Another interesting voice on the topic is growth fund manager Cathie Wood, who likewise foresees very elevated rates of productivity growth on the horizon, far beyond historical experience. She has stated that stocks linked to these technology trends have the potential to appreciate in value at a rate of 40% per year within the next decade.
Wood’s extremely optimistic projections for tech stocks are reminiscent of Michael Saylor’s bold claims that Bitcoin will reach or even exceed $13 million (from current levels around $60,000) by 2045, which we examined here.
Wood, like Saylor, who personally has billions of dollars of Bitcoin exposure, is no doubt “talking her own book.” As a fund manager who earns fees by getting people to invest in early-stage growth stocks, she stands to benefit personally from the perception that early-stage growth stocks will skyrocket.
Everyone should therefore take Cathie Woods’ predictions with a grain of salt. But the mere fact that someone’s viewpoint aligns with his or her financial self-interest does not make the viewpoint false.
The more extreme presentations of blue sky scenarios are at a minimum interesting in that they provide a full-fledged explanation of all the “pros” of a certain investment. Like listening to a lawyer speak in defense of a client, we can then critically analyze all the arguments and try to figure out the truth.
Extremely bullish forecasts like those of Saylor and Wood also do not have to be precisely correct. The real issue is whether or not they are directionally correct. If Bitcoin were to compound at 10% or 15% per year for the next 20 years, rather than 30%, an investment today would be more than justified.
Cognitive mistakes
While financial self-interest is a form of bias, there are also a number of cognitive biases, as described by behavioral economists, that prevent people from accepting scenarios that imply dramatic change. These include recency bias, which refers to the tendency of human beings to think that patterns or events that occurred more recently are more likely to repeat in the future.
The idea of a major unprecedented step-up in productivity growth (not to mention fundamentally transformative changes in how civilization operates) naturally seems “crazy” because it requires us to accept the idea that we are about to experience radical change.
The human mind has a tendency to assume the status quo until proven otherwise (“status quo bias” is also a recognized cognitive bias). This is perhaps a helpful default tendency in that it prevents us from falling prey to every exaggerated or apocalyptic claim that may come our way.
Every now and then, however, the status quo does break.
Hedge fund manager John Paulson became a multi-billionaire because he predicted such a break. Paulson made billions betting against subprime mortgages through various levered instruments.
These investments were, in retrospect, wildly mispriced because they heavily discounted the possibility of what was at the time seen as an extreme outcome. This created an opportunity for Paulson and others to earn outsized returns, like winning a bet at the track on a horse with 50:1 odds.
(John Paulson is incidentally in the mix to be named Treasury Secretary if Trump wins in November and, along with Elon Musk, was by Trump’s side at the recent rally in Butler, Pennsylvania. He has also been a very large investor in gold and gold-mining stocks. With gold having compounded at 12% annualized over the past 5 years, this trade is starting to look pretty smart as well.)
In the 2008 time frame, we witnessed firsthand how many investors downplayed credit risks within the banking system despite many smart, well-informed people pointing out that the sky would soon be falling.
The alarmists may have seemed crazy and extreme at the time, but they were right. The sky did indeed fall.
Financial melt-up?
Severe ruptures with the status quo can turn into great investment opportunities because they are not already priced in. The 2008 financial crisis was not captured in asset prices. The ones who saw it coming made a fortune, while most people brushed aside the risk and sustained major losses.
What if asset prices today fail to capture a radical change in circumstances that will instead lead to tremendous upside?
Among the predictions of Vinod Khosla is that the AI-driven productivity improvements coming our way will be deflationary. This is a critical observation, and the potential consequences of this important insight may be counterintuitive.
The reason productivity gains represent a deflationary force is that these gains, by definition, make the production of goods and services less expensive.
Nowhere is the deflationary impact of technology more visible than in consumer electronics. When the first high definition television sets came out around the turn of the century, they cost thousands of dollars and could not be lifted by one person. Today, a much larger flat screen television with far better picture quality can be purchased for a few hundred dollars and lifted with one hand.
Why AI is deflationary
AI has the potential to bring prices down by eliminating workers from the production process. One example we recently heard relates to what is known as “agentic AI", where you simply instruct the computer to accomplish a certain task and it does so more or less independently.
To build a website or e-commerce platform from scratch, for instance, is now a complicated process that requires a substantial amount of coding and integration. It can become quite expensive and involve hundreds of hours of skilled labor. Ultimately, however, it is a computer programming exercise.
What if one could simply ask AI, using plain language, to build an internet platform with a certain set of specifications, similar to how we might ask ChatGPT to write a poem on a specific topic? Within moments, a viable website could perhaps be produced, rendering web developers useless.
In the early stages, AI-generated websites may be clunky and flawed, but as AI continuously improves, such websites could conceivably be better than anything a human is capable of making—and at a fraction of the cost.
The world Vinod Khosla is describing is a world in which there is shrinking demand for human cognitive ability and production costs plunge. This suggests reduced demand for workers. This also suggests some combination of lower final prices to consumers and higher corporate profits.
Deflation is unacceptable
The problem with deflation is that modern economies, built on debt and fiat money systems, cannot handle it. The Federal Reserve has a 2% inflation target (rather than zero, or -2%) because our system requires steady inflationary drift.
Deflation is dangerous in that it discourages spending and investment. Why part with your money today if you can buy more with it tomorrow? As demand drops in the economy, prices fall even further. This is the “deflationary spiral” that economists fear most of all.
Our federal government currently also has some $35 trillion of debt (and growing). In a deflationary scenario, these liabilities become more valuable in real terms. To avoid financial catastrophe, our government needs them to become less valuable.
The U.S. central banking system, organized around the Federal Reserve, is also mandated to maximize employment. When unemployment starts to tick up, easier monetary policy is implemented, as we observed with the recent 50 basis point interest rate hike.
High levels of joblessness are politically unacceptable and can create all kinds of social dangers. In periods of declining employment, fiscal policy becomes more stimulative, in some cases automatically through mechanisms like unemployment insurance. Deficit spending occurs.
We may indeed be on the cusp of massively deflationary productivity increases, which have the potential to translate into broad-based prosperity. But, in our view, it is highly unlikely that these productivity gains will manifest as actual deflation.
Instead, monetary and fiscal policy will likely become highly aggressive to offset these deflationary forces. The productivity gains will therefore manifest as potentially large nominal increases in certain asset prices.
In an actual deflationary scenario, the small percentage of the population that has a lot of money tends to benefit (in the form of stronger purchasing power) while those who do not have much money find it more difficult to earn money.
The vast majority of voters in the United States do not have much wealth and rely on income from either an employer or the government for their survival. According to 2022 Federal Reserve data, the mean net worth of an American household exceeds $1 million, but the median net worth is less than $200,000.
The concentration of wealth in the United States is not necessarily high relative to most other countries, but it is still quite severe. There is unsurprisingly an enormous age component as well, with households under age 35 averaging less than $40,000 in median net worth. Households in the 65-74 year old bracket exceed $400,000 in median net worth.
Our political system is under constant pressure to keep money flowing to the broad population, either through robust labor markets or government transfer payments.
Even wealthy retirees, who may benefit from greater real purchasing power in a deflationary environment, are unlikely to support policies that align with deflation. They may have children and grandchildren whose interests they prioritize. They also are unlikely to favor policies that can lead to social unrest, crime and even political violence, which high rates of unemployment historically produce.
All roads lead to monetary debasement
To offset the deflationary impacts of a potential sustained jump in productivity growth, our government will likely (and probably appropriately) pursue monetary and fiscal policies that halt deflation and redirect money to those who do not have it.
More extreme measures like universal basic income are conceivable, but much of this redistribution could also be accomplished through less direct means, such as expansion of government payrolls or subsidies for higher education. Through a variety of potential fiscal and monetary mechanisms, the real value of dollars can be diluted in order to prevent newfound productivity gains from benefiting only the wealthy.
In a sustained deflation scenario (no government interference), a rational investment strategy may be to hold cash or bonds. Dollars will simply gain in purchasing power. But in a scenario in which deflationary forces are combatted with aggressive monetary and fiscal policy, a different approach is required.
An investor anticipating a tech-driven pick-up in productivity should have two overriding goals: (1) benefit from the trends in innovation; and (2) benefit from the potentially substantial monetary debasement that will need to occur to prevent U.S. dollars from naturally deflating (and to sustain inflation at the Fed’s 2% target).
AI-proof your portfolio
Stocks tend to be more volatile than bonds and tend to carry much greater risk of capital loss, which can be mitigated through diversification. But stocks have the distinct advantage of passing through economy-wide monetary inflation.
We do not have to go far back in time to make the case that stocks benefit from inflationary money creation. Since year-end 2020, cumulative inflation (as measured by the Consumer Price Index) has been approximately 20% as lavish Covid-era fiscal spending and loose monetary policy flooded the economy with U.S. dollars. Over the same time frame, the S&P 500 has risen by approximately 50%.
When more money circulating through the economy leads to generalized inflation, businesses face higher costs. However, businesses are also able to charge higher prices and earn higher profits, and if they have debt on their balance sheets, it depreciates in real terms. When it comes to stocks, monetary debasement tends to get passed through.
As mentioned, cost efficiencies driven by AI can result in either lower prices to customers or higher corporate profits (or some combination). Investors should prioritize stocks that are more likely to retain the benefit of cost efficiencies and generate higher profits.
The level of competitive intensity within any particular market should determine how much of the efficiency gains would be retained as corporate profits. In the highly competitive consumer electronics industry, for example, most of the cost savings accrued to customers in the form of lower prices.
As a general matter, investors should seek out companies with competitive moats and pricing power. They should also avoid businesses that play in intensely competitive areas. These qualitative attributes are emphasized across our Model Portfolios.
In an environment where AI may produce a windfall of cost savings throughout the economy, focusing on competitively advantaged firms is even more important. Businesses that have barriers to entry and pricing power stand to benefit most. Shareholders of companies that will have to pass the cost savings through to customers (because their competitors are doing so) might see no value creation at all.
When it comes to AI, disruption is the major risk factor investors in stocks should focus on. While many firms will benefit from AI through lower costs, many companies, if not entire industries, could be competitively damaged, if not destroyed, by new business models that are created by AI technology.
We recently highlighted how call centers, which employ some 3 million people in the United States alone, might be first in line for replacement by AI-based customer service systems. A hypothetical business that specializes in call center facilities is arguably an unappealing business to invest in at this moment.
Investors should think carefully about how companies they own might become victims of “creative destruction” from new, low-cost AI technologies. It is possible that some of the most AI-resilient investments will include many low-tech companies that are built around certain physical assets or commodities that will continue to have relevance (and appreciate in value as money is printed).
Gold and crypto
Gold is often described as an inflation hedge, so it may seem counterintuitive that gold would be an attractive asset in the context of a deflationary productivity boom. But gold truly shines as an investment when money is being debased to prevent deflation.
A good way to illustrate how gold performs in this context is to look at the experience of Japanese investors. With an aging population, low levels of family formation and a high savings rate, Japan has struggled with deflationary forces for decades.
To prevent deflation, the Bank of Japan has historically been extremely aggressive. Japan has been a pioneer in quantitative easing and has resorted to negative interest rates and direct purchases of securities by the central bank. Japan’s public debt to GDP now exceeds 250%, a ratio that is approximately twice that of the United States.
Gold has performed quite well in U.S. dollar terms over the past 20 years but even better in Japanese yen terms. In U.S. dollar terms, gold has appreciated approximately 500%, whereas in yen terms, it has appreciated approximately 750%.
