Archive for the ‘Good One’ Category

Anchors Aweigh!

October 19, 2015 2 comments

I think it is time to talk a little bit about “anchored inflation expectations.”

Key to a lot of the inflation modeling at the Fed, and in some sterile economics classrooms around the country, is the notion that inflation is partially shaped by the expectations of inflation. Therefore, when people expect inflation to remain down, it tends to remain down. Thus, you often hear Fed officials talk about the importance of inflation expectations being anchored, and that phrase appears often in Federal Reserve statements and minutes.

I have long found it interesting that with as much as the Fed relies on the notion that inflation expectations are anchored, they have no way to accurately measure inflation expectations. Former Fed Chairman Bernanke said in a speech in 2007 that three important questions remain to be addressed about inflation expectations:

  1. How should the central bank best monitor the public’s inflation expectations?
  2. How do changes in various measures of inflation expectations feed through to actual pricing behavior?
  3. What factors affect the level of inflation expectations and the degree to which they are anchored?

According to Bernanke, the staff at the Federal Reserve struggle with even the first of these questions (“while inflation expectations doubtless are crucial determinants of observed inflation, measuring expectations and inferring just how they affect inflation are difficult tasks”), although this has not deterred them from tackling the second and third questions. Economists use the Hoey survey, the Survey of Professional Forecasters, the Livingston survey, the Michigan survey, and inflation breakevens derived from the TIPS or inflation swaps markets. But all of these suffer from the fundamental problem that what constitutes “inflation” is a difficult question in itself and answering a question about a phenomenon that is hard to quantify viscerally probably causes people to respond to surveys with an answer indicating what they expect the well-known CPI measure to show. I talked about many of these problems in my paper on measuring inflation expectations (“Real-Feel” Inflation: Quantitative Estimation of Inflation Perceptions), but the upshot is that we don’t have a good way to measure expectations.

So, with that as background, consider this fact: next year, some Medicare participants will face a 0% increase in premiums while some Medicare participants will face increases of more than 50%.

I am skeptical of the notion of inflation anchoring. But I am really skeptical if it is the case that different segments of the population see totally different inflation pictures. Which anchor counts, if one large group of people expects 7% inflation and another large group expects 1% inflation?

I would argue that none of those anchors matter, because the whole notion is silly. Let’s think through the mechanism of “inflation anchoring.” So the idea is that when people expect lower inflation, they make decisions that tend to produce lower inflation. What decisions are those? If you expect 1% inflation, but Medicare costs go up 50%, what decision are you going to make that will cause that increase to be closer to your expectations? If eggs go up 25 cents per carton and you were expecting 5 cents…is the idea that no one will buy eggs and so the vendor will have to lower the price? What about his costs? Pretty clearly, the mechanism will have to work on the seller’s side, but since every seller is a buyer except for the original seller of labor, the idea must be that if people expect high inflation they argue for higher wages, which causes prices to rise.


I have put paid to that notion in this space before. It doesn’t make any sense to think that wages lead inflation, for if they did then we would all love inflation because we would always be ahead of it. But we know that’s not how it works – prices rise, and then we get higher wages. And sometimes we don’t.

Let’s try another hypothetical. Suppose the Federal Reserve literally drops $50 trillion, unexpectedly, from helicopters. And suppose that consumers did not change their expectations for inflation because they believed, much like the Fed does, that money doesn’t play a role in causing inflation – in other words, their expectations were “extremely well-anchored.” Does anyone think that the price level wouldn’t change, a lot, in contrast to the expectations of the crowd? (I sometimes wonder if Lewis Carroll’s Red Queen, who “sometimes…believed as many as six impossible things before breakfast,” was a Fed economist.)

The whole idea that inflation expectations matter is an effort to explain why parameterizations of inflation models have a regime break in the early 1990s. That is, you can fit a model to 1970-1992, or to 1994-present, but you need different parameters for almost anything you try in the Keynesian-modeling world. Econometricians know that outcome means that you are missing an explanatory variable somewhere; econometricians also know that a very convenient way to gloss over the problem is to introduce a “dummy” variable. In this case, the dummy variable is explained as “inflation expectations became anchored in the early 1990s.”

With all of the problems affecting the notion of expectations-anchoring, I find this solution to the modeling problem deeply unsatisfying. I do not believe that inflation expectations anchor for everybody collectively, but that different groups of people have different (and widely different) anchors. And I don’t think that these anchors themselves play much of a role at all in causing a certain level of inflation. There are better models, simpler models, which do not require you to believe six impossible things.

Unfortunately, they do require you to believe in monetarism. And to some people, that is a seventh impossible thing.

The New Fed Operating Framework Explained with Cheese

October 8, 2015 2 comments

This will be a brief but hopefully helpful column. For some time, I have been explaining that the new Fed operating framework for monetary policy, in which the FOMC essentially steers interest rates higher by fiat rather than in the traditional method (by managing the supply of funds and therefore the resulting pressure on reserves), is a really bad idea. But in responding to a reader’s post I inadvertently hit on an explanation that may be clearer for some people than my analogy of a doctor manipulating his thermometer to give the right reading from the patient.

Right now, there is a tremendous surplus of reserves above what banks are required to hold or desire to hold. With free markets, this would result in a Fed funds interest rate of zero, or even lower under some circumstances, with a substantial remaining surplus.[1] In this case, the Fed funds effective rate has tended to be in the 10-20bps range since the Fed started paying interest on excess reserves (IOER).

So what happens when there is a floor price established above the market-clearing price? Economics 101 tells us that this results in surplus, with less exchange and higher prices than at equilibrium. Consider a farm-price support program where the government establishes a minimum price for cheese (as it has, actually, in the past). If that price is below the natural market-clearing price, then the floor has no effect. But if the price is above the natural market-clearing price, as in the chart below where the minimum cheese price is set at a, then in the market we will see a quantity of cheese traded equal to b, at a price of a.


But what also happens is that producers respond to the higher price by producing more cheese, which is why the supply curve has the shape it does. In order to keep this excess cheese from pushing market prices lower, the government ends up buying c-b cheese at some expense that ends up being a transfer from government to farmers. It can amount to a lot of cheese.[2] This is the legacy of farm price supports: vast warehouses of products that the government owns but cannot distribute, because to distribute them would push prices lower. So the government ends up distributing them to people who wouldn’t otherwise buy cheese, at a zero price. And eventually, we get the Wikipedia entry “government cheese.”

Now, this is precisely what has happened with the artificial price support for overnight interest rates. Whatever the clearing interest rate is with the current level of reserves, it is lower than the 0.25% IOER (and we know this, among other ways, because there are excess reserves. If the price floating to the actual clearing price, then there would be no excess reserves, although the mechanism for this result is admittedly more confusing than it is for cheese). So the Federal Reserve is forced to “buy up the surplus reserves” by paying interest on these reserves; this amounts to a transfer from the government to banks, rather than to farmers in the cheese example.

You should realize too that setting the floor rate higher than the market-clearing rate artificially reduces the volume of trade in reserves. The chart below, which comes from this article on the New York Fed’s blog, illustrates this nicely.


Creating such a floor also causes the supply of excess reserves themselves to increase beyond what it would otherwise be. This confusing result derives because while the Fed supplies the total reserves number to the market, banks can choose to create more “excess” reserves by doing less lending, or can create fewer excess reserves by doing more lending. Of course, banks aren’t deciding to create excess reserves per se; they are deciding whether it is more advantageous to make a loan or to earn risk-free money on the excess reserves. A higher floor rate implies less lending, all else equal – and, as I have said in the past, this means the Fed could cause a huge increase in bank lending by setting IOER at a penalty rate. This would create the conditions necessary for these lines to cross in negative nominal interest rate territory, with much higher volumes of credit and much lower levels of excess reserves being the result.

In this environment, and as recognized by the Sack-Gagnon framework that is now the presumed operating framework for Fed policy, raising IOER is the only way to change the overnight interest rates unless the Desk undertakes to shift the entire supply curve heavily to the left, by draining trillions in reserves. But raising IOER, just like raising the floor price of cheese, will create more imbalances: bigger excess reserves, less lending, and a bigger transfer from government to banks.

(Note: this is subtly different from what I have said before, which is that raising IOER will have no effect on the growth rate of the transactional money supply. Depending on the shape of the supply curve, it will reduce lending which in turn may reduce the growth rate of the monetary aggregates that we care about, such as M2. My suspicion is that the supply curve is in fact pretty steep, meaning that banks are relatively insensitive to small changes in rates, and thus loans and hence the monetary aggregates won’t see much change in the rate of growth – or, more likely, any change will be the result of other effects beyond this one such as the effect of general economic prospects on the quality of credits and the demand for loans).

Price supports, as any economist can tell you, are an inefficient way to subsidize an industry. And in fact, I don’t think the Fed is really interested in subsidizing banks at this stage in the cycle: they seem to be doing just fine. But they are taking on all of these imbalances, creating all of this government cheese, because they believe the effects I talk about parenthetically above are quite large, rather than vanishingly small as I believe. And the ancillary effect, by raising interest rates, is to spur money velocity – an unmitigated negative in this environment, as it will push inflation higher.

Now, all of this discussion may be moot since the current betting is that the Fed won’t raise interest rates any time soon. But it is good to understand this mechanism as clearly as we can, so that we can prepare ourselves for those effects when they occur.

[1] It is really hard to say how low interest rates would go, and/or how much surplus would remain, because we have no idea at all what the supply and demand curves for funds look like at sub-zero rates. Most likely there is a discontinuity at a zero rate, but how much of one and the elasticities of supply and demand below zero are likely to be “weird.”

[2] In fact, in high school I won an economics prize for my paper “That’s a Lotta Cheese.” No joke.

Whither (Wither?) Profits

April 22, 2015 4 comments

Surprisingly, markets are treading water here. The dollar, interest rates, and stocks are all oscillating in a narrow range. In some ways, this is surprising. It does not shock me that interest rates are fairly boring right now, with the 10-year yield trading almost exclusively within 25bps of 2% since November. Market participants are divided between those who see the Fed’s cessation of QE as indicative that prices should decline to fair market-clearing levels (that is, higher yields) and those who see weakness economically both domestically and abroad. There is room for confusion here.

I am similarly not terribly shocked that the dollar is consolidating after a long run, especially when part of that run was fueled by the popular delusion that the Federal Reserve had suddenly become extremely hawkish and would preemptively hike rates before convincing signs of inflation arose. I am hard-pressed to think of a time when the Fed pre-emptively did anything, but that was the popular belief in any event. Now that it is becoming clear that a hike in rates in June is about as likely as the possibility that the Easter Bunny will deliver eggs at the same time, dollar traders who were relying on widening interest rate differentials are pausing to take stock of the situation. I will say that it certainly seems plausible to me that the dollar’s rally will continue for at least a little while, due to the volatility coming our way as the Greek drama plays out, but the buck is not an automatic buy either. Money growth in the U.S. continues to outpace money growth in most other economies (see chart, source Bloomberg), although it is a much closer thing these days.


An increase in relative supply, if the demand curves are similar, should provoke a decrease in relative price. Unless you believe that the Fed isn’t just going to increase rates but is also going to shrink its balance sheet so that money growth abates eventually, it is hard to envision the dollar launching continuously higher. More likely is that as more and more currencies see their supplies increase, the exchange rates meander but the whole kit-and-kaboodle loses ground to real assets.

One of those real assets is housing. An underpinning to my argument, for several years running now, that core prices were not going to be deflating any time soon was the observation that housing prices (and hence rents, with a lag) have been rising rapidly once again. The deceleration in the year/year growth rates in 2014 was a positive sign, but the increase in prices in 2012 and 2013 is still pressing rents higher now and any sag in rents is yet to be felt. However, today’s release of FHA price index data as well as the Existing Home Sales report suggests that it is premature to expect this second housing bubble to unwind gently. The chart below is the year/year change in the median price of existing homes (source: Bloomberg). The recent dip now seems to have been an aberration, and indeed the slowdown in 2014 may have merely presaged the next acceleration higher.


And that bodes ill for core (median) price pressures, which have been steady around 2.2% for a while but may also be readying for the next leg up. Review my post-CPI summary for some of the fascinating details! (Well, fascinating to me.)

This doesn’t mean that I am sanguine about growth, either domestic or global, looking forward. I thought we would get out of 2014 without a recession, but I am less sure about 2015. Europe is going to do better, thanks to weaker energy and a weaker currency (although the weaker currency counteracts some of the energy weakness), but the structural problems in Europe are profound and the exit of Greece will cause turmoil in the banks. But US growth is in trouble: the benefit from lower energy prices is diffuse, while the pain from lower energy prices is concentrated in a way it hasn’t been in the past. And the dollar strength pressures company earnings, as we have seen, on a broad basis. And that’s where it is a little surprising that we are seeing water-treading. It gets increasingly difficult for me to figure out what equity buyers are seeing. Profits are flattening out and even weakening, and they are already at a very high level of GDP so that any economic weakness is going to be felt in profits directly. Furthermore, I find it very interesting that the last time actual reported profits diverged from “Kalecki Profits” corresponded to the last equity bubble (see chart, source Bloomberg).


“Kalecki Profits” is a line that computes corporate profits as Investment minus Household Savings minus Government Savings minus Foreign Savings plus Dividends. Look up Kalecki Profit Equation on Wikipedia for a further explanation. The “Corp Business Prof After Tax” is from the Federal Reserve’s Flow of Funds Z.1 report and is measured directly. The implication is that if companies are reporting greater profits than the sum of the whole, then the difference is suspect. For example, leverage: by increasing financial leverage, the same top line creates more of a bottom line (in either direction). The chart below (source: Federal Reserve; Enduring Investments analysis) plots the 1-year percentage change in business debt outstanding (lagged 2 quarters to center it on the year in question) versus the difference between the two lines in the prior chart.


We might call this “pretty cool,” but in econometrics terms this is merely an explanatory relationship. That is, it doesn’t really help us other than to help explain why the two series diverge. It doesn’t, for example, tell us whether Kalecki profits will converge upwards to reported profits, or whether reported profits will decline; it doesn’t tell us whether it is a decline or deceleration in business debt outstanding that prompts that convergence or whether something else causes both things to happen. I think it’s unlikely that the divergence in the two profit measures causes the change in debt, but it’s possible. I will say that this last chart makes me more comfortable that the Kalecki equation isn’t broken, but merely that it isn’t capturing everything. And my argument, for what it is worth, would be that business leverage cannot increase without bound. At some point, business borrowing will decline.

It does not look like that is happening yet. I have been reading recently about how credit officers have been declining credit more frequently recently. That may be true, but it isn’t resulting in slower credit growth. Commercial bank credit growth, according to the Fed’s H.8 report and illustrated below, continues to grow at the fastest y/y pace since well before the crisis.


If credit officers are really declining credit more often than before, it must mean that applications are up, or that the credit is being extended on fewer loans (that is, to bigger borrowers). Otherwise, we can’t square the fact that there’s rapid credit growth with the proffered fact that credit is being declined more often.

There is a lot to sort through here, but the bottom line is this: I have no idea what the dollar is going to do. I am not sure what the bond market will do. I have no idea what stocks will do. But, if I have to invest (and I do!), then in general I am aiming for real assets and avoiding financial assets.

Summary (and Extension!) of My Post-CPI Tweets

February 26, 2015 3 comments

Below you can find a recap and extension of my post-CPI tweets. You can 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.

  • CPI -0.7%, core +0.2%. Ignore headline. Annual revisions as well.
  • Core +0.18% to two decimals. Strong report compared to expectations.
  • Core rise also off upwardly-revised prior mo. Changing seasonal adj doesn’t affect y/y but makes the near-term contour less negative.
  • y/y core 1.64%, barely staying at 1.6% on a rounded basis.
  • Core for last 4 months now 0.18, 0.08, 0.10, 0.18. The core flirting with zero never made a lot of sense.
  • Primary rents 3.40% from 3.38% y/y, Owners’ Equiv to 2.64% from 2.61%. Small moves, right direction.
  • Overall Housing CPI fell to 2.27% from 2.52%, as a result of huge drop in Household Energy from 2.53% to -0.06%. Focus on the core part!
  • RT @boes_: As always you have to be following @inflation_guy on CPI day >>Thanks!
  • A bit surprising is that Apparel y/y rose to -1.41% from -1.99%. I thought dollar strength would keep crushing Apparel.
  • Also New & Used Motor Vehicles -0.78% from -0.89%. Also expected weakness there from US$ strength. Interesting.
  • Airline fares, recently a big source of weakness, now -2.98% y/y from -4.71% y/y.
  • 10y BEI up 4bps at the moment. And big extension tomorrow. Ouch, would hate to have bet wrong this morning.
  • Medical Care 2.64% y/y from 2.96%.
  • College tuition and fees 3.64% from 3.43%. Child care and nursery school 3.05% from 2.24%. They get you both ends.
  • Core CPI ex-[shelter] rose to 0.72% from 0.69%. Still near an 11-year low.
  • Overall, core services +2.5% (was +2.4%), core goods -0.8% (was -0.8%). The downward pressure on core is all from goods side.
  • …and goods inflation tends to be mean-reverting. It hasn’t reverted yet, and with a strong dollar it will take longer, but it will.
  • That’s why you can make book on core inflation rising.
  • At 2.64% y/y, OER is still tracking well below our model. It will continue to be a source of upward pressure this year.
  • Thank you for all the follows and re-tweets!
  • Summary: CPI & the assoc. revisions eases the appearance that core was getting wobbly. Median has been strong. Core will get there.
  • Our “inflation angst” index rose above 1.5% for the 1st time since 2011. The index measures how much higher inflation FEELS than it IS.
  • That’s surprising, and it’s partly driven by increasing volatility in the inflation subcomponents. Volatility feels like inflation.
  • RT @czwalsh: @inflation_guy @boes_ using surveys? >>no. Surveys do a poor job on inflation. See why here:  …
  • 10y BEI now up 5.25bps. 1y infl swaps +28bps. Hated days like this when I made these markets. Not as bad from this side.
  • Incidentally, none of this changes the Fed outlook. Median was already at target, so the Fed’s focus on core is just a way to ignore it.
  • Once core rises enough, they will find some other reason to not worry about inflation. Fed isn’t moving rates far any time soon.
  • Median CPI +0.2%. Actually slightly less, keeping the y/y at 2.2%.

What a busy and interesting CPI day. For some months, the inflation figures have been confounding as core inflation (as always, we ignore headline inflation when we are looking at trends) has consistently stayed far away from better measures of the central tendency of inflation. The chart below (source: Bloomberg), some version of which I have run quite a bit in the past, illustrates the difference between median CPI (on top), core CPI (in the middle), and core PCE (the Fed’s favorite, on the bottom).


I often say that median is a “better measure of central tendency,” but I haven’t ever illustrated graphically why that’s the case. The following chart (source: Enduring Investments) isn’t exactly correct, but I have removed all of the food and beverages group and the main places that energy appears (motor fuel, household energy). We are left with about 70% of the index, about a third of which sports year-on-year changes of between 2.5% and 3.0%. Do you see the long tail to the left? That is the cause of the difference between core and median. About 12% of CPI, or about one-sixth of core, is deflating. And, since core is an average, that brings the average down a lot. Do you want to guide monetary policy on the basis of that 12%, or rather by the middle of the distribution? That’s not a trick question, unless you are a member of the FOMC.


Now, let’s talk about the dollar a bit, since in my tweets I mentioned apparel and autos. Ordinarily, the connection between the dollar and inflation is very weak, and very lagged. Only for terribly large movements in the dollar would you expect to see much movement in core inflation. This is partly because the US is still a relatively closed economy compared to many other smaller economies. The recent meme that the dollar’s modest rally to this point would impress core deflation on us is just so much nonsense.

However, there are components that are sensitive to the dollar. Apparel is chief among them, mainly because very little of the apparel that we consume is actually produced in the US. It’s a very clean category in that sense. Also, we import a lot of autos from both Europe and Asia, and they compete heavily with domestic auto manufacturers. As a consequence, the connection between these categories and the dollar is much better. The chart below shows a (strange) index of New Cars + Apparel, compared to the 2-year change in the broad trade-weighted dollar, lagged by 1 year – which essentially means that the dollar change is ‘centered’ on the change in New Cars + Apparel in such a way that it is really a 6-month lag between the dollar and these items.


It’s not a day-trading model, but it helps explain why these categories are seeing weakness and probably will see weakness for a while longer. And guess what: those categories account for around 7% of the “tail” in that chart above. Ergo, core will likely stay below median for a while, although I think both will resume upward movement soon.

One of the reasons I believe the upward movement will continue soon is that housing continues to be pulled higher. The chart below (source: Enduring Investments using Bloomberg data) shows a coarse way of relating various housing price indicators to the owners’ rent component of CPI.


We have a more-elegant model, but this makes the point sufficiently: OER is still below where it ought to be given the movement in housing prices. And shelter is a big part of the core CPI. If shelter prices keep accelerating, it is very hard for core (and median) inflation to decline very much.

One final chart (source Enduring Investments), relating to my comment that our inflation angst index has just popped higher.


This index is driven mainly by two things: the volatility of the various price changes we experience, and the dispersion of the price changes we experience. The distribution-of-price-changes chart above shows the large dispersion, which actually increased this month. Cognitively, we tend to overlook “good” price changes (declines, or smaller advances) and recall more easily the “bad”, “painful” price changes. Also, we tend to encode rapid up-and-down changes in prices as inflation, even if prices aren’t actually going anywhere much. I reference my original paper on the subject above, which explains the use of the lambda. What is interesting is the possibility that the extremely low levels of inflation concern that we have seen over the last couple of years may be changing. If it does, then wage pressures will tend to follow price pressures more quickly than they might otherwise.

Thanks for all the reads and follows today. I welcome all feedback!

Money, Commodities, Balls, and How Much Deflation is Enough?

January 22, 2015 2 comments

Money: How Much Deflation is Enough?

Once again, we see that the cure for all of the world’s ills is quantitative easing. Since there is apparently no downside to QE, it is a shame that we didn’t figure this out earlier. The S&P could have been at 200,000, rather than just 2,000, if only governments and central banks had figured out a century ago that running large deficits, combined with having a central bank purchase large amounts of that debt in the open market, was the key to rallying assets without limit.

That paragraph is obviously tongue-in-cheek, but on a narrow time-scale it really looks like it is true. The Fed pursued quantitative easing with no yet-obvious downside, and stocks blasted off to heights rarely seen before; the Bank of Japan’s QE has added 94% to the Nikkei in the slightly more than two years since Abe was elected; and today’s announcement by the ECB of a full-scale QE program boosted share values by 1-2% from Europe to the United States.

The ECB’s program, to be sure, was above expectations. Rather than the €50bln per month that had been mooted over the last couple of days with little currency-market reaction, the ECB pledged €60bln. And they promised to continue until September 2016, making the total value of QE around €1.1 trillion. (That’s about $1.3 trillion at today’s exchange rate, but of course if it works then it will be much less than $1.3 trillion at the September 2016 exchange rate). To be sure, a central bank always has the prerogative to change its mind, but on the risks of a sudden change in policy please see “Swiss National Bank”. It really is remarkable that Draghi was able to drag the Bundesbank kicking and screaming into this policy choice, and it is certain to end the threat of primary deflation in Europe just as it did in the U.S. and in Japan. It will likely also have similar effects on growth, which is to say “next to nothing.” But in Europe, deflation risks stemming from slow money growth had been a risk (see chart, source Bloomberg).


Interestingly, y/y money growth had already been accelerating as of late last year – the ECB releases M2 with a very long lag – but this puts the dot on the exclamation point. The ECB has said “enough!” There will be no core deflation in Europe.

Commodities: How Much Deflation is Enough?

Last week, in “Commodities Re-Thunk” and “Little Update on Commodities Re-Thunk”, I presented the results of using a generalization of the Erb & Harvey approach to forecast expected long-term real returns for commodities. It occurred to me that, since I have previously played with long-term real equity returns, and we have the real yield on 10-year TIPS as well, that it would be interesting to see if using these figures might produce a useful strategy for switching between assets (which doesn’t change the fact that I am a long-term investor; this is still based on long-term values. We merely want to put our assets in whatever offers the best long term value at the moment so as to maximize our expected long-term return).

The answer is yes. Now, I did a more-elegant version of what I am about to show, but the chart below shows the results of switching 100% of your assets between stocks, commodities, and TIPS based on which asset class had the highest expected real yield at a given month-end. Each line is an asset class, except for the blue line which shows the strategy result.


The labels at the top show the asset class that dominated for a long period of time. In 2005 there were a couple of quick crossovers that had little impact, but by and large there were three main periods: from 1999-2005, commodities offered excellent expected real returns; from mid-2005 through early-2008 the strategy would have been primarily in TIPS, and subsequent to that the strategy would have been primarily in equities. Fascinating to me is that the overall strategy does so well even though it would have been invested in equities throughout the crash in 2008. The crash in commodities was worse.

Now what is really interesting is that there is a vertical line at the far right-hand side of the chart. That is because at the end of December, the expected real return to commodities finally exceeded that of equities for the first time in a very long time. For this “selling out” strategy, that means you should be entirely out of stocks and TIPS and entirely in commodities.

As I said, that is the coarse version of this approach. My more-elegant version optimized the portfolio to have a constant expected risk in real terms. It was much less risky as a result (10.5% annualized monthly standard deviation compared to 15.5% for the strategy shown above), had lower turnover, but still sported returns over this period of 9.5% compounded compared to 11.2% for the strategy above. I am not, in other words, suggesting that investors put 100% of their assets in commodities. But this method (along with lots of other signals) is now suggesting that it is time to put more into commodities.

Balls: How Much Deflation is Enough?

Being a football fan, I can’t keep from weighing in on one mystery about deflate-gate (incidentally, why do we need to put ‘gate’ on the end of every scandal? It wasn’t Water-gate, it was the Watergate Hotel that proved Nixon’s undoing. “Gate” is not a modifier). Really, this part isn’t such a mystery but I have seen much commentary on this point: “How did the balls get deflated during the game since they were approved before the game?”

The answer is really simple in the real world: the official picked up one of the balls, said “fine”, and put them back in the bag. He has a million things to do before the championship game and in years of refereeing he has probably never found even one ball out of spec. This sort of error happens everywhere there are low reject rates, and it’s why good quality control is very difficult. (Now, if you fired the ref every time a bad ball got through, you damn betcha those balls would be measured with NASA-like precision – which is perhaps a bad metaphor, since similar issues contributed to the Challenger disaster). The real mystery to me is: if the Patriots truly think they are the better team, why would they cheat, even a little? As with the CHF/EUR cross that we discussed yesterday, the downside is far worse than the gain on the upside.

Or, is it? The NFL will have a chance to establish the cost of recidivism in cheating. Maybe the Patriots were simply betting that the downside “tail” to their risky behavior was fairly short. If the NFL wants to put a stop to nickel-and-dime cheats, it can do that by dropping the hammer here.

Commodities Re-Thunk

January 13, 2015 12 comments

I want to talk about commodities today.

To be sure, I have talked a lot about commodities over the last year. Below I reprise one of the charts I have run in the past (source: Bloomberg), which shows that commodities are incredibly cheap compared to the GDP-adjusted quantity of money. It was a great deal, near all-time lows this last summer…until it started creating new lows.


Such an analysis makes sense. The relative prices of two items are at least somewhat related to their relative scarcities. We will trade a lot of sand for one diamond, because there’s a lot of sand and very few diamonds. But if diamonds suddenly rained down from the sky for some reason, the price of diamonds relative to sand would plummet. We would see this as a decline in the dollar price of diamonds relative to the dollar price of sand, which would presumably be stable, but the dollar in such a case plays only the role of a “unit of account” to compare these two assets. The price of diamonds falls, in dollars, because there are lots more diamonds and no change in the amount of dollars. But if the positions were reversed, and there were lots more dollars, then the price of dollars should fall relative to the price of diamonds. We call that inflation. And that’s the reasoning behind this chart: over a long period of time, nominal commodities prices should grow with as the number of dollars increases.

Obviously, this has sent a poor signal for a while, and I have been looking for some other reasonable way to compute the expected return on commodities.[1] Some time ago, I ran across an article by Erb and Harvey called The Golden Dilemma (I first mentioned it in this article). In it was a terrific chart (their Exhibit 5) which showed that the current real price of gold – simply, gold divided by the CPI price index – is a terrific predictor of the subsequent 10-year real return to gold. That chart is approximately reproduced, albeit updated, below. The data in my case spans 1975-present.


The vertical line indicates the current price of gold (I’ve normalized the whole series so that the x-axis is in 2015 dollars). And the chart indicates that over the next ten years, you can expect something like a -6% annualized real return to a long-only position in gold. Now, that might happen as a result of heavy inflation that gold doesn’t keep up with, so that the nominal return to gold might still beat other asset classes. But it would seem to indicate that it isn’t a great time to buy gold for the long-term.

This chart was so magnificent and made so much sense – essentially, this is a way to think about the “P/E ratio” for a commodity” that I wondered if it generalized to other commodities. The answer is that it does quite well, although in the case of many commodities we don’t have enough history to fill out a clean curve. No commodities work as well as does gold; I attribute this to the role that gold has historically played in investors’ minds as an inflation hedge. But for example, look at Wheat (I am using data 1970-present).


There is lots of data on agricultural commodities, because we’ve been trading them lots longer. By contrast, Comex Copper only goes back to 1988 or so:


Copper arguably is still somewhat expensive, although over the next ten years we will probably see the lower-right portion of this chart fill in (since we have traded higher prices, but only within the last ten years so we can’t plot the subsequent return).

Now the one I know you’re waiting for: Crude oil. It’s much sloppier (this is 1983-present, by the way), but encouraging in that it suggests from these prices crude oil ought to at least keep up with inflation over the next decade. But do you know anyone who is playing oil for the next decade?


For the sake of space, here is a table of 27 tradable commodities and the best-fit projection for their next 10 years of real returns. Note that most of these fit a logarithmic curve pretty reasonably; Gold is rather the exception in that the historical record is more convex (better expectation from these levels than a pure fit would indicate; see above).


I thought it was worth looking at in aggregate, so the chart below shows the average projected returns (calculated using only the data available at each point) versus the actual subsequent real returns of the S&P GSCI Excess Return index which measures only the return of the front futures contract.


The fit is probably better in reality, because the actual returns are the actual returns of the commodities which were in the index at the time, which kept changing. At the beginning of our series, for example, I am projecting returns for 20 commodities but the 10-year return compares an index that has 20 commodities in 1998 to one that has 26 in 2008. Also, I simply equal-weighted the index while the S&P GSCI is production-weighted. And so on. But the salient point is that investing in spot commodities has been basically not pretty for a while, with negative expected real returns for the spot commodities (again, note that investing in commodity indices adds a collateral return plus an estimate 3-4% rebalancing return over time to these spot returns).

Commodities are, no surprise, cheaper than they have been in a long while. But what is somewhat surprising is that, compared to the first chart in this article, commodities don’t look nearly as cheap. What does that mean?

The first chart in this article compares commodities to the quantity of money; the subsequent charts compare commodities to the price level. In short, the quantity of money is much higher than has historically been consistent with this price level. This makes commodities divided by M2 look much better than commodities divided by the price level. But it merely circles back to what we already knew – that monetary velocity is very low. If money velocity were to return to historical norms, then both of these sets of charts would show a similar story with respect to valuation. The price level would be higher, making the real price of commodities even lower unless they adjusted upwards as well. (This is, in fact, what I expect will eventually happen).

So which method would I tend to favor, to consider relative value in commodities? Probably the one I have detailed here. There is one less step involved. If it turns out that velocity reverts higher, then it is likely that commodities real returns will be better than projected by this method; but this approach ignores that question.

Even so, a projected real return now of -2% to spot commodities, plus a collateral return equal to about 1.9% (the 10-year note rate) and a rebalancing return of 3-4% produces an expected real return of 2.9%-3.9% over the next decade. This is low, and lower than I have been using as my assumption for a while, but it is far higher than the expected real returns available in equities of around 1.2% annualized, and it has upside risk if money velocity does in fact mean-revert.

I will add one final point. This column is never meant to be a “timing” column. I am a value guy, which means I am always seen to be wrong at the time (and often reviled, which goes with the territory of being a contrarian). This says absolutely nothing about what the returns to commodities will be over the next month and very little about returns over the next year. But this analysis is useful for comparing other asset classes on similar long-term horizons, and for using useful projections of expected real returns in asset allocation exercises.

[1] In what follows, I will focus on the expected return to individual spot commodities. But remember that an important part of the expected return to commodity indices is in rebalancing and collateral return. Physical commodities should have a zero (or less) real return over time, but commodity indices still have a significantly positive return.

Call Off the Deflation Warning

January 7, 2015 9 comments

Today’s column is a brief one, as I need to post a correction. Not a correction to my stuff, mind you, but to others.

Pictures like the below have been circulating now for a couple of weeks. This is a chart of the 2-year inflation “breakeven” on Bloomberg, illustrating how a “deflation warning” is sounding as they go negative.


Unfortunately, it ain’t so. I wrote to the authors of the original Bloomberg piece referenced above, and called Bloomberg (more on that later), and figured that when I pointed out that 2-year inflation expectations are nowhere near zero, the story would at least die quietly even if pride prevented a retraction. Unfortunately, that hasn’t happened and other “analysts” and news outlets have picked up the story. So, I need to print a correction for them. Unconventional, I know, but I stand for Truth.

The simple fact is that 2-year inflation expectations have fallen deeply, but remain well above zero. The chart below, also from Bloomberg, shows 2-year inflation swaps over the same period. You will notice that it has fallen mightily but remains at about 0.70%.


It turns out that the difference between the Jan-17 TIPS (which have 2 years to maturity) and the Jan-17 nominal Treasuries that are their comparator bond – taking the difference between real and nominal rates gives you the “breakeven” inflation rate that makes them equivalent investments; thus the name – is also about 0.70%.

So why does Bloomberg say the 2-year breakeven is negative? Well, Bloomberg’s “policy” is to track the April-2016 TIPS as the “2-year” TIPS until the new April-2020 TIPS are auctioned in April At that time, they will roll to using the April-2017 TIPS, which will have two years to maturity, and will use that bond for a year. While I applaud Bloomberg for having a policy, that’s no excuse for a stupid policy. There is no place in this universe where the April-16s are a 2-year note. Not even close. And not the “best we can do.”

In truth, especially for short-dated inflation expectations there is no reason not to use inflation swaps. The 2-year inflation swap is evergreen each day with a new 2-year maturity, and there are no idiosyncrasies (such as the fact that the April issues often trade cheap because of the bad seasonality associated with them, so they will usually understate true inflation expectations if you use them) to worry about.

So the story is false. The market is not discounting two years of deflation. Indeed, the reality is quite a bit different. The chart below (source: Enduring Investments – we know stuff like this) shows the 1-year inflation rate, starting 1 year from now (the 1y1y or 1×2 if you like), derived from CPI swaps. While it has come down substantially since the summer, it is not particularly out of line. In fact, it’s pretty much right where core inflation is, which makes sense: the energy spike lower is not going to continue year after year, which means that once it stops then headline inflation will return to the neighborhood of core…unless there’s a rebound in gasoline, of course. But the point is that the best guess of inflation one year from now has little to do with gasoline.

1y1yswapActually, the even-deeper point is that it is appalling how little general knowledge there is about inflation, and how journalists and even many analysts have scant idea how to get to the real story. (Hint: calling an inflation expert is a good start.)


Get every new post delivered to your Inbox.

Join 1,847 other followers

%d bloggers like this: