This article focuses on long-term investment cycles, and specifically the shortages in energy and infrastructure that have become prevalent this year in part due to years of prior underinvestment, which have been further exacerbated by both war and drought.
Analyzing Investment Cycles
Each business cycle has a general “investment theme” about it.
To quantify that from a high level, this chart shows the performance ratio of value stocks vs growth stocks, with the gray bars representing recessions. When the line is going up, it means value stocks are outperforming, and when the line is going down, it means growth stocks are outperforming. The biggest pivots have generally occurred right before or after recessions:
This makes sense intuitively. Each cycle makes use of new technologies, renewed access to credit, some area of existing underinvestment, and/or sometimes new economic inflection points of specific countries, and continues until it runs hot and reaches some level of over-investment and excess valuations, which then results in a period of deleveraging and rightsizing in those industries and regions to clear out those excesses.
And in a broad sense, we can think of these investment cycles as oscillating between the value-oriented “hard asset” category (commodities, real estate, infrastructure, manufacturing), and the “growth/tech” category (fast-growing companies based on new technologies, during periods of relatively cheap and abundant commodities).
We can also quantify this by looking at the ratio of the Nasdaq 100 to oil in linear and log charts, with the latter being a lot more clear:
Or by looking at Stifel’s chart on commodity cycles, which I included back in my August 2021 newsletter, and that has indeed started to turn back up since then:
The 1960s had the “Nifty Fifty”, referring to several dozen growth stocks that could seemingly do no wrong and that investors would pay any price for. They were well-known companies like Disney (DIS), Coca-Cola (KO), and Xerox (XRX). The problem, of course, is that investors did overpay for these stocks. Oil had been flat in dollar terms for about two decades. Low commodity prices and low interest rates played a big role in allowing for company valuations to reach excessive levels.
The 1970s saw the opposite; US oil production structurally peaked in 1970, and the dollar’s backing by gold was removed, resulting in a significant period of inflation and commodity scarcity, as the US had to begin importing a lot of Middle Eastern oil. This was also a period of peak global population growth in year-over-year terms, meaning that demand was persistently strong. Avoiding overvalued growth stocks and focusing on value/commodity stocks was the way to go. Eventually, along with the petrodollar agreement for relative stability, producers around the world brought plenty of commodity supplies onto the market. Value outperformance extended into the 1980s, as financial companies took over as relative leaders after the commodity producers stalled, in part thanks to high positive real interest rates and increasing financialization of the economy.
During the 1990s, the tides turned again. After a long period of commodity supplies coming online, most commodities became relatively abundant and stable. The investment trend once again heavily shifted back into growth stocks, this time associated with the Internet. This led to rather infamous excess and malinvestment, which was cleared out in the early 2000s. Even during the height of the Dotcom Bubble, US real estate valuations and many value sector equity valuations were unremarkable; almost all of the excess was located in technology and consumer-oriented companies.
After that bubble unwound, the bulk of the 2000s decade was once again focused on the rise of value stocks and hard assets. In the US, it was mainly about the real estate market, along with the banking sector that financed it and the commodities that were used to build it. Globally, the world saw the rapid economic rise of China and the rest of the “BRICS” nations as the dollar weakened from its recent highs. But this, too, eventually ran into excess. The US housing market and banking sector become overleveraged and highly speculative. Emerging markets equities reached extremely high valuations, and their economic growth rates slowed down, in part due to a strengthening dollar.
2010s, Into 2020
With the hard asset and value category overdone to excess, the cycle yet again shifted back towards the growth-oriented category throughout the 2010s decade and into the early years of the 2020s decade. The proliferation of smart phones and fast mobile internet ushered in a new group of corporate titans, and we saw the rise of app-based companies, social media, ecommerce, cloud computing, and software as a service. In some ways, the transformative tech ideas of the 1990s became manifest in the 2010s. By the end of the decade and especially during the stimulus-fueled lockdowns in 2020, this cycle reached excessive levels just like all of the others have done. SaaS companies could IPO with almost any valuation, and investors seemingly had no concern for profits in the first decade or two of a tech business’s existence. Many of these companies instead relied on constant share dilution to pay their expenses, which worked as long as money kept pouring in and propping up the valuations.
My view for a while now is that we are once again running into a longer-term bullish trend of commodities, real assets, and infrastructure, all of which I expect to be a big theme (albeit a volatile one) throughout the 2020s decade due to a period of underinvestment. The world has invested quite heavily into technology, but not much in natural resources and heavy industry lately, and as the past year has shown us, we’re due for another shift of focus.
While policymakers may be able to suppress demand and tame energy prices for periods of time, energy security and energy-related inflation is likely to be a recurring issue in the years ahead until more capital expenditures occur in that area. We have yet to see a renewed capex cycle to fix the problem on a longer-term basis, although I think it will come in the years ahead.
Cheap energy and cheap labor were commonplace over the past decade, which I think for many governments, companies, and consumers, led to a sense of complacency.
We all tend to suffer from recency bias, where we assume that our recent experiences are the “new normal” and that we can extrapolate them indefinitely into the future.
But as past commodity capex cycles in particular have shown, that’s usually not the case. The situation can change quickly. Periods of commodity and commodity-infrastructure oversupply eventually transform into shortages, due to lack of ongoing investment to replenish depletion and depreciation. Plus, the possibility for war is always a wild card that can derail everyone’s plans in the global sense.
Limitations On Cheap Energy
The 2010s decade, despite being mostly focused on software/technology outperformance, also saw the rise of US shale oil.
Zero interest rates allowed for plenty of shale oil production to come online that was often not free cash flow positive. Pensions and other pools of capital would fund them with cheap debt, and plenty of oil was produced, but investors for the most part did not see good returns.
I often use the EOG Resources (EOG) free cash flow chart as an example of this (and they were one of the better-managed ones):
This allowed the world to, at least for quite a while, avoid the “peak oil” problem that some had been predicting. While conventional oil sources indeed ran into limitations roughly in line with the peak oil expectations, these unconventional oil sources allowed us to break those expectations. Here is a long-term chart of US oil production, where shale or “tight” oil was able to dramatically break out from the expected decline:
The result of this spike in US production was a period from 2008 to 2020 where commodity prices stagnated and fell, mostly across the board.
This was a huge disinflationary force, which could offset some of the inflationary fiscal and monetary policies of the US, Europe, Japan, and elsewhere.
However, without new investment, eventually those output levels started to stall and shrink. And shale oil output in particular tends to decline quickly without sufficient new investment. The mistake that many analysts and policymakers made in recent years was to assume that this 2010s period of energy oversupply was normal and sustainable forever, when in reality it was not, largely due to being free cash flow negative in reality, and requiring much higher capital expenditures at profitable levels if it is to continue.
I think the 2020-2022 energy/commodity bull run still has higher and longer to go before it is done. It’ll be volatile along the way, but until a lot more supply comes online globally, I don’t think it is over yet.
OPEC+ countries have nearly reached output capacity, and it will require capital expenditures to increase their production limits. US shale oil is estimated to be able to grow into the early 2030s, with several million more barrels of daily oil production possible compared to current levels, but not without significant investment (and likely profitable investment, with free cash flows, since the same investors won’t make the same mistakes of the 2010s). Africa also has a lot of potential to bring on new oil and gas supplies throughout this decade, but also with considerable investment. Europe has domestic shale resources that they could more aggressively try to tap into if energy prices remain sharply elevated.
Limitations On Cheap Labor
The 1990s through to the present saw a rapid period of globalization. The opening of the Chinese economy, and the fall of the Soviet Union and subsequent opening of many of its prior member states, connected a large number of impoverished people in those regions to global capital markets and developed market consumers.
For large corporations in the US and other developed markets, this new labor pool represented an increasingly viable alternative to their own domestic labor pools. Being in more impoverished starting economic conditions, this emerging pool of labor would work hard and for cheaper than their developed market peers. Combined with certain governments (specifically China’s and some others) that would maintain stability and organization in their jurisdictions, this became enticing to many global corporations. Even those that didn’t find it enticing, would find themselves undercut in terms of price by their competition that did find it enticing, and would have to adapt.
As a result, many developed countries shifted a portion of their labor-intensive and energy-intensive industries to emerging markets, with China taking center stage in that regard. These countries could build infrastructure faster and cheaper, and put people to work to supply goods to the world at low prices, and without the likely prospect of cyclical revolutions and asset seizures that would occur in some other parts of the developing world. Corporations and consumers would turn a blind eye towards authoritarianism if it meant cheap sneakers.
These countries, and once again China especially, also took on large environmental burdens for the rest of the world in order to increase their economic clout. They burned coal and developed rare-earth metals at a scale that was noticeable for its cities and landscape- tough smog, poisoned water, and damaged landscapes, due to the sheer concentration of this production and the speed with which it was developed. This trade-off was not unlike what the United States made in the early 1900s.
After so much growth, it is difficult to overstate how much electricity China currently generates. Their TWh/year numbers dwarf every other country including the United States.
China accounts for over a quarter of all global electricity production, and they use a big chunk of that to manufacture things for the rest of the world, rather than merely for their own domestic consumption.
However, this growth engine is potentially running out of steam. Multiple estimates suggest that China’s population either already peaked, or is in the process of peaking.
In addition to workforce limits, China has been running into some physical constraints as it relates to their power generation and industrial capacity; their electricity.
One thing that many people do not realize is that most types of electricity production require a lot of water. Hydroelectric dams obviously require a lot of it, but also coal plants, nuclear plants, and other types of facilities use water for cooling and/or burning. Power generation is the second biggest use case for water globally, after agriculture. And of course, factories are big consumers of electricity.
China, however, has been facing years of increasingly acute water shortages, which have been affecting the country’s ability to reliably generate the power it needs. There is a global drought occurring (for North America, Europe, China, and elsewhere), so it is not just China’s problem, but China’s economic importance for the world, and the level to which it has pushed its electrical capacity relative to its land area, are both significant factors.
Also of note is the fact that China makes 80% of the world’s solar panels. The process mainly involves using hydrocarbons to extract the resources and manufacture the panels, and then ship them to the rest of the world.
Any constraints that China runs into, whether it is water or demographics or geopolitics, trickle out into the rest of the world in more ways than most of us realize. The growth of China has given the world multiple disinflationary trends, and if these various limitations persist and result in a slower-growing China with more constrained industrial capacity, it is likely to shift some of these pressures back to the rest of the world in the form of shortages and price inflation.
The bright side or bull case here, is that if some of these drought conditions ease in 2023 or 2024 compared to this year, it would help the power systems that are being unusually stressed this year.
Section Summary: It’s Global
Europe presents an extreme example of what is happening globally at the current time. It began in late 2021, but got worse during 2022. This is what they are paying for natural gas:
Their natural gas prices, and by extension their electricity prices, have become completely unworkable. They’re correcting down now from an extreme level, but some combination of demand destruction and new supply arrivals are needed to more persistently break this price spike:
A key reason for this has to do with the lesser fungibility of natural gas relative to other energy sources. Commodities like oil and coal are pretty easy to ship globally; their transportation costs are a small part of their production cost, and the infrastructure to move them around is pretty cheap. However, natural gas is harder to transport; it either needs to flow through expensive pipelines, or it needs to be super-cooled into a liquid (requiring very expensive facilities), shipped in specialized vessels that can maintain that liquid state (of which there are a limited number), and then turned back into a gas at the destination (also requiring very expensive facilities). For that reason, if there are natural gas shortages in a specific region, they are slower and more costly to arbitrage and fix, because it’s much harder to take natural gas from somewhere cheaper and bring it over to where it is expensive, relative to oil or coal.
The extreme drought has then exacerbated the problem. Low river levels have made it harder to transport commodities and harder to cool nuclear reactors. Europe really couldn’t catch a break this year.
Some German industry has shut down out of necessity. Consumers are suffering from high energy bills, with some countries subsidizing their populace more than others, but often at the price of printed money and even more inflation down the line.
This leaks out into the rest of the world. Europe, being rather rich in a global sense, can pay high prices for marginal liquified natural gas “LNG” imports above their normal baseline amounts to offset their pipeline shortages from Russia. A significant amount of LNG ships that were originally destined for developing countries throughout Asia, including many with existing contracts, have shifted to Europe where much higher prices are available. This isn’t enough to fix the problem due to the limited capacity of LNG export terminals and ships that are available, but it is a part of the supply response to take the edge off, nonetheless. The dark result of this is that poor energy planning and energy diversification by Europe, combined with war from Russia, has resulted in acute energy shortages for some of the most impoverished people in developing countries.
Europe can also turn to other energy sources. As natural gas prices reach the energy equivalent of $500+ per barrel of oil, certain types of energy demand can shift towards firewood, coal, heating oil, or diesel fuel, which are cheaper in comparison. They then begin importing some these things in higher-than-normal volumes, which can result in other countries (even allies) applying protectionist policies in response to protect their own markets from shortages and inflation.
The Biden administration is effectively asking refiners to prioritize American consumers over maximizing profits by supplying fuel-starved Europe, which is facing an unprecedented energy crunch after the invasion of Ukraine triggered US sanctions on Russian oil supply.
An energy shortage anywhere, basically becomes an energy shortage everywhere. The world does not currently have the right mix of energy production and energy infrastructure/distribution in order to meet demand. In addition, for the first time in decades, there are pretty serious labor shortages as well.
Final Thoughts: Illusionary Wealth
This chart shows the ratio of total US household net worth to GDP in blue on the left axis, and short-term interest rates in red on the right axis:
For the past four decades, we have enjoyed a structural decline in interest rates, which has allowed us to price most assets such as stocks and real estate at higher valuations.
After all, if 1-year Treasuries pay you 10-15% per year to own them as they did in the late 1970s and early 1980s, then that creates a very high hurdle to own more volatile assets like equities, and so you would demand extremely low equity valuations to have potential huge returns to compensate for buying them rather than the 10-15% high yield safe Treasuries.
In contrast, if interest rates go all the way down to zero or nearly so, then there is very little hurdle rate for owning stocks and bonds, and so most investors are willing to pay much higher valuations with expectations of pretty low returns, because there is no yield from owning cash or Treasuries.
There were other factors too. Various investment plans and low-cost online brokers gave people greater access to the stock market, for example. But structurally declining interest rates and a rapid rate of globalization were likely the two biggest factors for this, as far as my analysis is concerned.
Here’s another way I like to visualize it. This chart shows household net worth in blue (just under $150 trillion), and shows GDP multiplied by 4x in red (just under $100 trillion). There are two versions of the chart below, one in linear and one in log form, because readers have different preferences for how they visualize this type of exponential data:
As we can see, household net worth historically spent most of its time under 4x of GDP. It rose above it during the 2000 dotcom bubble, rose above it even more so in the 2007 housing bubble, and then during the 2020/2021 stimulus bubble it reached over 6x.
This 1990s-2021 period of structurally high net worth relative to GDP coincided with falling interest rates, and perhaps just as importantly, rapid globalization. By keeping costs down, including domestic wages, asset valuations could pump up quite high.
Roughly speaking, total household net worth would need to fall by about a third in order to get back down to equaling 4x of GDP. Alternatively, if nominal GDP were to grow by 6% per year on average for the next seven years (including a sizable inflation component) while household net worth was to remain completely flat, it would also bring the ratio back to 4x. I’m not saying either of those two scenarios will happen or has to happen, but I think understanding both of those scenarios is a useful thing to do in terms of scenario modeling.
The problem is that due to the way economic cycles work, some of this wealth is illusionary. When we have large amounts of stored-up wealth, such as large stock portfolios, expensive homes, plenty of cash, and maybe some alternative assets as well, it represents a large implied claim on the world’s consumable resources. Each of us, when we look at our wealth, imagine that if we were so inclined, we could start drawing down that wealth by selling assets and using the proceeds to consume more. We could go on more vacations, we could go to more restaurants, we could buy more clothes and cars and other consumables, etc. Or we could retire early, and then draw down the wealth gradually for more normal spending.
However, if too many of us were to do that together, we would quickly find the prices of all of those consumables starting to rise, and perhaps quite sharply. There’s only so much airline capacity, hotel capacity, manufacturing capacity, energy production capacity, logistics infrastructure, and so forth. Therefore, in the unlikely event that we ever decided to all increase our consumption at the same time, or reduce our workload and live off some of our wealth, we would quickly find prices rising rapidly, and we would be able to buy a lot less with our wealth than our initial calculations suggested. Any one of us drawing down our wealth alone to buy more stuff or work less, is very different than a large number of us at the same time trying to do that.
The Translation of Financial Wealth To Real Consumption
The extremes inform the means, so when we bring this back to reality, what it means is that there are real-world constraints that affect what our wealth means in terms of actual consumptive power. And it can change very rapidly based on shifts in the supply/demand imbalances of energy, other commodities, and infrastructure capacity. And wildcards like war and global-scale droughts.
If you have a million dollars in assets, while oil is $40/barrel and there are no shortages of car production or globalized manufacturing or service labor, then your money will go a lot further in terms of consumption than a world where oil is suddenly $140/barrel and various electricity limitations or labor limitations or logistics limitations result in shortages and/or higher prices for much of the experiences and consumables you were intending to buy. Even if your wealth jumped 20% to $1.2 million in this environment partially due to inflation, you would probably find yourself with less consumptive power.
During disinflationary periods, where there is no significant shortage of natural resources or manufacturing capacity or infrastructure or labor, our net worth relative to GDP tends to be higher. During inflationary periods, where we do run into significant shortages of one or many of those things, then prices tend to go up, corporate margins tend to decline, interest rates go up, and thus net worth relative to total economic output declines.
Whenever household net worth reaches an unusually high ratio relative to economic output, then it is probably in part because investors and other economic participants are overestimating the spare capacity for things like energy and manufacturing, and pricing forward inflation too low. In those periods, most economic participants assume no geopolitical shocks for the foreseeable future that would add frictions to global supply chains, and assume that our prior investment into energy supplies and other commodity supplies was sufficient for the foreseeable future.
With the rapid rate of globalization likely behind us, and with more energy supply limitations going forward, and with the zero bound of interest rates already having been hit (and even mildly negative for some countries), I think inflation and interest rates are generally set to stop going down structurally, and to start going sideways or up structurally, for quite a while.
If we suppose a mild case of interest rates going choppy and sideways for the next decade or two, oscillating between 0% and 5%, then that represents a major trend change compared to the prior four decades of structurally declining interest rates.
With that being the case, my expectation is that this 6x level of household net worth to GDP that we saw at the end of last year and the start of this year, was the high water mark for a long time. Whether net worth falls or stays flat (correcting relative to GDP mainly in terms of either price or time) will depend in part on policymaker decisions and various geopolitical events. In my view, the least likely outcome is for the prior rising trend to continue, with US household net worth reaching 7x, 8x, or 9x of US GDP.
That doesn’t mean asset prices can’t go higher nominally, but if it’s correct, it means that asset prices won’t rise much more quickly than nominal GDP like they have been.
The Inability To “Normalize” Policy
The US Federal Reserve is currently trying to reign in demand to quell inflation, since it can’t do much about the supply-side limitations. This can work for a period of time, by way of causing recessionary-like conditions to slow down inflation, but the problem is that unless we see signs of new energy production and other solutions coming online, then the fix is likely to be short-lived.
The yield curve is already inverted, and flashing yellow. Whenever a recession becomes problematic enough, or liquidity conditions in the Treasury market become severe enough, policymakers may need to pivot or pause their tightening policy. And when they do, those supply-side constraints will still be there, fundamentally unresolved.
In other words, we can have no growth and less inflation, or some growth and a lot of inflation, but if we want ample growth with low inflation, then fixing the energy supply side is necessary.
Compounding that issue is the fact that, at 370% total (public and private) debt to GDP, interest rates that are persistently higher than inflation would be nearly impossible for policymakers to maintain. They could hit those levels for periods of time, but they’re very unlikely to be able to reach positive real yields and maintain them at those levels persistently, which would represent actual “normalization” of policy.
These high levels of debt are commonplace throughout the developed world, and extending into China as well. The Bank of Japan has given up on trying to get to positive real rates when its sovereign debt is 250% of GDP. The European Central Bank has nearly given up trying to get to positive real yields for its member states, especially for example with Italy when its sovereign debt is 150% of GDP at a time when they’re experiencing a crisis in energy supply.
I continue to be preferential to real assets in my investment mix when looking out over multiple years. Energy pipelines, energy producers, certain types of real estate with fixed-rate debt attached, inexpensive healthcare stocks, cheap stocks with good dividends, select emerging market equities, and allocations to both gold and bitcoin. Some of those assets have paid off earlier this year, and other ones I expect to continue to have some near-term turbulence until the Federal Reserve reaches their limits on their current monetary tightening cycle.
I also like to maintain some cash and cash-equivalents for rebalancing purposes, for moments where we get disorderly and illiquid market conditions.