Everyone Is Thinking About Inflation Wrongly

|
Includes: DDM, DIA, DOG, DXD, EEH, EPS, EQL, FEX, FWDD, HUSV, IVV, IWL, IWM, JHML, JKD, OTPIX, PSQ, QID, QLD, QQEW, QQQ, QQQE, QQXT, RSP, RWM, RYARX, RYRSX, SCAP, SCHX, SDOW, SDS, SFLA, SH, SMLL, SPDN, SPLX, SPUU, SPXE, SPXL, SPXN, SPXS, SPXT, SPXU, SPXV, SPY, SQQQ, SRTY, SSO, SYE, TNA, TQQQ, TWM, TZA, UDOW, UDPIX, UPRO, URTY, UWM, VFINX, VOO, VTWO, VV
by: Joe Kuefler, CFA

Summary

Inflation is like gravity to long-dated interest rates and thus extremely important for investors to consider.

Academic macroeconomic theories dominate the models used by today’s policy makers, but they consistently come up short due to reliance on unrealistic axioms.

Policy makers around the world have particularly struggled to explain our low inflation environment, despite near zero interest rates, exceptional monetary accommodations, and low unemployment.

I offer a better, more inductive, bottom-up approach to thinking about inflation over the medium and long term.

I’ve never considered myself an economist, certainly not in the classical sense. When I think about the economy I tend to think on a more micro level, about how consumers, individual companies, and industries affect each other and our economy from a bottom-up basis. So, it is with great surprise to myself that I am writing an article on such a macro topic as long-term inflation expectations.

Part of the reason I’m throwing myself into the inflation debate is that I believe most of the commentators and policy makers have little idea what is really driving inflation. There are no shortage of economic models, each of which have worked during one period or another, but none of them consistently. Central banks put into practice academic models created by Nobel laureates that were seduced into relying on unrealistic axioms (e.g. consumers are rational, utility maximizing decision makers with stable preferences) so that their models could grow into beautiful mathematical specimens, fully supported by “proofs.” Unfortunately, they are rarely supported by reality. As a wise man once said, “In theory there is no difference between theory and practice, in practice there is.”

These economists and academics tend to approach the world through deductive thinking, that is the reliance on general axioms combined with logic and mathematics to solve problems. This works great in hard sciences like physics, but poorly in the social sciences where the subjects of study (in this case markets) rely on complex human networks.

I’ve typically approached finance and market forecasting with an inductive thinking process. Inductive thinking relies on looking at specific cases, drawing generalizations from them, using those generalizations to formulate arguments and theses, and ultimately make decisions. Inductive thinking also relies on learning from experience (or the experience of others), utilizing multiple, interdisciplinary mental models (thank you Charlie Munger), and changing models and paradigms on the fly as new information is obtained. There is no formula for this approach and it inherently means dealing with all kinds of ambiguity.

Richard Bookstaber uses a great deductive vs. inductive analogy in his 2017 thought-provoking book “The End of Theory” in describing the hard way and the easy way to catch a baseball. Deductive thinking in economics is like trying to catch a baseball the hard way. The hard way to catch a baseball is to calculate the ball velocity, spin and trajectory out of the pitcher's hand, as well as the mass, speed and angle of the bat at time of contact.

Then, adjusting for gravity, wind speed and atmospheric resistance, run a complex mathematical calculation to determine where to position oneself in the outfield to make the catch. This has widely been the economic forecasting approach of academics and our academic-led central bankers around the world.

The easy way to catch a baseball is to do it the inductive way. That is, to watch the ball come off the bat, fix your gaze on the ball while beginning to run backwards (or forwards) and adjusting your running speed in order to keep the angle of your gaze constant, until you lift your mitt and close it around your ball. This is a much easier, and more intuitive approach. While I may not know exactly where I will be in the outfield when I catch the ball, and I may run into a wall every once and again, I’m going to catch a lot more balls.

So far in my career, this inductive approach to forecasting financial markets has served me well and I will utilize it to establish and support my thesis that the world economies will experience more disinflationary forces than inflationary forces over the next 10-20 years, keeping inflation in check and much lower than most professionals expect.

Debunking a couple of common inflation myths:

  1. Low unemployment results in increasing inflation: The Phillip’s Curve is a seemingly intuitive theory from the 1950s stating that there is an inverse relationship between the unemployment rate and inflation. Macro economists later expanded on this theory claiming that there is a “natural rate” of unemployment, and only unemployment rates below the natural rate spur wage inflation. This is very plausible sounding, but problematic in practice. First of all, no one knows what the natural rate of unemployment is. It is a continual guess. The lower the unemployment rate goes without spurring inflation or wage growth, the lower the estimates for the natural rate go (the economists simply move the goal posts!). Just look at the Federal Reserve’s estimates for the natural rate of employment over the last several years. Remaining at roughly 5% since the early 2000s, the estimates are slowly trending down now that the unemployment rate has plunged through this level without spurring meaningful wage inflation. Expect this to continue until we get a result that confirms their models! There are several reasons why the Phillips Curve may not be working as macro economists expect. For example, technological advances (leading to higher productivity and lower costs) and globalization (increasing the benefits of the division of labor) have nothing to do with the unemployment rate, but greatly affect the price of goods. Demographics also affect unemployment and inflation in ways not captured by this model. Just look at Japan (with their aging population) seemingly barely holding off deflation despite a now 2.3% unemployment rate.
  2. Low interest rates spur economic growth and inflation: It continually surprises me how hard it is for academics and policy makers to kick this paradigm. Ultra low interest rates are the opposite of inflationary; they are disinflationary. Banks make money by making loans; when interest rates are low, and the term structure is flat, banks have less of an incentive to lend because it will be less profitable. Less lending means less money supply and lower economic growth. This shouldn’t be news…Japan is once again a case study with their ultra low growth and inflation the last 20 years despite an ultra low (and now negative) interest rate environment. Same for the world economy during the last 10 years with zero or negative interest rates in developed countries.

30 years of disinflation:

So, if all the conventional wisdom about inflation has been wrong, what has led to the tremendous disinflation we’ve seen over the past 30 years? The biggest two factors are probably rapid technological advances, particularly in information technology, and demographics.

  • Revolution in information technology: The information technology revolution that kicked into high gear in the 90s and gave us the internet and the smartphone has increased our computational power and increased productivity in ways that aren’t being captured well in the economic data, but are captured in the CPI data. Consider Moore’s Law, coined after Gordon Moore’s predication that computing power would double every couple years. It isn’t really a law, but it turned out to be a very good prediction and it has resulted in computing power growing exponentially and at lower prices. My smartphone is hundreds of times more powerful than my first computer, yet it cost me less. This is deflationary and the trend doesn’t appear to be ending anytime soon. The same phenomenon goes for the cost of software, data storage, and data transfer speeds.
  • Demographics: Aging populations are disinflationary, despite putting downward pressure on unemployment rates. As the older age cohorts leave the workforce, the knowledge and experience they’ve gained over the course of their careers leaves the economy and is replaced by younger workers that are fewer in number and less experienced. Again, I use Japan as a case study. Japan’s percentage of workers over 65 years old has increased from under 8% in 1975 to 27% in 2017. I believe this is a major factor that has kept the country’s inflation and economic growth so low over the past 20 years. Japanese inflation has bounced around zero since 1998, and GDP has grown only 3.7% in the same period (that’s 3.7% total, not annually!). In the U.S., even though the millennial cohort is bigger than the boomers, we can experience a disinflationary environment all the while having low unemployment because the workers leaving the workforce are much more experienced and productive than the younger millennials entering. The US, however, is in a better position than much of the rest of the world because the millennial generation is larger than the boomers and will continue to accumulate knowledge and skills, eventually contributing more than what is lost by the retiring boomers, thus someday turning into a tailwind for productivity and inflation.
  • Globalization: The world economy has experienced two great waves of globalization in the past 30 years. First with the fall of the Berlin Wall, opening up the Soviet territories to the world economy during the 1990s, and second with China opening for business with the rest of the world. Naturally, production moved to where it was best cost, driving down pricing on goods around the world while dramatically increasing the standard of living for billions of people. It should be noted that globalization has had a less impactful affect on the inflation of services. Services tend to be more localized (i.e. less easily outsourced), and as such have not experienced the full disinflationary effects of globalization.

(Source: Bureau of Economic Analysis)

So, with the conventional theories debunked, and an understanding of what has driven the last 30 years of disinflation, let’s discuss what will impact the next 10-20 years.

The Disinflationary

  • Cloud Computing: Instead of companies purchasing and managing their own servers and software, they can now rent exactly what they need. In-house servers must be underutilized by default in order to allow for growth and spikes in usage. Not only is renting server space more cost effective from a pure division of labor standpoint, it greatly reduces this wasted capacity. In aggregate, this wasted capacity is tremendous and the continued migration of these servers to the cloud will eliminate the extra costs associated with them.
  • Energy: Cheap batteries and technological improvements in electrical generation, such as solar, will disrupt the energy industry as we know it. Similar to private servers, but on a much grander scale, our electrical generation apparatus requires excess, idle capacity to meet the highest peaks in demand, otherwise serious problems and outages ensue. This results in “peaker” plants being built that hardly ever run, or are drastically under-utilized. When battery technology improves sufficiently, industrial-sized battery plants (referred to as power packs) will be built to replace peakers for these peaks in demand. These are integrated, modular systems of batteries that can be scaled up and down. While this is already happening now, in most markets, it isn’t clearly the lowest cost solution, but the costs continue to drop. While I have no prediction on when this occurs, there will be a critical mass moment where the economics drive a cascading overhaul to the energy industry, disrupting the energy cost landscape. The result will be a deflationary force throughout the world economy. If favorable industrial battery economics coincide with sufficient advances in solar generating technology, we could see lollapalooza effects (another hat tip to Charlie Munger in pointing out the power of lollapalooza effects).
  • Electric Vehicles (EVs): EV economics are very dependent on the technological advances of batteries as just discussed, but once the inflection point is hit, growth will be exponential. Within a few years it is widely expected that electric vehicles with over 250 miles of range will be less than $30,000 (compared to the $33,000 average price tag for a new internal combustion engine (ICE) vehicle). EVs have only dozens of moving parts compared to the thousands found in ICE vehicles. This reduces manufacturing complexity, reduces maintenance costs and increases vehicle lifetime, from roughly 200K miles today to perhaps over 1M miles or more for EVs. Fives times more vehicle for less money than an ICE vehicle is certainly disinflationary.
  • Autonomous Transportation: The granddaddy of all capacity underutilization is in transportation. Our cars and trucks sit idle in driveways and parking lots most of the time. Ride-hailing fleets of autonomous vehicles will, over time, reduce car ownership purely on an economic superiority basis. Why would a family voluntarily choose to make a huge capital investment that mostly sits idle if they have a better choice? Autonomous vehicles will make the economics of these ride-hailing fleets, or transportation-as-a-service (TaaS), incredibly more cost effective than private ownership. Not only will the capacity of these assets be better utilized (they won’t sit idle in garages and parking lots all day), but insurance costs per vehicle will plummet (because they will be safer), with the savings passed onto the consumer. Again, a paradigm shift this big may not happen overnight, but once the inflection point is hit, things will take off exponentially. Just look at how quickly smartphones changed our lives. People have a tendency to underestimate things that are non-linear, and the adoption rate of TaaS once the technology is viable will most definitely be non-linear. I don’t know when the inflection point will be, but I know I’ll probably underestimate the speed of adoption once it hits. For more specific predictions on timing, see Tony Selba’s thought-provoking presentation on Clean Disruption of Energy and Transportation on YouTube (I promise you won’t be disappointed). With Alphabet’s (NASDAQ:GOOGL) (NASDAQ:GOOG) Waymo already ordering over 60K autonomous Chrysler mini vans with plans to open an autonomous ride sharing service to the public in Arizona in 2018 (that’s this year folks!), this reality may be closer than most realize. (As a side note, I still believe Alphabet shareholders are getting a free option on the future success of Waymo. See my 2017 article, not currently behind the Seeking Alpha paywall, for my take on the attractiveness of Alphabet as an investment).
  • The digital “free” model: Google, Facebook (NASDAQ:FB), PayPal (NASDAQ:PYPL), AliBaba (NYSE:BABA), RobinHood, Tencent (OTCPK:TCEHY), Dropbox (NASDAQ:DBX) and the like provide us services for free that we never would have dreamed about a decade ago. We are in the “there’s a free app for that” world. In exchange we of course provide our data, but it is deflationary in the way that we calculate inflation… the value of the data we give isn’t inputted into the equation, driving down consumer prices.
  • Online retailing: Competition from online retailers and the ease of comparison shopping has clearly driven down retail prices. Just the change in mix of online retail spending vs. brick and motor will be disinflationary since online has a lower cost structure. The most expensive part of the online model is the freight, which will eventually be disrupted by autonomous freight fleets.
  • AI and Robotics: Automation is deflationary if it results in lower costs. Even more so if it results in higher unemployment (although I’m not convinced it will). As Ed Yardini put it in his April 2013 commentary, as robots proliferate “manufacturing productivity will soar. Labor costs and the prices of all goods manufactured on an assembly line will plunge. Such productivity-led deflation would boost the purchasing power of all workers around the world.”

The Inflationary

While the disinflationary forces are numerous, I’m not suggesting I expect outright deflation in the next 10-20 years. Rather, these forces are off-setting other inflationary forces, keeping headline inflation in check. That said, it is worth briefly discussing some inflationary forces we may encounter.

  • Accelerating economic expansion: This is part of the business cycle, which will always wax and wane. Since we are in the later stages of a world-wide expansion, there are many interrelated factors that are inflationary. You can lump in inflationary forces like credit booms, and leveraging into the business cycle… these are more transitory than secular.
  • Government deficit spending: Deficit spending can help boost economic activity and inflation. I won’t try to forecast any government’s deficit spending. Once a government becomes tapped out, it can go one of two ways… spending can be pulled back, which is deflationary, or money can be printed, which is inflationary.
  • Protectionism: Anything that leads to less globalization and less division of labor is likely to be inflationary. Protectionist policies or trade wars would create inflationary pressure, particular on goods. We are currently in an environment where policy mistakes could be made on this front, leading to higher inflation.
  • Health care: U.S. healthcare expenditures have risen from 5% of GDP in 1960 to 18% of GDP in 2016. I certainly hope that the Berkshire Hathaway (NYSE:BRK.A) (NYSE:BRK.B), JPMorgan (NYSE:JPM), and Amazon (NASDAQ:AMZN) joint-venture to lower healthcare costs and increase transparency will be successful, but the system is set up in a way that industry stakeholders have powerful incentives to resist change. For the time being, healthcare looks to continue to be an inflationary force. The aging world populations will also keep demand for health care high.
  • Low unemployment: While low unemployment doesn’t mean we will get rising inflation, it isn’t disinflationary either. If labor markets become competitive enough, wages will rise, at least wages per unit of productivity. Still, without a secular demographic or immigration shift driving it, this should probably be lumped in with the business cycle.
  • Real (world) war: Prices have a history of rising sharply during war and falling sharply as soon as peace ensues. By default, war creates trade barriers, reducing globalization, which is inflationary. WWIII would lead to more inflation.

Conclusion:

Inflation is a difficult (and dangerous, if you are concerned about your ego) thing to forecast. The combination of monetary policy, fiscal policy, and microeconomic developments like the ones discussed in this article interact in unpredictable ways. That said, I hope this article broadened your understanding of the secular inflationary forces at work now and potentially in the future, and at a minimum made you think differently than the macro economist academics and our central bankers around the world.

Disclosure: I am/we are long GOOGL.

I wrote this article myself, and it expresses my own opinions. I am not receiving compensation for it (other than from Seeking Alpha). I have no business relationship with any company whose stock is mentioned in this article.