Is This 1999?
Investment Management
The technology can change the world and the stocks can still disappoint. Three strong years have quietly moved allocations, and rebalancing requires no prediction about AI.
By Trey Touchstone · July 31, 2026 · Updated August 6, 2026 · 9 min read
Is today's AI-driven market another 1999?
Today's market shares real similarities with 1999, including concentration, high expectations, and heavy infrastructure spending, but today's largest companies are far more profitable. Investors do not need to predict the outcome; they can rebalance to their agreed risk targets and use higher-quality bond yields to support the plan.
Key takeaways
Transformative technology does not guarantee attractive stock returns at any price.
Index concentration increases both upside and downside sensitivity.
Strong markets can silently push portfolios beyond intended risk.
Rebalancing is a risk decision, not a forecast.
High-quality bonds now offer meaningful income compared with 2021.
My thermostat has been set to 74 since April. That is not the same as saying the house has been 74 since April. On July 22, it reached 106 degrees in DFW with a heat index of 116, and by mid-afternoon the upstairs had drifted to 79. We have agreed as a family not to discuss it.
Some of us fool ourselves into believing that setting the thermostat to 65 will cool the house faster. It will not. Most air conditioners have one speed, and the fancy ones still have a ceiling. The lower number does not summon colder air; it only removes the moment at which the system is allowed to stop. The house reaches 74 at precisely the hour it would have anyway, and then the equipment keeps running toward a number it has no hope of reaching before October. August presents the invoice.
Nobody here consults the ten-day forecast before deciding where to set the dial, either. You pick a number you can live with and let the equipment chase it. The forecast is interesting, but it does not get a vote.
What we are hearing from clients this summer has been more about the heat and less about the market. Understandably so, the market has been excellent. The S&P 500 returned 26.29% in 2023, 25.02% in 2024, 17.88% in 2025, and about 10% through July 22 this year. The gains are appreciated and not entirely trusted. More than a few of you have reached for the same comparison: this feels like 1999.
Our answer is no, this is not 1999 but you do not need us to be right about it. Three strong years have a way of quietly moving an allocation, and bringing it back requires no view on AI whatsoever.
What 1999 actually was
1999 is a fair and useful comparison. The internet was not a fad. It proved to be every bit as transformative as its advocates promised, and then some. So why did the Nasdaq peak in March 2000, fall roughly 78% over the following two and a half years, and take fifteen years to recover?
Because the technology arriving and the stock working are two separate events, and in that instance they were separated by a decade and a half. The prices had sprinted ahead of earnings, and it took the better part of a generation for earnings to catch up to what investors had already paid. Being right about the revolution and being right about the share price are not the same wager.
What rhymes
The comparison got louder this month when chip stocks, the center of the AI trade, fell more than 20% from their late-June record, crossing the commonly used threshold for a bear market. They are still up more than 60% for the year. A decline of that size barely dented the gain, which tells you how far and how fast this market's leadership has run.
Three things genuinely rhyme. Seven companies now account for roughly a third of the S&P 500's market value, a level of concentration the index has not seen in fifty years. The index trades near 20 times expected earnings, above both its five-year and ten-year averages. And enormous sums are being committed to infrastructure well ahead of the revenue that spending is meant to produce, which is precisely what the fiber-optic companies were doing in 1999.
The third one is the echo worth taking seriously. The fiber got laid in the nineties, carried the internet for decades, and several of the companies that laid it went to zero. The railroads got built and bankrupted many of their builders. The infrastructure can succeed and the shareholders can still lose.
What does not
The class of 2000 was priced on ambition. Many of those companies had thin revenue and no profits at all. Pets.com never made a dime, though it did buy a Super Bowl ad starring a sock puppet. It went from IPO to liquidation in 268 days.
Today's leaders are among the most profitable enterprises ever assembled, and their earnings continue to grow. Earnings growth across the S&P 500 is running near 25% this reporting season; FactSet reported 24.7% for Q2, with the largest seven expected to grow 31.1% versus 22.8% for the other 493. That does not make this market cheap. It makes the question different.
In 2000, you had to ask whether the companies would ever earn anything. In 2026, they earn plenty. The open question is whether hundreds of billions of dollars a year in data center spending will yield a return sufficient to justify what investors have already paid. Those are not the same question, and the second is harder than it sounds. A technology can change the world, the companies building it can prosper, and the stocks can still disappoint if their prices already assume an even better outcome.
Concentration sharpens that in both directions. If spending pays off beyond expectations, having a third of the index in seven names works in your favor. If it disappoints, those same seven names take the index down with them. Owning the S&P 500 is still diversification. It is simply less evenly diversified than the number 500 suggests.
What we are doing
None of this will be resolved soon, and we are not going to position your portfolio as though we know how it ends.
What we can see is concrete. Three strong years have pushed equity allocations above where they started. A portfolio that began 2023 at 60% stocks and was never rebalanced could now be north of 70%, depending on what it held. Nobody decided to take that additional risk. It arrived one good day at a time, which is how risk usually arrives.
Therefore, where allocations have moved beyond their agreed ranges, we are rebalancing toward target, consistent with each client's plan. This is not because we think the run is over, but because the amount of risk you carry should be something you choose on purpose.
Why bonds are worth owning again
The money we trim has somewhere worth going, which was not true for most of the past fifteen years.
High-quality bond yields: July 22, 2021 vs. July 22, 2026
| Maturity / index | July 22, 2021 | July 22, 2026 |
|---|---|---|
| 3-month Treasury | 0.05% | 3.89% |
| 2-year Treasury | 0.20% | 4.31% |
| 10-year Treasury | 1.28% | 4.67% |
| 30-year Treasury | 1.92% | 5.15% |
| Investment-grade corporate bonds | 1.95% | 5.39% |
Sources: U.S. Treasury; Federal Reserve Bank of St. Louis (ICE BofA US Corporate Index).
Five years ago, the high-quality end of a portfolio paid you almost nothing to own it. We wrote in 2023 that earning 4.25% on long Treasuries moved clients toward their goals, whether we got a recession or an AI boom, and we were called old-fashioned for it, mostly by ourselves. The math is better today.
The trade-off is worth stating plainly. Inflation is still running above the Fed's target, energy prices have turned higher after easing in June, and the direction of rates from here is genuinely uncertain. If rates rise, longer bonds lose value in the short run. That risk is real, and so is the income. What the table changes is the starting point, and the starting point matters more in bonds than in most things, because a bond's yield is closer to a contract than a forecast. Five years ago, that contract was barely worth signing. Today it is worth it.
Summer ’99
The summer of 1999 ran 33 days at or above 100 degrees in DFW, including 24 straight, and topped out at 107 on August 10. This week we hit 106. Whatever separates that year from this one, the weather is not it.
What separates them is on the income statements, and that is where we spend our attention rather than on the forecast. We cannot tell you whether the AI build-out rewards shareholders on the market's timetable, and we would be suspicious of anyone who claims otherwise. What we can tell you is that your portfolio strategy should not require it to do so.
Your money is not a bet. It is your retirement, your family's plans, and whatever you intend to leave behind. It does not need to win a race. It needs to arrive.
Stay cool out there.
Trey
Important information
Advisory services are offered through Lifeway Financial Corporation, an SEC-registered investment adviser. All content is for information purposes only and should in no way be construed or interpreted as a solicitation to sell or offer to sell advisory services to any residents where it is not appropriately registered, excluded or exempted from registration or where otherwise legally permitted. It is also not intended to provide any tax or legal advice or provide the basis for any financial decisions. Nor is it intended to be a projection of current or future performance or indication of future results. Moreover, this material has been derived from sources believed to be reliable but is not guaranteed as to accuracy and completeness and does not purport to be a complete analysis of the materials discussed. Purchases are subject to suitability. This requires a review of an investor’s investment goals, preferences, and constraints. Investing always involves risk and possible loss of capital.
Common questions
Frequently asked questions
-
Because prices had sprinted ahead of earnings. The technology arriving and the stock working are two separate events, and in that case they were separated by roughly fifteen years while earnings caught up to what investors had already paid.
-
Seven companies now account for roughly a third of the index's market value. If spending on AI pays off, that concentration works in your favor; if it disappoints, those same names take the index down with them. It is still diversification, just less evenly diversified than the number 500 suggests.
-
Three strong years pushed equity allocations higher without anyone deciding to take more risk. A portfolio that started 2023 at 60% stocks could now be north of 70%. Rebalancing toward the agreed target restores the risk level you chose on purpose, and it requires no view on AI.
Sources and further reading
Where these figures come from
Rules, thresholds, and figures change. Confirm current details with the primary source for the year you are planning.
U.S. Equities Market Attributes
S&P Dow Jones Indices
S&P Dow Jones Indices
Magnificent 7 contribution to Q2 2026 earnings growth
FactSet
Earnings Insight, March 27, 2026
FactSet
Daily Treasury Par Yield Curve Rates, July 2026
U.S. Department of the Treasury
ICE BofA US Corporate Index Effective Yield
Federal Reserve Bank of St. Louis
100-degree day data for Dallas-Fort Worth
National Weather Service Fort Worth/Dallas
Nasdaq-100 historical perspective
Nasdaq
Chipmakers and other high-flying stocks slide as the AI trade wobbles
Reuters, syndicated by Investing.com
Insights are educational and are not investment, tax, or legal advice. See our Form ADV Part 2A and Form CRS for important details.
Keep reading.
Can You Retire? The Question Behind the Question
The Withdrawal Order That Quietly Decides Your Tax Bill
Social Security at 62, 67, or 70
Let's start with a conversation.
Tell us what is on your mind. We will listen, answer what we can, and help you decide whether Lifeway is the right fit.
