Archive

Archive for the ‘Economy’ Category

Housing Disinflation Isn’t Happening Yet

June 19, 2017 8 comments

Before everyone gets too animated about the decline in core inflation, with calls for central banks to put the brakes on rate normalization, let’s realize that the main drivers of lower inflation over the last few months – zero rise in core CPI over three months! – are not sustainable. I’ve written previously about the telecommunications-inflation glitch that is a one-off effect. Wireless telephone services fell -1.38% month-over-month in February (not seasonally adjusted), -6.94% in March, and -1.73% in April. In May, the decline was -0.06%. Here is a chart, courtesy of Bloomberg, showing the year-to-date percentage declines for the last decade. The three lines at top show the high, average, and low change over the prior decade, so you can see the general deflationary trend in wireless telecom services and the historical outliers in both directions. The orange line is the year-to-date percentage change. Again, the point here is that we cannot expect this component of inflation to deliver a similar drag in the future.

The other main drag comes from a less-dramatic decline in a much-larger component: Owners’ Equivalent Rent. In this month’s CPI tweetstorm, I pointed out that this decline is mostly just returning the OER trend to something closer to our model (see chart below), but many observers (who don’t have such a model) have seen this as a precursor to a more-significant decline in rents.

This is actually a much more-important question than the dramatic, and easy-to-diagnose, issue of wireless telecommunications, because OER is a ponderous category. You can’t get high inflation without OER rising, and you can’t get deflation or even significant disinflation without OER declining. It’s just too big. So what are the prospects for OER rolling over?

Here are two reasons that I think it’s very unlikely that this is a precursor to a significant decline in housing inflation.

First, while I understand that rent increases in some parts of the country are moderating, they are always moderating somewhere in the country. Owners’ Equivalent Rent tends to parallel primary rents (“Rent of Primary Residence,” which measures the actual price of a rental unit as opposed to implied rent of an owner-occupied dwelling) reasonably well, and when home prices are rising it tends to imply that rents – as the price of a substitute, at least for the consumption part of home prices – are also rising. (A house is both an investment asset and a consumption good, and the BLS’s method for separating these two components of a home recognizes that the consumption component should look a lot like the substitute). And the fact is that Primary Rents are not (yet?) decelerating much (see chart, source Bloomberg).

Yes, I understand and agree that home prices are already too high to be sustainable in the long run. Either incomes need to outpace home prices for a while, or home prices need to decline again, or we need to become accustomed to housing becoming a permanently larger part of our consumption and asset mix (see chart, source Enduring Investments).

But is that going to happen? Well, here are two charts that should make you somewhat skeptical that at least on the supply side we are about to see a decline in home prices. First, here is the index of Housing Starts, which last month took a nasty drop. Even without the nasty drop, though, notice that the level of starts was not only far below the level of the last few peaks in the housing market, but actually not far above the troughs reached in the recessions of the mid-1970s, early 1980s, and early 1990s. The only reason the current level of starts looks high is because homebuilders basically stopped building for a few years after the housing bubble.

Homebuilders stopped building because there was suddenly plenty of inventory on the market! In the immediate aftermath of the bubble, the homes that were available for sale were often distressed sellers and as prices rose, more and more of the so-called “shadow inventory” (people who wanted to sell, but were now underwater and couldn’t sell) was freed. This kept a lid on overall housing starts, but the net effect is that even now, when most of that shadow inventory has presumably been liquidated (a decade after the bubble and at new price highs), the inventory of existing homes available for sale has become and has remained quite low (see chart, source Bloomberg).

The supply side, then, doesn’t seem to offer much cause to expect home prices to moderate, even if their prices are relatively high. I’d want to see an overreaction of builders, adding to supply, before I’d worry too much about another bust, and we haven’t seen that yet. So we have to turn to the demand side if we expect home prices to decline. On that side of the coin, there are two arguments I sometimes hear: 1) household formation in the era of the Millennial is low, or 2) households don’t buy as much housing as they used to.

There is no evidence that household formation has slowed in recent years. As the chart below (source Bloomberg) shows, household formation has been rising since 2009 or so, and is back in line with long-term trends. Millennials may have weird notions of home life (I don’t judge!), but they still form households of their own.

As for the second point there…notice that I phrased the question as whether Millennials are buying less housing, rather than as buying fewer homes. I think it’s plausible to suggest that Millennials might demand fewer homes to buy, but it’s hard to imagine that they’re neither going to rent nor buy homes – and if they do either, they are demanding shelter as a consumption item. It just becomes a question of whether they’re demanding rental housing or owned housing.

The upshot of this is that there’s no sign yet of a true ebbing in housing/rental inflation. And until there is, there’s scant need to fear a disinflationary trend taking hold.

Summary of My Post-CPI Tweets (June 2017)

June 14, 2017 1 comment

Special request: if you get value from these pieces, please take a moment to respond to my survey/market research – it will take about three minutes – and forward to everyone you know! I am looking for 1,000 responses. Please help, and remember that these posts are all free to you!

Below is a summary of my post-CPI tweets. You can (and should!) follow me @inflation_guy or sign up for email updates to my occasional articles here. Investors with interests in this area be sure to stop by Enduring Investments. Plus…buy my book about money and inflation. The title of the book is What’s Wrong with Money? The Biggest Bubble of All; order from Amazon here.

  • CPI day! People looking past CPI at 8:30 but…not me!
  • Last 2 CPI prints were very low. The first was a 1-off wireless telecom debacle, read about that effect here.
  • Last month’s CPI weakness was in core services – in medical care & rent of shelter. Harder to ignore but unlikely to be in freefall.
  • Consensus core CPI is for another weak print, only 0.16% or so. Economists believe disinflation is upon us. I think that’s premature.
  • Last May’s core CPI was 0.21%, so that’s the hurdle to get acceleration in y/y figures.
  • WOW! At this rate I will have to change my Twitter handle. Each month is more shocking. m/m core 0.1%, not sure on the rounding yet.
  • 06% m/m on core CPI, so again incredibly weak. y/y at 1.74%, producing the scary optic of a drop from 1.9% to 1.7% on the rounded core
  • This is an amazing chart.

  • waiting for the data dump, but housing, medical care, apparel subcomponents all decelerated.
  • So the upshot is…core prices overall are unchanged from February. That’s right, 0% core inflation over 3 months.
  • Yes, it was telecom that made 0% possible and that won’t be repeated. But still striking. Here is the index itself.

  • So Dec, Jan, Feb core inflation is rising at a 3% annualized pace. next 3 months, zero. That’s not supposed to happen to core.
  • Breakdown now. In Housing, Primary rents remain solid at 3.85% y/y, unch. But Owners’ Equiv plunged (for it) to 3.25% from 3.39%.
  • Picture of OER: this is a dramatic shift in this index, and frankly hard to explain given home price increases.

  • Medical Care decelerated to 2.66% from 2.95%. But w/in MC, drugs rose to 3.34% vs 2.62%. Professional svcs flopped to 1.00% from 1.58%
  • CPI/Med Care/Professional Services, y/y. Doctors suddenly don’t need to be paid.

  • Apparel had been at 0.45% y/y, fell to -0.94%.
  • The Fed funds rate is too low and almost certainly rises today. But with a sudden zig in CPI…it wouldn’t SHOCK me if they delayed.
  • Back to housing – we’ve believed OER was ahead of itself for awhile. Adjustment is just really sudden.

  • in the biggest-pieces breakdown, core goods is at -0.8% y/y while core services is down to 2.6%.
  • US$’s recent decline (2y change in trade-weighted $ is only +7%) means core goods are losing the downward pressure of last few yrs.
  • But the dollar’s effect is lagged significantly. We’re still seeing effect of prior strength.

  • Here are the four pieces of CPI, most volatile to least. Starting with Food & Energy (21% of CPI)

  • Core goods (33%)

  • Core services less Rent of Shelter. Yipe!

  • Got my percentages wrong. Food & Energy is 21%. Core goods is 19%, core services less ROS is 27%. Rent of Shelter is 33%.
  • Rent of Shelter. 27% of overall CPI. I still find it hard to believe this is going to collapse, but as I tweeted earlier it was ahead.

  • My early estimate of Median CPI is 0.18% m/m, 2.28% y/y down from 2.37%.
  • One thing to keep in mind is that in June and July we drop off 0.15% and 0.13% from y/y core. So core should bounce back some. (??)
  • I mean, we can’t average 0% core going forward, right?!? Otherwise @TheStalwart and @adsteel will never have me on again.
  • core ex-shelter down to 0.59% y/y. Lowest since JANUARY 2004!

  • Interestingly, the weight of categories inflating more than 3% remains high. The pullback is in the far left tail.

Well, it’s getting harder to put lipstick on this pig. The telecom-induced drop of a couple of months ago was clearly a one-off. But the slowdown in owners’-equivalent rents is merely putting it back in line with our model, and so it’s hard to believe that’s going to be reversed. And I’m really, really skeptical that there has been an abrupt collapse in the rate of increase of doctors’ wages.

Except, what if there is a shift happening from higher-priced doctors to lower-priced doctors? This sort of compositional shift happens all the time in the data and it’s devilishly hard to tease out – for example, in the Existing Home Sales report it is sometimes difficult to tell if a change in home prices is coming from a broad change in home prices, or because more high-priced or low-priced homes are being sold this month, skewing the average. So this kind of composition shift is possible, in which case each individual doctor could see his wages increasing while the average declines due to the composition effect. I have no idea if this is what is happening – I’m just making the point that if it is, then this effect could be more persistent and not the one-off that the telecom change was. However, I am skeptical.

I do not believe that we have seen a turn in the inflation cycle. With money growth persistently above 6%, it would take a further collapse in money velocity from already-record-low levels to get that to happen. Forget about the micro question, about whether movements in this index or that index look like they’re rolling over. The macro question is that it is hard to get disinflation if there’s too much money sloshing around, whether or not the economy is growing.

But that being said, the Fed doesn’t necessarily believe that. There is a tendency to believe one’s own fable, and the fable the FOMC tells itself is that raising interest rates causes growth to slow and inflation to decline. Although the effect is spurious, we are currently seeing somewhat slower growth (for example, in the recent slowing of payrolls) and we are seeing lower core inflation. It is a low hurdle for the Fed to believe that their policy moves are an important part of the cause of these effects. Of course, they’re not – the tiny changes the FOMC has made in the overnight rate, even if it had been propagated to significant changes in longer rates – which it hasn’t been – or resulted in slower month growth – it hasn’t, especially if you look globally – would not have had much effect at all. But that won’t stop them from thinking so. Ergo, the chance that the Fed skips today’s meeting, while small, are non-zero. And there is a much greater chance that the “dot plot” shifts lower as dovish members of the Fed (and that’s most of them) back away from the feeble pace of increases they’d been anticipating.

Summary of My Post-CPI Tweets (May 2017)

May 12, 2017 4 comments

Below is a summary of my post-CPI tweets. You can (and should!) follow me @inflation_guy or sign up for email updates to my occasional articles here. Investors with interests in this area be sure to stop by Enduring Investments. Plus…buy my book about money and inflation. The title of the book is What’s Wrong with Money? The Biggest Bubble of All; order from Amazon here.

This month, I am making sure to include my comments before the actual number, since my suspicions about the upside risk were exquisitely wrong. This is why you shouldn’t put a lot of weight on monthly figures, folks!

  • Step right up ladies and gents. The CPI circus is about to commence.
  • Last month’s circus crazier than usual, including an unprecedented (and inexplicable) 11+% drop in wireless telecom services. [Editor’s note: it was only 7%. I corrected this in a later tweet]
  • This caused more diversion in core and median CPI. Median (better measure) remains steady at 2.5%.

  • PPI y’day was broadly strong. I don’t pay much attention to PPI but it does create upside risk.
  • Also note that European inflation saw a drop and then big jump from the early Easter. Not sure we have an analog but…
  • Point is that consensus is for 0.17% or so. There’s a lot of upside risk to that number I think.
  • Over next few months, core will rise regardless as we drop off 0.18, 0.21, 0.15, and 0.13. Easy hurdles.
  • Wow! Core only 0.1% again! Even a low 0.1%…0.07%.
  • I cannot WAIT to get a look at the breakdown.
  • ..Medical Care ebbed from 3.5% y/y to 3.0% y/y, wanna look inside that one. Recreation and yes, communication also soft.
  • Core drops to 1.89% y/y. Lowest since late 2015. Of course, remember that median is a better measure – we’ll see that later.
  • [I retweeted this, look at Matthew’s yellow line here]

  • Wireless telecom services fell another 1.7%. Incidentally I earlier said 11% m/m was last mo…it was only 7% m/m, the 11% decline was y/y.
  • so wireless telecom services now down 12.9% y/y, 9.9% over the last 3 months. This really warrants explanation from BLS.
  • In Medical Care, Medical Drugs fell to 2.62% from 3.97% y/y. Professional svcs, which is twice the weight, fell to 1.58% from 2.50%.
  • Health insurance fell to 2.72% vs 3.34%. Lowest since 2015.
  • Medical decel seems implausible but remember is a rate of change measure. So rising from high level, but at slower rate, is lower CPI.
  • Let’s get to housing. Primary rents 3.84% vs 3.88%. OER fell to 3.39% from 3.49%, that’s a big drop for 25% of the index.
  • So overall, Housing rose from 3.1% to 3.2%, but that’s on the strength of a 1% rise in household energy y/y.
  • This is OER. The decline is actually welcome – it had been running well ahead of even our optimistic models.

  • Core goods steady at -0.6% y/y. So the deceleration in last two months is all from core services, from 3.1% to 2.9% to 2.7%.
  • I don’t see that slowdown in core services as sustainable unless housing rolls over…
  • …and I don’t see that happening while home prices keep rising at 6-7% as they have been.
  • Weakness in services outside of housing s/b taken with grain of salt though…a lot of that is wireless services!
  • But doing core-less-housing-and-wireless is cheating. We take out housing to look @ the wiggly stuff. Can’t also take out wiggly stuff.
  • OK, four-pieces CPI look. From most-wiggly to least. They tell the story. Food & Energy:

  • Core goods (about 19% of CPI)

  • Core services less rent-of-shelter (26% of CPI). <<BOOM>>

  • And Rent of Shelter (33.3%)

  • And within core services less ROS, a lot of that is wireless but medical care ebbing is also in there. That’s the story of this month.
  • On Median…I have 0.13% m/m, but the median category is an OER piece and the BLS seasonally adjusts those.
  • But my best guess on median is 0.13%, dropping y/y to 2.4%.
  • Maybe I’m wrong and inflation pressures are ebbing after all. You know who else is thinking that? Janet Yellen.
  • Forgot to tweet this chart earlier.

  • Also interesting. Core<median b/c of big weight in left tail. But also starting to be more weight in general left of mode.

  • Last routine chart: the weight of categories inflating faster than 3% is still almost half. It’s that left tail draggin’ stuff down.

It was easy to ignore last month’s negative core print. It was obviously tied to a ridiculous (and still not explained by the BLS) plummet in the price of wireless telecommunications services. A 7% fall, nationwide, in one month, that no one seems to have noticed, is something the BLS really needs to comment on (my best guess is that some data plans got uncapped, and the BLS assumed a large increase in the data taken at zero dollars and therefore a big drop in the price per gig. That’s effectively a hedonic adjustment, and a not unreasonable one if you really saw a dramatic increase in data being taken. Since I have yet to talk to anyone who saw anything that resembled this huge effect, I remain skeptical.) But in any event, it was easy to ignore March’s number released in April.

Now we have two months in a row, and while wireless telecom contributed this month as well, there was also softness in medical care and in owner’s equivalent rent. That’s harder to ignore. And while median CPI was steady after last month’s debacle, it should downtick today.

I don’t think inflation is done rising; I think this is just a pause. But as I said above, I am sure that the decline in core CPI and core PCE will not go unremarked at the next FOMC meeting – the one where they are supposed to hike rates again. I think we’ll learn a lot about the stomach the Fed has for continuing the rate normalization regime by whether they go through with the next hike.

Categories: CPI, Tweet Summary

Tariffs are Good for Inflation

The news of the day today – at least, from the standpoint of someone interested in inflation and inflation markets – was President Trump’s announcement of a new tariff on Canadian lumber. The new tariff, which is a response to Canada’s “alleged” subsidization of sales of lumber to the US (“alleged,” even though it is common knowledge that this occurs and has occurred for many years), ran from 3% to 24% on specific companies where the US had information on the precise subsidy they were receiving, and 20% on other Canadian lumber companies.

In related news, lumber is an important input to homebuilding. Several home price indicators were out today: the FHA House Price Index for new purchases was up 6.43% y/y, the highest level in a while (see chart, source Bloomberg).

The Case-Shiller home price index, which is a better index than the FHA index, showed the same thing (see chart, source Bloomberg). The first bump in home price growth, in 2012 and 2013, was due to a rebound to the sharp drop in home prices during the credit crisis. But this latest turn higher cannot be due to the same factor, since home prices have nearly regained all the ground that they lost in 2007-2012.

Those price increases are in the prices of existing homes, of course, but I wanted to illustrate that, even without new increases in materials costs, housing costs were continuing to rise faster than incomes and faster than prices generally. But now, the price of new homes will also rise due to this tariff (unless the market is slack and so builders have to absorb the cost increase, which seems unlikely to happen). Thus, any ebbing in core inflation that we may have been expecting as home price inflation leveled off may be delayed somewhat longer.

But the tariff hike is symptomatic of a policy that provokes deeper concern among market participants. As I’ve pointed out previously, de-globalization (aka protectionism) is a significant threat to inflation not just in the United States, but around the world. While I am not worried that most of Trump’s proposals would result in a “reflationary trade” due to strong growth – I am not convinced we have solved the demographic and productivity challenges that keep growth from being strong by prior standards, and anyway growth doesn’t cause inflation – I am very concerned that arresting globalization will. This isn’t all Trump’s fault; he is also a symptom of a sense among workers around the world that globalization may have gone too far, and with no one around who can eloquently extol the virtues of free trade, tensions were likely to rise no matter who occupied the White House. But he is certainly accelerating the process.

Not only do inflation markets understand this, it is right now one of the most-significant things affecting levels in inflation markets. Consider the chart below, which compares 10-year breakeven inflation (the difference between 10-year Treasuries and 10-year TIPS) to the frequency of “Border Adjustment Tax” as a search term in news stories on Google.

The market clearly anticipated the Trumpflation issue, but as the concern about BAT declined so did breakevens. Until today, when 10-year breakevens jumped 5-6bps on the Canadian tariff story.

At roughly 2%, breakevens appear to be discounting an expectation that the Fed will fail to achieve its price inflation target of 2% on PCE (which would be about 2.25% on CPI), and also excluding the value of any “tail outcomes” from protectionist battles. When growth flags, I expect breakevens will as well – and they are of course not as cheap as they were last year (by some 60-70bps). But from a purely clinical perspective, it is still hard to see how TIPS can be perceived as terribly rich here, at least relative to nominal Treasury bonds.

Summary of My Post-CPI Tweets

Below is a summary of my post-CPI tweets. You can (and should!) follow me @inflation_guy or sign up for email updates to my occasional articles here. Investors with interests in this area be sure to stop by Enduring Investments. Plus…buy my book about money and inflation, published in March 2016. The title of the book is What’s Wrong with Money? The Biggest Bubble of All; order from Amazon here.

  • Pretty big market day! Importantly, CPI: remember last month was big upside surprise, and driven by unusual suspects – core goods.
  • There’s a decent base effect hurdle today, as last Feb was 0.25% on core CPI. Consensus today is for a very weak 0.2% (almost 0.1%).
  • The consensus forecast clearly says that most economists see last month’s shocking 0.31% on core as one-offs.
  • Consensus expectation is for core to slip back to 2.2% from 2.3%. But then, last month they thought we’d fall to 2.1%.
  • Hurdles get easier next month: March ’16 saw 0.09% core CPI, and then a series of low 0.2s & 0.1s. So core is going up this summer.
  • Here is what I said about last month’s figures: https://mikeashton.wordpress.com/2017/02/15/summary-of-my-post-cpi-tweets-36/
  • 5 mins to CPI. Sources say the headline number is trading 243.34 (which would be -0.04% on headline) in the CPI derivs mkt.
  • core at 0.21%, higher than consensus expectations of 0.15% or so. Keeps y/y at 2.22%, down from 2.26%. But next month is an easy comp.
  • Monthly core CPI prints.

  • I don’t pay much attention to headline but it was a little high, y/y up to 2.74%. Only matters if it affects tenor of Fed discussion.
  • In major subgroups: Housing rose to 3.18% vs 3.12%. Need to see if that’s energy. Apparel fell back, as did health care.
  • w/in housing, Primary Rents slipped to 3.91% from 3.93%, Owner’s Equiv to 3.53% from 3.54%. So the housing bump was elsewhere.
  • Looks like the housing increase was mostly household energy, 4.46% from 3.51%. So no biggie as the kids say.
  • Apparel 0.42% vs 0.99%. The big jump last month was mostly reversed. Overall core services 3.1% and core goods dropped back to -0.5%
  • Last month the big story was that core goods had caused the jump in core CPI. Looks like these were mostly seasonal issues after all.
  • Transportation 6.3% vs 4.8%. That’s mostly gasoline. New & used cars slipped. But rising: parts, maintenance, insurance, airfares.
  • In Medical Care, big drop in medicinal drugs 4.19% vs 4.85%. Also drop in prof svcs (2.68% v 2.94%). THOSE are the one-offs this month.
  • Here are y/y med care & housing, source of the big upward pressure recently. But remember this month the housing is mostly energy.

  • Four more major subcategories. Recreation is the only one moving higher, but it’s a heterogeneous group & hard to decipher.

  • Quick estimate of Median is 0.21% m/m, 2.52% y/y, not quite a new high. Official figure will be out later.
  • Next month we should have core back over 2.3% and a shot at 2.4%, thanks to easy comp in March.
  • 10y inflation swaps still below current median inflation.

  • Mkt pretty confident in Fed: CPI mkt pricing: 2017 2.0%;2018 2.2%;then 2.2%, 2.1%, 2.2%, 2.2%, 2.3%, 2.4%, 2.5%, 2.7%, & 2027:2.5%.
  • This CPI report takes inflation off the boil, but not off the burner.
  • One more chart: weight of CPI categories over 3% inflation y/y.

Let’s face it. While this month’s CPI held some intrigue because of last month’s surprising spike, nothing about the figure was likely to change the outcome of today’s FOMC meeting and probably not the tenor of the statement or post-meeting presser. So, in that sense, this was a much less-significant report than last month’s release.

At the same time…let’s not lose sight of the fact that this was still an above-consensus CPI report. While the consensus was broadly correct that some of the jumps in core goods categories from last month were one-offs, and at least partially retraced this month, it’s still the case that y/y core inflation is going to keep rising through the summer merely on base effects. If the Fed wants to be hawkish and tighten more than the market currently expects (I think that nothing could be further from the truth, with Yellen at the helm, but she seems to dislike President Trump enough that she might forget some of her dovish leanings), then they will continue to have cover from inflation reports for a while.

Going forward from that, there are two inflation questions that will be resolved: (1) Will core goods recover and rise, indicating a broadening of inflation impulses that could produce a longer-tail upside? And (2) will housing inflation flatten out or decline since rent inflation is currently rising faster than even our most-generous models? If it does, then core inflation might stabilize near the current level, or even decline.

I have trouble figuring out what the mechanism would be for inflation to flatten out at these levels, from the macro-monetary perspective. Money growth remains brisk and higher interest rates should eventually goose velocity. I don’t see much prospect of money growth rolling over while banks are neither capital- nor reserve- constrained. And it’s hard to see interest rates heading back down while central banks shift into less-accommodative stances. I have more confidence in the macro-monetary (“top down”) model at longer time frames, and more confidence in the bottom-up analysis at shorter time frames. And for years they’ve told the same story: inflation should be rising, and it has. But there is a conflict between these perspectives that is coming later this year. How it resolves will be the story of the next 3-6 months.

Good Models and Bad Models

I have recently begun to spend a fair amount of time explaining the difference between a “good model” and a “bad model;” it seemed to me that this was a reasonable topic to put on the blog.

The difference between a good model and a bad model isn’t as obvious as it seems. Many people think that a “good model” is one that makes correct predictions, and a “bad model” is one that makes bad predictions. But that is not the case, and understanding why it isn’t the case is important for economists and econometricians. Frankly, I suspect that many economists can’t articulate the difference between a good model and a bad model…and that’s why we have so many bad models floating around.

The definition is simple. A good model is one which makes good predictions if high-quality inputs are given to the model; a bad model is one in which even the correct inputs doesn’t result in good predictions. At the limit, a model that produces predictions that are insensitive to the quality of the inputs – that is, whose predictions are just as accurate no matter what the inputs are – is pure superstition.

For example, a model of the weather that depends on casting chicken bones and rat entrails is a pretty bad model since the arrangement of such articles is not likely to bear upon the likelihood of rain. On the other hand, a model used to forecast the price of oil in five years as a function of the supply and demand of oil in five years is probably an excellent model, even though it isn’t likely to be accurate because those are difficult inputs to know. One feature of a good model, then, is that the forecaster’s attention should shift to the forecasting of the inputs.

This distinction is relevant to the current state of practical economics because of the enormous difference in the quality of “Keynesian” models (such as the expectations-augmented Phillips curve approach) and of monetarist models. The simplest such monetarist model is shown below. It relates the GDP-adjusted quantity of money to the level of prices.

This chart does not incorporate changes in money velocity (which show up as deviations between the two lines), and yet you can see the quality of the model: if you had known in 1948 the size of the economy in 2008, and the quantity of M2 money there would be in 2008, then you would have had a very accurate prediction of the cumulative rate of inflation over that 60-year period. We can improve further on this model by noting that velocity is not random, but rather is causally related to interest rates. And so we can state the following: if we had known in 2007 that the Fed was going to vastly expand its balance sheet, causing money supply to grow at nearly a 10% rate y/y in mid-2009, but at the same time 5-year interest rates would be forced from 5% to 1.2% in late 2010, then we would have forecast inflation to decline sharply over that period. The chart below shows a forecast of the GDP deflator, based on a simple model of money velocity that was calibrated on 1977-1997 (so that this is all out-of-sample).

That’s a good model. Now, even solid monetarists didn’t forecast that inflation would fall as far as it did – but that’s not a failure of the model but a failure of imagination. In 2007, no one suspected that 5-year interest rates would be scraping 1% before long!

Contrariwise, the E-A-Phillips Curve model has a truly disastrous forecasting history. I wrote an article in 2012 in which I highlighted Goldman Sachs’ massive miss from such a model, and their attempts to resuscitate it. In that article, I quoted these ivory tower economists as saying:

“Economic principles suggest that core inflation is driven by two main factors. First, actual inflation depends on inflation expectations, which might have both a forward-looking and a backward-looking component. Second, inflation depends on the extent of slack (or spare capacity) in the economy. This is most intuitive in the labor market: high unemployment means that many workers are looking for jobs, which in turn tends to weigh on wages and prices. This relationship between inflation, expectations of inflation and slack is called the “Phillips curve.”

You may recognize these two “main factors” as being the two that were thoroughly debunked by the five economists earlier this month, but the article I wrote is worth re-reading because it describes how the economists re-calibrated. Note that the economists were not changing the model inputs, or saying that the forecasted inputs were wrong. The problem was that even with the right inputs, they got the wrong output…and that meant in their minds that the model should be recalibrated.

But that’s the wrong conclusion. It isn’t that a good model gave bad projections; in this case the model is a bad model. Even having the actual data – knowing that the economy had massive slack and there had been sharp declines in inflation expectations – the model completely missed the upturn in inflation that actually happened because that outcome was inconsistent with the model.

It is probably unfair of me to continue to beat on this topic, because the question has been settled. However, I suspect that many economists will continue to resist the conclusion, and will continue to rely on bad, and indeed discredited, models. And that takes the “bad model” issue one step deeper. If the production of bad predictions even given good inputs means the model is bad, then perhaps relying on bad models when better ones are available means the economist is bad?

The Fed Needs More Inflation Nerds

January 30, 2017 5 comments

Earlier today I was on Bloomberg<GO> when the PCE inflation figures were released. As usual, it was an enjoyable time even if Alix Steel did call me a ‘big inflation nerd’ or something to that effect.

The topic was, of course, PCE – as well as inflation in general, how the Fed might respond (or not), and what the effect of the new Administration’s policies may be. You can see the main part of the discussion here, although not the part where Alix calls me a nerd. A man has some pride.

My main point regarding the PCE report was that PCE isn’t terribly low, but rather right on the long-run average as the chart below (all charts source Bloomberg) shows. Of course, PCE has been lagging behind the rise in CPI, but because it had been “too tight” previously this isn’t yet abnormal.

spread

However, in the interview I didn’t get to the really nerdy part. Perhaps my ego was still stinging and so I didn’t want to highlight the nerdiness?[1] No matter. The nerdy part is that the reason PCE is low is actually no longer because of Medical Care, but because of housing. This next chart plots the spread of core CPI over core PCE, through last month’s figures, versus Owners’ Equivalent Rent (OER).

vs-housing

Housing has a much higher weight in the CPI than in the PCE, and as you can see the plodding nature of OER means that the correlation is somewhat persistent because housing inflation is somewhat persistent. Right now, OER (which, frankly, I thought would have leveled off by now) is rising and showing no signs of slowing, and this fact has served to widen the CPI-PCE spread back to its historical average and likely will cause it to widen to an above-average level. I suppose the good news there is that it is still true that outside of housing, core inflation is still not rising aggressively. Core services ex-housing are looking perkier, but core goods continue to languish as the dollar remains strong. The strength of the dollar almost beggars belief if it’s true that the rest of the world hates us now, but it is what it is.

The bigger point, for markets, is “so what?” There is nothing about a 1.7% core PCE that presents any urgency for Chairman Yellen. As I said on the program: as Yellen approaches the end of her chairmanship (in January 2018, since she insists Trump won’t chase her out before), I believe it is much more likely that she wants to be remembered for pushing the unemployment rate very low – because she believes inflation is easily controlled – than that she wants to be remembered for being a hawk that stopped inflation from getting going. She isn’t worried about inflation, and so the question is whether she wants to be criticized for adding “too many” jobs, or not adding enough. Not that monetary policy has much to do with that, but I believe she clearly will err on the side of keeping policy too loose. The Fed isn’t tightening this week, and I find it unlikely that they will tighten in March, unless inflation expectations rise considerably further than they have already (see chart of 5y5y inflation forward from CPI swaps, below). Even after the big rally since late last year, 5y5y is well below the long-term average through 2014.

5y5y

And even if inflation expectations do rise further, the excuse from the chair will be easy: expectations are rising because the end (and possible reversal) of the globalization dividend and the imposition of tariffs will lead to higher prices. But there is nothing that Fed policy can do about this – it is a supply-side effect, just as high oil prices due to OPEC production restraints would represent a supply-side effect that the Fed shouldn’t respond to. So the excuses are all there for Dr. Yellen. History will show that she missed a chance to shrink the Fed’s balance sheet and avert the worst of the next inflationary upturn, but that history will not be written for some time.

[1] Ridiculous, of course. I embrace my nerdiness, at least when it comes to inflation.

Categories: CPI, Economy, Federal Reserve Tags:
%d bloggers like this: