEP tell you anything about individual FOMC participant's views on the efficacy, desirability, or likelihood of further monetary actions, one way or the other.

David AltigBy Dave Altig, executive vice president and research director at the Atlanta Fed

August 22, 2012 in Federal Reserve and Monetary Policy, Monetary Policy | Permalink


TrackBack URL for this entry:

Listed below are links to blogs that reference The (Unfortunately?) Consistent Record of the Recovery: Duy and Thoma Respond:


Tim Duy has done a great job of getting himself noticed as a Fed observer. However, as is commonly the case in the pundit game, he is not being judged by the quality of his views, but seems instead to have done what is necessary to get noticed, and doesn't have to earn his grades.

Duy has sunk into the habit of personalizing his monetary policy commentary, featuring his own likes and dislikes as if his preferences are the important issue. He has also personalized his commentary in another way - basing his views of monetary policy on an implicit ability to see deeply into Bernanke's motivations.

Yesterday's FOMC minutes suggest Duy's recent analysis has been wrong. He has been big enough to admit that, to his credit. However, being wrong and getting to the wrong answer by way of bad analytics ought to be reason to give Duy's views less credence.

Of course, I've been banned from Thoma's place for pointing out that his expertise is in economics, not politics or psychology. The ban coincides with Thoma's arrival as a paid pundit, in which position he wanders into politics and psychology pretty frequently. Were that not the case, I'd make my point about Duy's critical abilities there rather than here...

Posted by: kharris | August 23, 2012 at 08:21 AM

Really liked the last part about not inferring policy views from the SEP figures. I agree that the current recovery can be consistent with the Fed's mandate and monetary policy, to date. (http://bubblesandbusts.blogspot.com/2012/08/fomc-projections-provide-no-hint-of.html)

Posted by: Joshua Wojnilower | August 23, 2012 at 11:36 AM

Post a comment

Comments are moderated and will not appear until the moderator has approved them.

If you have a TypeKey or TypePad account, please Sign in

August 17, 2012

The (Unfortunately?) Consistent Record of the Recovery

In his last two posts (here and here), economist Tim Duy has done some yeoman work displaying and discussing the economic context of monetary policy decisions past and prospective. Though Wednesday's self-titled post "Data Dump" focuses on the incoming data as a set-up to the next meeting of the Federal Open Market Committee (FOMC), what strikes me is the consistency of the broad macroeconomic outcomes over the course of the recovery. Gross domestic product (GDP) growth has pretty clearly clocked in at about 2 percent...

...and, looking through the quarterly ups and downs, payroll employment growth has clearly trended near 150,000 jobs per month after a slower start in 2010:

The inflation picture shows more variation...

...but in my view, that sort of variation is why it makes sense to think in terms of medium-term performance. "Medium-term" is more a measure of art than science, and I would concede the point that the recovery as a whole would be on the shorter end of that time frame. Suffice it to say that the pace of price-level growth over the past two and a half years wouldn't contradict the presumption that inflation is pretty close to the FOMC's stated longer-run objective.

Duy looks at this performance and sees pretty clear evidence of failure:

The economy continues to settle into a path that is not consistent with either part of the Fed's dual mandate. Moreover, there are very real downside risks to even a tepid outlook...

This is frustrating. What in the world is the point of making a big claim to affirm the nature of the dual mandate and then subsequently ignore any forecasts that indicate you have no faith the elements of the dual mandate will be met anytime soon?

That complaint is not really about the inflation part of the mandate, but the employment/growth part of it. But if you are willing to accept that employment growth remains on a pace of 150,000 jobs per month—and I see no clear evidence to the contrary—it is not at all obvious that the pace of the recovery is inconsistent with the FOMC's view of achieving its dual mandate. Here, for example, are the central tendency ranges of the unemployment rate projections from the FOMC's June Summary of Economic Projections (SEP) and the employment growth that would be required to meet those objectives (with some important assumptions, such as the labor force participation rate remaining at the current level).

Here is the important statement of conditionality, as described in the SEP document:

The charts show actual values and projections for three economic variables [GDP growth, the unemployment rate, and PCE inflation] based on FOMC participants' individual assessments of appropriate monetary policy.

Under appropriate policy—which pretty clearly means mandate-consistent outcomes—the majority of FOMC participants don't seem to think that the unemployment rate will improve that quickly. And, to my point, it is not clear that the trend in payroll employment is inconsistent with that pace of improvement.

Of course, individual contributors to the SEP may have different assumptions about things like the labor force participation rate. More importantly, the SEP is silent on what, in each contributor's view, constitutes "appropriate policy."

And I am certainly begging the important issues. Would the economy have achieved even the somewhat unspectacular pace of 2 percent GDP growth, 150,000 jobs per month, and average inflation near the long-run objective absent large-scale asset purchases ("QE2"), forward guidance (statements indicating that policy rates are expected to be exceptionally low through at least late 2014), and maturity extension programs ("Operation Twist")? Does "appropriate policy" imply that more must be done to achieve even the modest progress in the unemployment rate implied in my calculations above? And could we have (looking backward) or can we (looking forward) do even better with an even more aggressive approach, as many Fed critics argue?

Good questions, those.

David AltigBy Dave Altig, executive vice president and research director at the Atlanta Fed


August 17, 2012 in Employment, GDP, Inflation | Permalink


TrackBack URL for this entry:

Listed below are links to blogs that reference The (Unfortunately?) Consistent Record of the Recovery:


Post a comment

Comments are moderated and will not appear until the moderator has approved them.

If you have a TypeKey or TypePad account, please Sign in

August 10, 2012

Deflation Probabilities on Our Radar Screen

In the latest Wall Street Journal Economic Forecasting Survey, conducted August 3–6, economists were asked whether they "expect[ed] the Fed to start another round of large-scale bond buying in 2012?" Sixty-three percent answered yes, and 49 percent expected a program would be announced in September, presumably at the end of the next meeting of the Federal Open Market Committee (FOMC) on September 12–13. Obviously this question is of interest to more than just business economists. For example, at his July 17 testimony before the Senate Committee on Banking, Housing, and Urban Affairs, Sen. Mike Crapo asked Fed Chairman Ben Bernanke whether the FOMC should seriously consider more quantitative easing going forward. As part of his response, the Chairman said that "we would certainly want to react against any increase in deflation risk." The entire video exchange can be viewed at the 52–55 minute mark here.

As part of the Atlanta Fed's Inflation Project, we regularly update probabilities of deflation in the Consumer Price Index (CPI) estimated from Treasury Inflation-Protected Securities (TIPS) prices, described here and here. The basic idea is that a recently issued 5-year TIPS has less "deflation protection" than a 10-year TIPS maturing about the same date as the 5-year TIPS. The yield spread between the 5-year TIPS and 10-year TIPS can be used to help estimate the probability of deflation.

The most recent (August 8) estimate puts the 5-year probability of deflation from early 2012 to early 2017 at around 15 percent. As seen in the figure below this probability is up slightly from May, but only about half the readings of the 5-year (2010–15) deflation probability seen in the late summer and early fall of 2010 and considerably below readings seen during the height of the financial crisis in late 2008 and early 2009.

It is important to note that these deflation probabilities are estimates based on a relatively simple model that uses a number of assumptions that not everyone may agree on. Jens Christensen, Jose Lopez, and Glenn Rudebusch at the Federal Reserve Bank of San Francisco have built an alternative model for estimating deflation probabilities that also uses TIPS yields. At the time of the publication of their paper in 2011, their model's probabilities were somewhat lower than—but highly correlated with—ours.

The TIPS market has a number of features that make inferring both inflation expectations and deflation probabilities from them tricky. Most notably, there are unknown liquidity differences between TIPS and nominal Treasury securities. A more direct way of estimating deflation probabilities—or in fact the entire probability distribution of future CPI inflation—using so-called inflation caps and floors has recently been explored by economists Yuriy Kitsul and Jonathan Wright. Inflation caps and floors are essentially options on the Consumer Price Index. We used the Kitsul and Wright method for constructing the implied probabilities of (annualized) CPI inflation over the next five years. As seen in the figure below, this method implies a 13 percent probability that inflation will be 0 percent or negative on average over the next five years. This probability is about at the midpoint of the range that prevailed between October 2009 and March 2012. (See figure 3, on page 29 of Kitsul and Wright's working paper.)

As Kitsul and Wright explain, the market for inflation caps and floors is still quite small relative to the TIPS market. So the deflation probabilities from their model should be considered suggestive, as should our own.

What's the takeaway from all this? Well, readings from the financial market indicate the likelihood of a sustained deflation is currently about 15 percent, or a bit less. That's up from earlier in the year, but not nearly as high as in 2010.

Should we be concerned about the prospect for deflation in the years ahead? This is obviously an important policy question. But I'm not a policymaker; I merely put up the numbers for you to consider. And, of course, we will continue to follow these indicators closely—as can you. Our deflation probability estimates are updated every Thursday and posted on our Inflation Project.

Partick Higgins By Patrick Higgins, an economist in the Atlanta Fed's research department

August 10, 2012 in Deflation, Inflation | Permalink


TrackBack URL for this entry:

Listed below are links to blogs that reference Deflation Probabilities on Our Radar Screen:



What relevance would you give to market demand for government securities? I have generally equated high demand with hedging against perceived deflation risk, even though the data suggest the two are actually diverging. Can bond market signalling be used reliably to determine probability of phase change from inflationary to deflationary environment or vice versa?

Posted by: Ben Johannson | August 11, 2012 at 01:59 AM

Sustained deflation seems unlikely. A far more likely explanation/interpretation (that btw agrees with 1 and 2-yr breakeven-implied TIPS inflation over the June/July time period) is that there is a significant chance that one year of the next five will see deflation. In other words: be very afraid.

Posted by: dwb | August 11, 2012 at 01:32 PM

How could one go about assessing the same calculation using PPI vs. CPI? I am curious if there are any notable differences between the underlying index for inflation as it could potentially have implications for profit margins.

Thanks for your time and efforts.

Posted by: Danny | August 15, 2012 at 10:43 AM

Post a comment

Comments are moderated and will not appear until the moderator has approved them.

If you have a TypeKey or TypePad account, please Sign in

Google Search

Recent Posts



Powered by TypePad