Editor's Note: Against the backdrop of AI giants continuously pushing up U.S. stock indices and increasing market concentration, Ray Dalio revisits a classic question in his latest note: How should investors allocate their assets when a revolutionary technology is changing the world?

Dalio's core reminder is that technological advancement itself does not equate to the stocks of related companies being equally attractive. Major technological cycles in history have often gone through phases of excitement, crowding, volatility, and shakeouts. Even long-term winners like Microsoft and Apple suffered significant pullbacks within these cycles. Today's AI industry similarly faces multiple uncertainties including overinvestment, intensifying competition, geopolitics, tax policies, anti-AI sentiment, and disruption by next-generation technologies.

The most important point of the article is not to judge whether AI will change the world, but to discuss how investors should cope with a market structure characterized by 'high concentration.' Dalio believes that as a handful of tech companies account for an ever-increasing weight in major indices, investors need to be vigilant about whether they are inadvertently holding a highly correlated, high-risk concentrated exposure. Compared to continuing to chase a few leading stocks, a truly more robust approach is to build a diversified portfolio composed of high-quality, low-correlation assets and adjust the volatility level according to one's own risk tolerance.

In his view, knowing what you don't know is as important as judging what you think you know. In the current market environment driven by AI, characterized by high valuations and concentrated risks, investors should not translate their excitement about new technology directly into a concentrated allocation to a few AI stocks. Diversification is the 'holy grail of investment' in Dalio's eyes for navigating this technological cycle.

The following is the original text:

This note is about: how to play the investment game given the current environment.

Imagine you're playing a game like bridge, poker, backgammon, or chess, and it's your turn to move, and you have a computer next to you to assess the situation with you and suggest your next move. To me, the investment game is like that. Whether or not you have a computer to help you, I think you should:

Given the existing characteristics of the board in front of you, ask yourself what you should do next. That is to say, decide how to act based on the existing characteristics of the market and the various forces affecting the market now.

I've been in this investment game for a very long time. At this stage, my objective is to pass on how I would play this game; furthermore, I also want to create a platform where all sorts of people can explore the investment game the way they want to, learn how they would have done in the past by back-testing, and actually do it well. I believe there are right and wrong ways to play the hand you're dealt. So when you encounter a combination of circumstances XYZ, you should ask yourself, "How should I bet given these circumstances?" and be able to give a good answer.

Now, I'd like to share with you what I see as the current characteristics of the market and what I think should be done and what I am actually doing.

How to Handle This Particular Set of Circumstances

What are the most important circumstantial influences at play, and how should a person bet given them?

It seems to me, and probably to most people, that the environment we are in now is one in which a major new technology, primarily AI, is driving the market in an industry dominated by a very few companies. These companies represent a high percentage of the total market capitalization and are having a huge impact on the market and the economy. All such periods have in common a lot of excitement, uncertainty, and volatility concentrated in the new technology industry and transmitted from it to the global stock market. As a result, the volatility and uncertainty around this industry is important.

In addition, there are uncertainties related to other big drivers. I call these drivers "the five big forces": 1) what is happening with debt and money, 2) what is happening with political/social issues that could have big implications for taxes and other politically driven market factors, 3) what geopolitical forces such as wars are doing to the markets, 4) what natural forces are doing, and 5) what new technology is doing. I'll put these circumstances into my investment system to have it think about how to bet given them, and I'll also think independently about what to bet on.

When thinking about how to bet given these circumstances, the most important question is: Which choice do you want to make? a) to make a bigger bet on the new technology relative to the broad equity index like the S&P 500, i.e., to overweight this new industry or overweight what you think are the best companies in this industry, b) to have your exposure roughly where the index weights are, or c) to diversify away from this concentration?

Nearly everyone wants to own the best investments and works to do that, and there certainly appears to be a new technology that seems to be changing nearly everything. Yet history shows that, at this stage of the cycle, most people fail by putting a large percentage of their chips into the stocks of a few leading technology companies. There is a logical set of reasons for this, and it has always played out in the same way. While this AI technology is indeed unique, there have been many other "unique" new technologies in the past that serve as analogous cases and references. People should study these cases; if they choose to ignore them, they must have a good explanation as to why this time is different.

Risk is Certainly High

All cases of big new technologies in the past play out in similar ways for the same logical reasons. High risks and great uncertainties are inherent characteristics of these new technology companies. Looking at how these companies performed in analogous historical circumstances, we see that even the best revolutionary new technology companies that prospered over the long run, such as Microsoft and Apple, got killed during similar stages of their development. Also, at the time these new technology companies emerge, not in hindsight, it's not easy to tell which ones will succeed and which will fail—think about IBM. If you look at all these cases, you see: big new technology companies naturally have highly uncertain futures.

For example, they either overinvest or underinvest. That's because they have to invest enough to win the competition or they will lose for sure; but they can't know with enough precision what will happen to know if they are overinvesting. Either overinvesting or underinvesting is costly.

Also, they can't foresee all the changes, including exogenous changes, such as monetary tightening, wars, big tax changes, etc., which will impact them. So they all go through wild up and down cycles: first exciting investors, then scaring them and shaking out the fragile ones, leading to exaggerated swings in the market. Furthermore, just as these new technologies and new technology companies have disrupted those before them, most of them will eventually be disrupted by newer technologies and newer technology companies in ways we can't pre-imagine. So we should also consider whether the same risks apply to these new technologies and technology companies now. The impact of quantum computing is one of the known knowns. What about the unknowns that aren't yet imagined?

What about the risks from competitors? For example, China is producing and distributing AI technology, and Chinese policymakers have completely different views about the economy and AI. We are in a new technology war, and national leaders believe they must win it. Their perspective on AI and its impact on the economy and the well-being of their people will prompt them to offer this technology for free or at a low price because it has great productivity benefits and raises the general standard of living. In their view, profits are not as important as the overall benefits derived from many people using these new technologies. I think they will compete in international markets as they have in cars, solar panels, batteries, and many other products.

This particular set of circumstances is a lot like many cases in history that provide lessons. I can't help but think about how, at the end of the Dutch Empire and the beginning of the British Empire, the British beat the Dutch in shipbuilding and other important industries. Also, there is a geopolitical conflict around Taiwan, which at least should make us consider the possibility that, as a tool of geopolitical war, China might prevent chips from leaving Taiwan. AI stocks face other risks as well, such as the risk of a wealth tax or other taxes going up that could force holders with a lot of their wealth concentrated in these stocks to sell them; or the risk that rising anti-AI sentiment could limit the space for companies to push the technology forward.

I could list many more concerns to worry about, and I could also list an equally long list of great opportunities that AI will create that I'd like to bet on. I'm not saying how these risks will play out, nor am I saying that one shouldn't bet on AI companies. I'm just saying that there is a lot of concentrated risk in the markets, which is indisputable; and people should know what to do in such an environment. Based on my study of all analogous cases and the logical reasons for them, I'm certain the risks are high and the best way to handle this environment is:

Be Well Diversified

As you probably know, my mantra is "diversify." My "holy grail of investing" is to try to have 15 good, uncorrelated bets that are risk-balanced. In other words:

A well-diversified portfolio of good bets will outperform a concentrated bet. It has a better risk/return ratio and can be engineered to produce better returns at the same level of risk. The more risk is concentrated in one area of the market, the more a person should diversify; especially when the market is driven by a revolutionary new technology, which inherently creates enormous uncertainty.

This is not an opinion; it is a mathematical certainty. For example, if I compare an investment with a risk/return ratio of 0.3, assuming a return of 6% and a standard deviation of 18%, which are typical assumptions for stocks; then, if I hold 5, 10, or 15 uncorrelated investments, I can get the same 6% return but the risk, as measured by standard deviation, drops to 8%, 6%, and 5% respectively. So if I hold 15 good, uncorrelated investments, my risk/return ratio improves by 4.3 times, from 0.3 to 1.29. If you want, you can then apply leverage to that and get a much higher return at the same level of risk. That is a fact.

I have great confidence in this. That confidence comes from my back-testing, the returns I have actually delivered over more than 50 years of investing, and the logic of probabilities that fits with it: a diversified portfolio of good bets, adjusted to the volatility one wishes to bear, will produce much better returns over time than the concentrated bets that most investors tend to hold. More specifically, through good diversification, a person can achieve a better risk/return ratio than any concentrated bet; and by then adjusting it to the level of risk one wishes to take, one can achieve a higher return at one's target risk level than through any other process.

Because I am passing this on, it is now my "not-so-secret" way to successful investing. Nonetheless, I rarely come across investors who think about investment strategy this way. That is, I rarely come across people who think from a portfolio construction perspective, i.e., how a well-structured, diversified portfolio of bets will perform compared to being concentrated in the stocks of companies in a great new transformative industry. Most people are just thinking about whether these stocks and this industry will do well and how to bet on them. The performance of those who think about portfolio construction and those who don't will be vastly different. So I'll articulate my thinking about how to do this well more completely at another time.

For all these reasons, given this particular set of circumstances, thinking about how to play the hand you're dealt should lead a person to ask themselves: how much concentration should I have? And then to diversify.

Returns Look Like They Will Be Low

The risk being high is indisputable. Now I'll give you an opinion that might be wrong: future expected returns are low. My perspective on future expected returns comes from analytical work related to valuations and from the readings of my bubble gauge: expected real returns for stocks over the next 5 to 10 years look to be about -5% to -10%, though there is considerable uncertainty around these numbers. These stocks look to me like long-duration assets that are risky because it is hard to reliably see into the distant future; they look expensive; and the holder base does not look stable.

A Question My Research Team Raised on This Topic

In a recent meeting, a member of my research team asked me: Why do you think the market is wrong to be configured the way it is today? How do you know that the lack of diversification in the market today isn't for rational reasons? For example, some investors think expected returns for AI stocks are very high; or, when an industry represents such a high percentage of the total market cap, such concentration in the index occurs naturally; or, when an industry is met with a lot of enthusiasm, many investors buy these stocks without making smart, reliable calculations about what future earnings will look like and how those earnings should be reflected in prices.

My Answer

Prices rise for various reasons, and not all of those reasons are good. Some investors think about prices and push them higher because they think prices are attractive relative to the fundamentals; some investors hold these stocks for the long term because they recognize it's a great new technology and take the rising stock prices as confirmation that they are good stocks; and some investors have index exposures, which give them large weights in these stocks passively. It seems to me you can struggle with these questions in order to decide what you want to do; or you can realize that you don't have to struggle with this question because you simply don't have enough information to bet with confidence. You can simply say, "I don't know enough to bet" and not bet.

What gets people into trouble is thinking that they have to have a view and that their view is valuable; but the more likely case is that they can't form a view that is reliable enough to be worth betting on.

Footnote: To be clear, I am not recommending avoiding bets. Besides, you can't avoid bets because you have to put your money into some investment or cash. Most people think cash is the lowest-risk investment, but over the long run it is almost certainly the worst investment. What I recommend is knowing how to diversify your bets well, even when you don't have tactical views about which markets are good and which are bad. The way to do that is to have a well-balanced strategic asset allocation portfolio and to hold it when you don't have tactical views you have enough confidence to bet on. But that's a topic for another time.

So I believe that: Knowing what you don't know and therefore deciding when not to bet is as important as knowing what you do know and therefore deciding when to bet.

More simply put, I believe in this principle: Because it's typically hard to know enough to justify concentrated bets, the best approach is to have only a diversified portfolio of bets that you have the most confidence in that are uncorrelated with each other and to engineer that portfolio to the level of risk you wish to take. That is my "holy grail of investing."

At this moment, given this particular set of circumstances, I don't think anyone knows clearly enough what will happen next in this technology-driven market to be able to make a big, concentrated bet. To me, avoiding concentration and staying diversified is the best way to deal with this "not knowing." I know this is contrary to what you might read in textbooks. Textbooks essentially say that markets are efficient, so you should "trust the market."

To sum up, the market is unusually concentrated and revolves around a revolutionary new technology. This fact should remind us: not to confuse our excitement about a new technology with whether the stocks of the new technology are attractive; and not to hold a set of high-risk, highly correlated concentrated bets, throwing caution to the wind. Especially when we can achieve similarly attractive returns with much lower risk through intelligent diversification.

P.S. I won't share my specific holdings or tactical views with you because I don't want to be your investment advisor. But I will shortly share with you some key perspectives behind these views, including the readings of my bubble gauge and the logic behind them.