Recently, I went through the data of the S&P 500 and found something quite shocking—the equity risk premium (ERP) has actually turned negative. The last time it was this ridiculous was during the internet bubble in 2000. To put it simply: the expected returns from buying stocks now are even lower than just lying down and buying government bonds.

**Let's talk about how exaggerated this number is**

The forward earnings yield of the S&P 500 is now about 3.35%, the yield on the 10-year U.S. Treasury is 4.16%, which gives an ERP of -0.81% when subtracted. Negative! Such a situation is rare in history; it was only worse in 2000 (when it briefly reached -2%). And then what happened? The Nasdaq evaporated 80%, and the S&P was almost cut in half.

Generally speaking, the ERP should be maintained between 2% and 4% (historical average around 3.5%). This premium is meant to compensate you for taking on the risk of stock market volatility. Now that the premium is gone and there is even a deficit, it means that market pricing has already deviated from the fundamentals. What’s even more critical is that currently, there are additional overlays of AI bubble controversies, geopolitical tensions, and the Federal Reserve maintaining a tight stance—any black swan could become the straw that breaks the camel's back.

**How is a bubble blown up?**

In terms of valuation, the cyclically adjusted price-to-earnings ratio of the S&P 500 (Shiller CAPE) has soared to 37.8-39.8, far exceeding the historical average of 21.2, second only to the peak level of 2000.

Even more distorted is the structural issue: the seven major tech giants account for one-third of the index weight, with market expectations for their profit growth reaching 18%, while the remaining 493 companies only expect 11%. The forward price-to-earnings ratio of these seven companies is 25 times, even higher than during the bubble in 2000. Everyone is betting that the AI revolution will deliver, but this concentration itself represents a huge risk exposure.

However, to put it another way, there is still a difference between now and the year 2000. Back then, many internet companies were not profitable at all and relied purely on storytelling; now these tech giants at least have real cash on their books, such as Nvidia with a gross margin exceeding 70%. Coupled with a relatively low interest rate environment (although it has already risen to 4%), and the FOMO sentiment among retail investors (fear of missing out on the opportunity for a price increase), these factors have collectively pushed up valuations. But these factors cannot change one fact: you are taking on risk without receiving appropriate compensation.

The macro level is also unfriendly. The Federal Reserve may slow down the pace of interest rate cuts (the market expects only 1-2 cuts in 2025), and the yield on the 10-year U.S. Treasury may continue to rise to 3.75%-4.25%, further squeezing the attractiveness of stocks. If inflation rebounds, coupled with trade protectionist policies, volatility will only increase.

**What will happen next?**

In the short term, within 3-6 months, the optimistic sentiment around AI may drive the index upward, and the S&P 500 could possibly reach 4800-5000 points (currently around 4700). However, one must be cautious of the "double kill" risk in stocks and bonds, which has shown signs earlier this year.

Looking at the medium term of 1-2 years, historical data shows that when ERP is below 0, the average return of the S&P 500 in the subsequent 1-3 years is -3% to -14%. I expect a pullback of 10%-20%, although it won’t lead to a crash, but if the profit growth of tech giants falls short of expectations (for example, dropping from 18% to 11%), that would be the trigger for a correction. By then, value stocks will significantly outperform growth stocks.

For more than 3 years, US stocks remain high-quality assets globally. Even if you bought at the peak of the 2000 bubble and held for 25 years, the annualized return is still 8.1%. You just need to be patient now and wait for valuations to return to a reasonable range.

**What should I do?**

It is recommended to reduce holdings in those overvalued tech giants, lowering the position ratio to 15%-20%, and shifting towards the S&P 500 Quality Index—this type of index has outperformed the market by 20% during bubble periods.

Increase the allocation to value and cyclical sectors, such as banking, energy, and small-cap stocks (Russell 2000 Index). The ERP of these assets is still around 2%-3%, and their volatility is also lower. At the same time, a portion can be diversified into Europe or emerging markets, where their valuations are only 60% of those in the US stock market.

In terms of fixed income, the yield on 10-year U.S. Treasury bonds has exceeded 4%, and it may be worth considering allocating to Treasury Inflation-Protected Securities (TIPS) to hedge against inflation risk.

In terms of portfolio allocation, the classic 60/40 stock-bond combination is still reliable, and it's also advisable to keep 10% in cash (to purchase short-term government bonds). If you're concerned about tail risks, you can use options tools for hedging, such as when the VIX index breaks above 25.

Finally, focus on a few key signals: if the ERP rises above 1% (possibly due to earnings exceeding expectations or a decline in U.S. Treasury yields), then you can increase your position; conversely, if not, reduce your stock position to 50%.

In short: now is not the time to bet recklessly. The market may still rise in the short term, but the risks have become very high. Adjust your position structure, prepare for defense, and don't wait until the bubble actually bursts to regret it.