Archive

Archive for the ‘Causes of Inflation’ Category

What “Transitory Factors” Might Tell Us About Inflation

May 23, 2019 2 comments

There is a lot of buzz around inflation these days. Some people are explaining why we shouldn‘t worry and some people why we should, but regardless – it’s a topic of conversation for the first time in ages. And despite this (or rather, because of it, because I find myself very busy these days), I haven’t written in a long time despite the fact that I have a few things worth writing about. I keep trying, though.

Today I am cheating a bit and taking a column from the quarterly inflation outlook that my company (Enduring Investments) just sent to customers. But I think it is fair to include it here, because the musing was provoked by a recent exchange I had on Twitter while doing my monthly CPI analysis/tweetstorm (follow me @inflation_guy).

As readers know, I tend to focus on Median CPI, rather than Core CPI, as my forecast target variable. The reason is that price changes are rarely distributed randomly. If they were, then the choice of core or median CPI would be irrelevant because they would normally be the same, or roughly the same. But when a distribution has long tails, the ends of the distribution exert a lot of pressure on the average and so median can differ substantially from the mean simply because one tail is much longer than the other even if most of the distribution is similar.

Consider a playground see-saw and imagine that on one side of the see-saw are seated several small children. Think of the “average” of the see-saw system as the point where the see-saw balances. Well, there are lots of ways to balance the system with weight on the other side of the see-saw: a very large weight close to the fulcrum will do it. But the further away from the fulcrum one places the weight, the smaller the weight necessary to balance the scale. As Archimedes said, “give me a lever long enough, and a place to stand, and I can move the world.” The point is that an influence far from the middle of the distribution can have a very significant effect on the average because it is far away from the distribution:  the effect on the mean is (weight * distance from the mean).

Example: the mean of 98% of a distribution is 12. The remaining 2% of the distribution is 28. The weighted average is [(12 * 0.98) + (118 * .02)] / (0.98 + 0.02) = 14.12.  That little 2% caused the mean to go from 12, without the tail, to 14.12 with the tail…a movement of 17.7%! Notice that .02 * 118 = 2.12, which is the amount the mean moved. And if that tail is 228 rather than 118, the mean goes to 16.32. So you see, the length of the tail matters. In both cases, the median was 12, which I would argue is a better indicator of the “central tendency” of the distribution.

(If the distribution is approximately normal, then the tails roughly balance and so the mean and median are about the same. But many economic indicators are not normally distributed, especially ones like income or home prices which are bound by zero on one side. Thus, for many economic series the median, rather than mean, is a better measure. Even though CPI is not bound by zero, it is not normal because prices are not set in a continuous process but instead to have jump-discontinuities.)

The chart below, which I often show in my CPI tweetstorm, shows the see-saw of CPI, where I’ve broken up the index into its lowest-level components and placed those weights on a number line representing the most-recent year-over-year changes. The height of the bar indicates the amount of the basket that sits in that bucket. As you can see, nearly half of the CPI is inflating faster than 3% (which is why Median CPI is 2.8%), and the mode of the distribution is between 3.5% and 4.0%. But because of the far left tail, the mean – which is what core CPI is – is just barely over 2%. Because we have much longer left-hand tails than right-hand tails, the average is biased lower relative to median.

But is this “normal?” Some people have occasionally accused me of picking Median CPI because it tends to be higher, and so the number makes it look like there is more inflation. If the spread were constant, then it would be a bit academic which we chose as the forecast variable, and in fact Core would have a better claim since after all, as consumers purchasing that basket we are in fact paying the average price and not the median.

In fact, though, I think that the tendency of core in recent years to trade below median really is its own interesting story about how prices evolve. If we have 3% inflation, it does not mean that all prices are going up at 3% per year, 0.75% per quarter, 0.25% per month. The price of any given good doesn’t move smoothly but rather episodically, sporadically, spastically. When we are in a disinflationary period, or anyway a low-inflation period, what is happening is that those episodes involve periods of slower prices and “transitory factors” that tend to be on the downside.[1] In that sense, it may be that the Fed, and me/Enduring, both err when they try to look through ‘transitory factors’ because transitory factors may be part of the process. The argument for that perspective is similar to the argument I myself make about why “ex-items” measures make sense when you are looking at an individual company’s earnings but not when you look at the aggregate earnings of the economy. Because bad stuff, or “items,” are always happening to someone somewhere. You can throw them out of any one analysis but if you own the index, you’ll get some of those “items.” You just don’t know from where. Perhaps inflation is the same way.

However, I should point out that median inflation is not always below core. The chart below shows median and core CPI going back to 1983, which is when the Cleveland Fed’s series for Median CPI begins. Notice that from 1983 until 1993, Median CPI was generally lower than core CPI. In 1994, this changed and it has been the opposite ever since.

The year 1994 is significant because that is also the year that most models for inflation that are calibrated on pre-1994 data break down (or, conversely, it is the year prior to which a model calibrated on post-1994 data breaks down). I have written previously about this phenomenon and the fact that the Fed believes this is when inflation expectations abruptly became “anchored,” whatever that means – but *I* believe that this discontinuity is when globalization kicked into high gear with an explosion in the number of bilateral and multilateral trade agreements. It strikes me as plausible that these items are related. When markets are suddenly opened to global competition, affected markets will suddenly show slower price appreciation due to the pressure from that competition (and the replacement of high-cost domestic goods with lower-cost imports). But which market is currently being affected will not stay constant, but change over time. In other words, I think the fact that core has been persistently below median for a long time is a symptom of the globalization “dividend.”

If I am right, and if if I am also right about the arrow of globalization changing direction, then it follows that core and median might flip positions at some point over the next couple of years. And then the “transitory effects” will be mostly on the high side.


[1] It could also be indicative of a bias from the measurers that their improved methods are always looking for lower inflation – not in the “the BLS is making up this *@&$^” sense but in the sense that for lots of reasons the CPI appears to be overstated because of technical details about the functional form, the way measurement errors happen, etc. And so researchers may spend more time looking for ways that inflation is overstated. I don’t think that research bias is actually much of a problem. But I figured I ought to mention that that is one possible interpretation.

Advertisements
Categories: Causes of Inflation, CPI

What if ‘Excess Reserves’ Aren’t Really Excess?

March 4, 2019 1 comment

One intriguing recent suggestion I have heard recently is that the “Excess” reserves that currently populate the balance sheet of the Federal Reserve aren’t really excess after all. Historically, the quantity of reserves was managed so that banks had enough to support lending to the degree which the Fed wanted: when economic activity was too slow, the Fed would add reserves and banks would use these reserves to make loans; when economic activity was too fast, the Fed would pull back on the growth of reserves and so rein in the growth of bank lending. Thus, at least in theory the Open Markets Desk at the New York Fed could manage economic activity by regulating the supply of reserves in the system. Any given bank, if it discovered it had more reserves than it needed, could lend those reserves in the interbank market to a bank that was short. But there was no significant quantity of “excess” reserves, because holding excess reserves cost money (they didn’t pay interest) – if the system as a whole had “too many” reserves, banks tended to lend more and use them up. So, when the Fed wanted to stuff lots of reserves into the system in the aftermath of the financial crisis, and especially wanted the banks to hold the excess rather than lending it, they had to pay banks to do so and so they began to pay interest on reserves. Voila! Excess reserves appeared.

But there is some speculation that things are different now because in 2011, the Basel Committee on Banking Supervision recommended (and the Federal Reserve implemented, with time to comply but fully implemented as of 2015) a rule that all “Systematically Important Financial Institutions” (mainly, really big banks) be required to maintain a Liquidity Coverage Ratio (LCR) at a certain level. The LCR is calculated by dividing a bank’s High Quality Liquid Assets (HQLA) by a number that represents its stress-tested 30-day net outflows. That is, the bank’s liquidity is expressed as a function of the riskiness of its business and the quantity of high-quality assets that it holds against these risks.

In calculating the HQLA, most assets the bank holds receive big discounts. For example, if a bank holds common equities, only half of the value of those equities can be considered in calculating this numerator. But a very few types of assets get full credit: Federal Reserve bank balances and Treasury securities chief among them.[1]

So, since big banks must maintain a certain LCR, and reserves are great HQLA assets, some observers have suggested that this means the Fed can’t really drain all of those excess reserves because they are, effectively, required. They’re not required because they need to be held against lending, but because they need to be held to satisfy the liquidity requirements.

If this is true, then against all my expectations the Fed has, effectively, done what I suggested in Chapter 10, “My Prescription” of What’s Wrong with Money? (Wiley, 2016). I quote an extended section from that book, since it turns out to be potentially spot-on with what might actually be happening (and, after all, it’s my book so I hereby give myself permission to quote a lengthy chunk):

“First, the Federal Reserve should change the reserve requirement for banks. If the mountain will not come to Mohammed, then Mohammed must go to the mountain. In this case, the Fed has the power (and the authority) to, at a stroke, redefine reserves so that all of the current “excess” reserves essentially become “required” reserves, by changing the amount of reserves banks are required to hold against loans. No longer would there be a risk of banks cracking open the “boxes of currency” in their vaults to extend more loans and create more money than is healthy for an economy that seeks noninflationary growth. There would be no chance of a reversion to the mean of the money multiplier, which would be devastating to the inflation picture. And the Open Markets Desk at the Fed would immediately regain power over short-term interest rates, because when they add or subtract reserves in open market operations, banks would care.

“To be sure, this would be awful news for the banks themselves and their stock prices would likely take a hit. It would amount to a forcible deleveraging, and impair potential profitability as a result. But we should recognize that such a deleveraging has already happened, and this policy would merely recognize de jure what has already happened de facto.

“Movements in reserve requirements have historically been very rare, and this is probably why such a solution is not being considered as far as I know. The reserve requirement is considered a “blunt instrument,” and you can imagine how a movement in the requirement could under normal circumstances lead to extreme volatility as the quantity of required reserves suddenly lurched from approximate balance into significant surplus or deficit. But that is not our current problem. Our current problem cries out for a blunt instrument!

“While the Fed is making this adjustment, and as it prepares to press money growth lower, they should work to keep medium-term interest rates low, not raise them, so that money velocity does not abruptly normalize. Interest rates should be normalized slowly, letting velocity rise gradually while money growth is pushed lower simultaneously. This would cause the yield curve to flatten substantially as tighter monetary conditions cause short-term interest rates in the United States to rise.

“Of course, in time the Fed should relinquish control of term rates altogether, and should also allow its balance sheet to shrink naturally. It is possible that, as this happens, reserve requirements could be edged incrementally back to normal as well. But those decisions are years away.”

If, in fact, the implementation of the LCR is serving as a second reserve requirement that is larger than the reserve requirement that is used to compute required and “excess” reserves, then the amount of excess reserves is less than we currently believe it to be. The Fed, in fact, has made some overtures to the market that they may not fully “normalize” the balance sheet specifically because the financial system needs it to continue to supply sufficient reserves. If, in fact, the LCR requirement uses all of the reserves currently considered “excess,” then the Fed is, despite my prior beliefs, actually operating at the margin and decisions to supply more or fewer reserves could directly affect the money supply after all, because the reserve requirement has in effect been raised.

This would be a huge development, and would help ameliorate the worst fears of those of us who wondered how QE could be left un-drained without eventually causing a move to a much higher price level. The problem is that we don’t really have a way to measure how close to the margin the Fed actually is; moreover, since Treasuries are a substitute for reserves in the LCR it isn’t clear that the margin the Fed wants to operate on is itself a bright line. It is more likely a fuzzy zone, which would complicate Fed policy considerably. It actually would make the Fed prone to mistakes in both directions, both over-easing and over-tightening, as opposed to the current situation where they are mostly just chasing inflation around (since when they raise interest rates, money velocity rises and that pushes inflation higher, but raising rates doesn’t also lower money growth since they’re not limiting bank activities by reining in reserves at the margin).

I think this explanation is at least partly correct, although we don’t think the condition is as binding as the more optimistic assessments would have it. The fact that M2 has recently begun to re-accelerate, despite the reduction in the Fed balance sheet, argues that we are not yet “at the margin” even if the margin is closer than we thought it was previously.


[1] The assumption in allowing Treasuries to be used at full value seems to be that in a crisis the value of those securities would go up, not down, so no haircut is required. Of course, that doesn’t always happen, especially if the crisis were to be caused, for example, by a failure of the government to pay interest on Treasuries due to a government shutdown. The more honest reason is that if the Fed were to haircut Treasuries, banks would hold drastically fewer Treasuries and this would be destabilizing – not to mention bad for business on Capitol Hill.

Inflation-Related Impressions from Recent Events

September 10, 2018 2 comments

It has been a long time since I’ve posted, and in the meantime the topics to cover have been stacking up. My lack of writing has certainly not been for lack of topics but rather for a lack of time. So: heartfelt apologies that this article will feel a lot like a brain dump.

A lot of what I want to write about today was provoked/involves last week. But one item I wanted to quickly point out is more stale than that and yet worth pointing out. It seems astounding, but in early August Japan’s Ministry of Health, Labour, and Welfare reported the largest nominal wage increase in 1997. (See chart, source Bloomberg). This month there was a correction, but the trend does appear firmly upward. This is a good point for me to add the reminder that wages tend to follow inflation rather than lead it. But I believe Japanese JGBis are a tremendous long-tail opportunity, priced with almost no inflation implied in the price…but if there is any developed country with a potential long-tail inflation outcome that’s possible, it is Japan. I think, in fact, that if you asked me to pick one developed country that would be the first to have “uncomfortable” levels of inflation, it would be Japan. So dramatically out-of-consensus numbers like these wage figures ought to be filed away mentally.

While readers are still reeling from the fact that I just said that Japan is going to be the first country that has uncomfortable inflation, let me talk about last week. I had four inflation-related appearances on the holiday-shortened week (! is that an indicator? A contrary indicator?), but two that I want to take special note of. The first of these was a segment on Bloomberg in which we talked about how to hedge college tuition inflation and about the S&P Target Tuition Inflation Index (which my company Enduring Investments designed). I think the opportunity to hedge this specific risk, and to create products that help people hedge their exposure to higher tuition costs, is hugely important and my company continues to work to figure out the best way and the best partner with whom to deploy such an investment product. The Bloomberg piece is a very good segment.

I spent most of Wednesday at the Real Return XII conference organized by Euromoney Conferences (who also published one of my articles about real assets, in a nice glossy form). I think this is the longest continually-running inflation conference in the US and it’s always nice to see old friends from the inflation world. Here are a couple of quick impressions from the conference:

  • There were a couple of large hedge funds in attendance. But they seem to be looking at the inflation markets as a place they can make macro bets, not one where they can take advantage of the massive mispricings. That’s good news for the rest of us.
  • St. Louis Fed President James Bullard gave a speech about the outlook for inflation. What really stood out for me is that he, and the Fed in general, put enormous faith in market signals. The fact that inflation breakevens haven’t broken to new highs recently carried a lot of weight with Dr. Bullard, for example. I find it incredible that the Fed is actually looking to fixed-income markets for information – the same fixed-income markets that have been completely polluted by the Fed’s dominating of the float. In what way are breakevens being established in a free market when the Treasury owns trillions of the bonds??
  • Bullard is much more concerned about recession than inflation. The fact that they can both occur simultaneously is not something that carries any weight at the Fed – their models simply can’t produce such an outcome. Oddly, on the same day Neel Kashkari said in an interview “We say that we have a symmetric view of inflation. We don’t mind if it’s 2.1 or 1.9, but in our practice, in what we actually do, we are much more worried about high inflation than we are low inflation. And I think that that is the scar from the 1970s.” That’s ludicrous, by the way – there is no way in the world that the Fed would have done the second and third QEs, with the recession far in the rear view mirror, if the Fed was more concerned with high inflation. Certainly, Bullard showed no signs of even the slightest concern that inflation would poke much above 2%, much less 3%.
  • In general, the economists at the conference – remember, this is a conference for people involved in inflation markets – were uniform in their expectation that inflation is going nowhere fast. I heard demographics blamed (although current demographics, indicating a leftward shift of the supply curve, are actually inflationary it is a point of faith among some economists that inflation drops when the number of workers declines. It’s actually a Marxist view of the economic cycle but I don’t think they see it that way). I heard technology blamed, even though there’s nothing particularly modern about technological advance. Economists speaking at the conference were of the opinion that the current trade war would cause a one-time increase in inflation of between 0.2%-0.4% (depending on who was speaking), which would then pass out of the data, and thought the bigger effect was recessionary and would push inflation lower. Where did these people learn economics? “Comparative advantage” and the gain from trade is, I suppose, somewhat new…some guy named David Ricardo more than two centuries ago developed the idea, if I recall correctly…so perhaps they don’t understand that the loss from trade is a real thing, and not just a growth thing. Finally, a phrase I heard several times was “the Fed will not let inflation get out of hand.” This platitude was uttered without any apparent irony deriving from the fact that the Fed has been trying to push inflation up for a decade and has been unable to do so, but the speakers are assuming the same Fed can make inflation stick at the target like an arrow quivering in the bullseye once it reaches the target as if fired by some dead-eye monetary Robin Hood. Um, maybe.
  • I marveled at the apparent unanimity of this conclusion despite the fact that these economists were surely employing different models. But then I think I hit on the reason why. If you built any economic model in the last two decades, a key characteristic of the model had to be that it predicted inflation would be very low and very stable no matter what other characteristics it had. If it had that prediction as an output, then it perfectly predicted the last quarter-century. It’s like designing a technical trading model: if you design one that had you ‘out’ of the 1987 stock market crash, even if it was because of the phase of the moon or the number of times the word “chocolate” appeared in the New York Times, then your trading model looks better than one that doesn’t include that “factor.” I think all mainstream economists today are using models that have essentially been trained on dimensionless inflation data. That doesn’t make them good – it means they have almost no predictive power when it comes to inflation.

This article is already getting long, so I am going to leave out for now the idea I mentioned to someone who works for the Fed’s Open Market Desk. But it’s really cool and I’ll write about it at some point soon. It’s an idea that would simultaneously be really helpful for investors and help the Fed reduce a balance sheet that they claim to be happy with but we all really know they wish they could reduce.

So I’ll move past last week and close with one final off-the-wall observation. I was poking around in Chinese commodity futures markets today because someone asked me to design a trading strategy for them (don’t ask). I didn’t even know there was such a thing as PVC futures! And Hot Rolled Coils! But one chart really struck me:

This is a chart of PTA, or Purified Terephthalic Acid. What the heck is that? PTA is an organic commodity chemical, mainly used to make polyester PET, which is in turn used to make clothing and plastic bottles. Yeah, I didn’t know that either. Here’s what else I don’t know: I don’t know why the price of PTA rose 50% in less than two months. And I don’t know whether it is used in large enough quantities to affect the end price of apparel or plastic bottles. But it’s a pretty interesting chart, and something to file away just in case we start to see something odd in apparel prices.

Let me conclude by apologizing again for the disjointed nature of this article. But I feel better for having burped some of these thoughts out there and I hope you enjoyed the burp as well.

Developed Country Demographics are Inflationary, not Deflationary

July 17, 2018 4 comments

I’m a relatively simple guy. I like simple models. I get suspicious with models that seem overly complicated. In my experience, the more components you add to a model the more likely it is that one of them ceases having explanatory power and messes up your model’s value. In this it is like (since tonight is Major League Baseball’s All-Star Game I thought I’d use a baseball analogy) bringing in relievers to a game. Every reliever you bring in has some chance that he just doesn’t have it tonight, so therefore you ought to bring in as few relievers as you can.

Baseball managers don’t seem to believe this, so they bring in as many relievers as they can. Similarly, economists don’t seem to believe the rule of parsimony. The more complexity in the model, the better (at least, for the economist’s job security).

Let’s talk about demographics and inflation.

Here’s how I think about how an aging population affects inflation:

  1. Fewer workers in the workforce implies a lower unemployment rate and higher wages, c.p.
  2. A higher retiree/active worker ratio implies lower saving, which will tend to send interest rates higher and equity prices lower, and tend to increase money velocity, c.p.
  3. A higher retiree/non-retiree ratio probably implies lower spending, c.p.

It seems to me that people who argue that aging populations are disinflationary don’t really have a useful model in mind. If they do, then it revolves only around #3, and the idea that spending will diminish over time; if you believe that inflation is related to growth then this sounds like stagnation and deflation. But if there’s lower spending, that doesn’t necessarily indicate a wider output gap because of #1. The best you could say about the effect on the output gap of an aging population is that it is indeterminate: potential output growth should decline because of workforce decline (potential output growth » growth in the # of workers + growth in productivity per worker), while demand growth should also decline, leading to uncertain effects on the output gap.

I think that most people who think the demographic situation of developed nations is disinflationary are really just extrapolating from the single data point of Japan. Japan had an aging population; Japan had deflation; ergo, an aging population causes deflation. But as I’ve argued previously, the main cause of deflation in Japan was overly tight monetary policy.

The decrease in potential growth rates due to the graying of the population is real and clearly inflationary on its face, all else equal. Go look at our MVºPQ calculator and see what happens when you lower the annual real growth assumption, for any other set of assumptions.

So, my model is simple, and you don’t need to have a lot of extraneous dynamics if you simply say: slower potential growth implies higher potential inflation, and demographics implies lower potential future growth. Qed.

One other item I would point out about the three points above: all three are negative for stock markets. If you truly believe that the dominant effects are lower spending, less savings, and higher wages the you can’t possibly think that demographics are anything other than disastrous for equity valuations in the future.

Potpourri for $500, Alex

June 1, 2018 5 comments

When I don’t write as often, I have trouble re-starting. That’s because I’m not writing because I don’t have anything to say, but because I don’t have time to write. Ergo, when I do sit down to write, I have a bunch of ideas competing to be the first thing I write about. And that freezes me a bit.

So, I’m just going to shotgun out some unconnected thoughts in short bursts and we will see how it goes.


Wages! Today’s Employment Report included the nugget that private hourly earnings are up at a 2.8% rate over the last year (see chart, source Bloomberg). Some of this is probably due to the one-time bumps in pay that some corporates have given to their employees as a result of the tax cut, and so the people who believe there is no inflation and never will be any inflation will dismiss this.

On the other hand, I’ll tend to dismiss it as being less important because (a) wages follow prices, not the other way around, and (b) we already knew that wages were rising because the Atlanta Fed Wage Tracker, which controls for composition effects, is +3.3% over the last year and will probably bump higher again this month. But the rise in private wages to a 9-year high is just one more dovish argument biting the dust.

As an aside, Torsten Slok of Deutsche Bank pointed out in a couple of charts today that one phenomenon of recent years has been that people staying in the same jobs increasingly see zero wage growth. Although this is partly because wage growth in general has been low, the spread between wage growth for “job switchers” and “job stayers” is now about 1.25% per year, the highest rate in about 17 years. His point is that as we see more switchers due to a tight labor market, that implies more wage growth (again, the Atlanta Fed Wage Growth Tracker does a better job, so this just means average hourly earnings should increasingly converge with the Atlanta Fed figure).


Today I was on the TD Ameritrade Network and they showed a chart that I’d included in our Quarterly Inflation Outlook (which we distribute to customers). I tweeted the chart back on May 22 but let me put it here, with some brief commentary lifted from our quarterly:

“As economic activity has started to absorb more and more unemployed into the workforce, a shortage has developed in the population of truck drivers. This shortage is not easy to overcome, since it takes time to train new truck drivers (and the robo-truck is still no more than science fiction). Moreover, recent advances in electronically monitoring the number of hours that drivers are on the road – there have been rules governing this for a long time, but they relied on honest reporting from the drivers – have artificially reduced the supply of trucker hours at just the time when more were needed because of economic growth…As a result of this phenomenon, total net-of-fuel-surcharge truckload rates are 15% higher than they were a year ago, which is the highest rate of increase since 2004. As the chart (source: FTR Associates and BLS) illustrates, there is a significant connection between truckload rates lagged 15 months and core inflation (0.74 correlation).”

According to FTR Transportation Intelligence, the US is short about 280,000 truck drivers compared to what it needs.


Remember when everyone said Europe was about to head back into deflation, thanks to that surprise dip in core inflation last month? Here is what I had to say about that on my private Twitter feed (sign up here if this stuff matters to you) at the time.

As Paul Harvey used to say, the rest of the story is that core European CPI printed this month at 1.1%, shocking (almost) everyone for a second month.


I had a conversation recently with a potential client who said they didn’t want to get into a long-commodity strategy because they were afraid of chasing what is hot. It’s a reasonable concern. No one wants to be the pigeon who bought the highs.

But some context is warranted. I didn’t want to be impolite, but I pointed out that what he was saying was that in the chart below, he was afraid it was too late to get on the orange line because it is too hot.

Incidentally, lest you think that I chose that period because it flatters the argument…for every period starting June 30, XXXX and ending June 1, 2018, the orange line is appreciably below the white line and has never been meaningfully above it, for XXXX going back to 2002. For 2002-2011, the two indices shown here were pretty well correlated. Since 2011, it has been a one-way underperformance ticket for commodities. They are many things, but “hot” is not one of them!

I haven’t heard back.

Central Banking Tragedy: The Case of Japan

Today I want to talk about one of the real tragedies of monetary policy and inflation: Japan.

The tragedy is that the mystery of the deflation in Japan is no mystery at all. The cure also was no mystery. So the tragedy is that these were both treated as mysteries by the central bank, which stumbled on the right response and then stumbled right back out of it again.

The chart below shows the money supply and core inflation history of Japan going back into the 1990s. Core inflation is in red (I’ve interpolated through the sales-tax-induced spike) and M2 growth is in blue. The cause of the disinflation is pretty plain: between 1998 and 2013, year/year money growth in Japan never exceeded 4%. From 1999 to 2013, Japanese M2 rose 38% in aggregate; in the US it rose 138% over the same period. It is very hard to get inflation, especially in an environment of declining interest rates, if the money supply is increasing at or somewhat less than the rate of potential GDP growth.

However, in the middle of 2013 Japanese Prime Minister Shinzo Abe persuaded Bank of Japan governor Haruhiko Kuroda to promise to double the money supply in two years, by pursuing massive QE. Although that turned out to be an exaggeration, M2 growth did peek out from behind 4%, and inflation started to perk up as well. It wasn’t a lot, but inflation in 2013 reached new 14-year highs and the economy was officially out of deflation. While QE made very little sense, at least the QE2 and later versions, in the US where inflation was positive and money growth was adequate, it made a ton of sense in Japan. In fact, if Japan had been the only country pursuing QE, I can make the argument that the yen would have likely depreciated substantially and caused inflation in that country.

In the US, the central bank pursued QE and risked inflation to hopefully spur GDP growth. But in Japan, both of those were desirable outcome. There was no reason not to pursue QE in Japan.

But the Bank of Japan lost faith in the power of money growth to cause inflation. While the BoJ continued asset purchases, those purchases diminished and the rate of increase in the central bank’s balance sheet declined from roughly 50% y/y in 2014 to around 8% today (see Chart, source Bloomberg).

Instead, the Bank decided to deploy negative interest rates, dropping the policy rate to -0.10% from +0.10% in early 2016. They did this partly because it was a central bank fad in 2015 and 2016 to experiment with negative interest rates, despite the fact that we have no idea (and no guidance from theory) about what happens to monetary velocity at negative interest rates. I wrote the following in our quarterly piece on February 20, 2016 (and I think it’s been long enough that it’s fair to put it in the public domain as our customers have had plenty of time to read it in private!). I quote at length because, frankly, it was pretty good:

 “…the question of what happens to monetary velocity at negative interest rates is one worth considering since several central banks (the latest one being the Bank of Japan) are now implementing monetary policy with a negative policy rate.

“Why the Bank of Japan is doing so is beyond us. Abstracting from the sales-tax related hike, it has successfully eliminated deflation, driving core inflation from -1.5% to near +1.0% since mid-2010. It has done so, very simply, by working to accelerate money supply growth from the 1.5%-2.0% growth that was the standard in the late pre-crisis period to over 4% by 2014 and 2015.

“This isn’t rocket science; it’s monetary science.

“Now, recently money supply growth has begun to fall off, so the BoJ likely was concerned by that development and wanted to find a way to ensure that inflation doesn’t slip back. If that was their intention, then cutting rates was probably the wrong thing to do. As Friedman explained decades ago, and we have illustrated here repeatedly, money velocity is strongly tied to the level of interest rates. Lower interest rates imply less reason to not hold cash; ergo real cash balances rise and the inverse of the demand for real cash balances is velocity.

“But we said “probably” the wrong thing to do for one reason: we don’t really know whether this relationship holds when interest rates cross the negative bound. It may be the case that this relationship ceases to apply at negative rates even though Friedman’s idea is based on the relative difference between cash yielding zero and longer-term investments or consumption alternatives. The reason that velocity might behave differently at sub-zero rates is that people respond asymmetrically to losses and gains. That is, the pleasure of a gain is dominated by the pain of the same-sized loss, in most people. This cognitive bias may cause savers/investors to behave strikingly different if they are charged for deposits than if they are merely paid zero on those deposits (even if zero is lower than other available rates). In that case, we might see a spike in money velocity once rates go through zero as cash balances become hot-potatoes, just as they would if investment opportunities suddenly appeared. Now, we need to make several observations.

“First, there is no data to suggest this effect exists. What we have posited in the preceding paragraph is rank speculation. (But, unlike the various central banks, we’re not betting our entire monetary policy on that speculation).

“Second, while we don’t really think this effect exists, if it does exist then we would expect it to be a spot discontinuity in the relationship between rates and velocity. That is, the behavior should change between 0% and some negative rate, but then be somewhat linear thereafter. Cognitively, the reaction is both a general loss aversion, which is linear but no different at negative rates from zero, and a behavioral “endowment” reaction that is to the “taking” of money from a person and not necessarily related to the size of the theft.

“Third, if this effect does exist it still doesn’t mean that cutting rates to a negative rate was wise for the Bank of Japan. After all, quantitative easing has done a fine job of pushing up inflation, and so there is no reason to take a speculative gamble like this to keep inflation moving higher. Just do more of the same.

“Fourth: more likely, the BoJ is doing this because it believes that negative rates will stimulate growth. This is much more speculative than you might think, and we may be overgenerous in phrasing the point that way. In any case, any growth benefit would stem either from weakening the currency (which QE would also do, with less risk) or from provoking investment in more marginal ventures that become acceptable at lower financing rates. We call that malinvestment, and it isn’t a good thing.

“Inflation was moving higher in Japan. As long as QE continues (with, or without negative rates), then inflation should continue to move higher. But if the BoJ abandons its successful anti-deflation policy because it was not effective at increasing growth, then it is likely to end up with neither inflation nor growth.”

So, this is the tragedy. As the first chart above illustrates, increasing asset purchases seemed to increase the money supply and inflation; however, cutting interest rates had the evident effect of decreasing monetary velocity enough to push the Japanese economy back to near-deflation. It’s tragic, because they had the right policy and essentially for the right reasons, but changed it to the wrong policy for reasons that are tied to economic dogma which happens to be incorrect (that the act of lowering interest rates causes inflation).

It isn’t the only tragedy to be visited on global economies by too-smart-by-half central bankers, but it is one of the most tragic because they were on the right track before their buddies in the central bankers’ club persuaded them to change tack. It’s like the alcoholic who, having suffered into sobriety, is tempted by friends to have a ‘social drink’ and ends up back in the bottle. At this stage, there is no longer any way for this to end well.

Categories: BOJ, Causes of Inflation, Japan

Why the M2 Slowdown Doesn’t Blunt My Inflation Concern

April 12, 2018 1 comment

We are now all good and focused on the fact that inflation is headed higher. As I’ve pointed out before, part of this is an illusion of motion caused by base effects: not just cell phones, but various other effects that caused measured inflation in the US to appear lower than the underlying trend because large moves in small components moved the average lower even while almost half of the consumption basket continues to inflate by around 3% (see chart, source BLS, Enduring Investments calculations).

But part of it is real – better central-tendency measures such as Median CPI are near post-crisis highs and will almost certainly reach new highs in the next few months. And as I have also pointed out recently, inflation is moving higher around the world. This should not be surprising – if central banks can create unlimited amounts of money and push securities prices arbitrarily higher without any adverse consequence, why would we ever want them to do anything else? But just as the surplus of sand relative to diamonds makes the former relatively less valuable, adding to the float of money should make money less valuable. There is a consequence to this alchemy, although we won’t know the exact toll until the system has gone back to its original state.

(I think this last point is underappreciated. You can’t measure an engine’s efficiency by just looking at the positive stroke. It’s what happens over a full cycle that tells you how efficient the engine is.)

I expect inflation to continue to rise. But because I want to be fair to those who disagree, let me address a potential fly in the inflationary ointment: the deceleration in the money supply over the last year or so (see chart, source Federal Reserve).

Part of my thesis for some time has been that when the Fed decided to raise interest rates without restricting reserves, they played a very dangerous game. That’s because raising interest rates causes money velocity to rise, which enhances inflation. Historically, when the Fed began tightening they restrained reserves, which caused interest rates to rise; the latter effect caused inflation to rise as velocity adjusted but over time the restraint of reserves would cause money supply growth (and then inflation) to fall, and the latter effect predominated in the medium-term. Ergo, decreasing the growth rate of reserves tended to cause inflation to decline – not because interest rates went up, which actually worked against the policy, but because the slow rate of growth of money eventually compounded into a larger effect.

And so my concern was that if the Fed moved rates higher but didn’t do it by restraining the growth rate of reserves, inflation might just get the bad half of the traditional policy result. The reason the Fed is targeting interest rates, rather than reserves, is that they have no power over reserves right now (or, at best, only a very coarse power). The Fed can only drain the inert excess reserves, which don’t affect money supply growth directly. The central bank is not operating on the margin and so has lost control of the margin.

But sometimes they get lucky, and they may just be getting lucky. Commercial bank credit growth (see chart, source Federal Reserve) has been declining for a while, pointing to the reason that money supply growth is slowing. It isn’t the supply of credit, which is unconstrained by reserves and (at least for now) unconstrained by balance sheet strength. It’s the demand for credit, evidently.

Now that I’ve properly laid out that M2 is slowing, and that declining M2 growth is typically associated with declining inflation (and I haven’t even yet pointed out that Japanese and EU M2 growth are both also at the lowest levels since 2014), let me say that this could be good news for inflation if it is sustained. But the problem is that since the slowing of M2 is not the result of a conscious policy, it’s hard to predict that money growth will stay slow.

The reason it needs to be sustained is that we care about percentage changes in the stock of money plus the percentage change in money velocity. For years, the latter term has been a negative number as money velocity declined with interest rates. But M2 velocity rose in the fourth quarter, and my back-of-the-envelope calculation suggests it probably rose in Q1 as well and will rise again in Q2 (we won’t know Q1’s velocity until the advance GDP figures are reported later this month). If interest rates normalize, then it implies a movement higher in velocity to ‘normal’ levels represents a rise of about 12-14% from here (see chart, source Bloomberg.[1])

If money velocity kicks in 12-14% over some period to the “MVºPQ” relationship, then you need to have a lot of growth, or a pretty sustained decline in money growth, to offset it. The following table is taken from the calculator on our website and you can play with your own assumptions. Here I have assumed the economy grows at 2.5% per year for the next four years (no mean feat at the end of a long expansion).

The way to read this chart is to say “what if velocity over the next four years returns to X. Then what money growth is associated with what level of inflation?” So, if you go down the “1.63” column, indicating that at the end of four years velocity has returned to the lower end of its long-term historical range, and read across the M2 growth rate row labeled “4%”, you come to “4.8%,” which means that if velocity rises to 1.63 over the next four years, and growth is reasonably strong, and money growth remains as slow as 4%, inflation will average 4.8% per year over those four years.

So, even if money growth stays at 4% for four years, it’s pretty easy to get inflation unless money velocity also stays low. And how likely is 4% money growth for four years? The chart below shows 4-year compounded M2 growth rates back thirty or so years. Four percent hasn’t happened in a very long time.

Okay, so what if velocity doesn’t bounce? If we enter another bad recession, then it’s conceivable that interest rates could go back down and keep M2 velocity near this level. This implies flooding a lot more liquidity into the economy, but let’s suppose that money growth is still only 4% because of tepid credit demand growth and velocity stays low because interest rates don’t return to normal. Then what happens? Well, in this scenario presumably we’re no longer looking at 2.5% annual growth. Here’s rolling-four-year GDP going back a ways (source: BEA).

Well, let’s say that it isn’t as bad as the Great Recession, and that real growth only slows a bit in fact. If we get GDP growth of 1.5% over four years, velocity stays at 1.43, and M2 grows only at 4%, then:

…you are still looking at 2.5% inflation in that case.

I’m going through these motions because it’s useful to understand how remarkable the period we’ve recently been through actually is in terms of the growth/inflation tradeoff, and how unlikely to be repeated. The only reason we have been able to have reasonable growth with low inflation in the context of money growth where it has been is because of the inexorable decline in money velocity which is very unlikely to be repeated. If velocity just stops going down, you might not have high inflation numbers but you’re unlikely to get very low inflation outcomes. And if velocity rises even a little bit, it’s very hard to come up with happy outcomes that don’t involve higher inflation.

I admit that I am somewhat surprised that money growth has slowed the way it has. It may be just a coin flip, or maybe credit demand is displaying some ‘money illusion’ and responding to higher nominal rates even though real rates have not changed much. But even then…in the last tightening cycle, the Fed hiked rates from 1% to 5.25% over two years in 2004-2006, and money growth still averaged 5% over the four years ended in 2006. While I’m surprised at the slowdown in money growth, it needs to stay very slow for quite a while in order to make a difference at this point. It’s not the way I’d choose to bet.


[1] N.b. Bloomberg’s calculation for M2 velocity does not quite match the calculation of the St. Louis Fed, which is presumably the correct one. They’re ‘close enough,’ however, for this purpose, and this most recent print is almost exactly the same.

%d bloggers like this: