Stock Talk - Did Warsh Just “Light this Candle”? Stock Talk Update August 14, 2026

Episode Date: August 14, 2026

  Is the AI boom beginning to rhyme with the late 1990s? Strong earnings, record market highs, massive semiconductor investment, and an unusual shift in global currency conditions are creating a market setup investors should be watching closely. In this video, we examine the growing comparison between today’s AI-driven market and the final stages of the dot-com boom. The comparison became even more interesting after the United States and Japan stepped in to support the Japanese yen—an unusual development that raises an important question: could changing global financial conditions provide another source of liquidity at the same time earnings and AI investment remain strong? That doesn’t mean 2026 has to repeat 1999. Today’s largest AI companies are generating substantial revenue, profits, margins, and cash flow. But history shows that strong businesses, strong earnings, improving liquidity, and increasing investor enthusiasm can sometimes combine to produce powerful—and increasingly speculative—late-cycle markets. In this video, we break down: ✔ Why the 1999 dot-com comparison is becoming more interesting ✔ What strong corporate earnings are telling us about the market ✔ Why semiconductor earnings remain critical to the AI investment cycle ✔ How massive AI infrastructure spending could affect future returns ✔ Why the U.S.-Japan yen intervention matters ✔ What happened during the Y2K liquidity window in 1999 ✔ How interest rates and inflation expectations affect stock valuations ✔ Why liquidity could become an important market driver ✔ Four indicators investors should be watching from here ✔ What this environment could mean for retirees and people approaching retirement The goal isn’t to predict exactly what the market will do next. Instead, we believe investors should focus on the forces underneath the market: earnings, interest rates, AI investment, liquidity, and the dollar—and make investment decisions within the context of their own financial plan, income needs, time horizon, and ability to tolerate risk. If you’re retired or getting close to retirement and would like a second opinion on your investment and retirement plan, use the link below to schedule a free consultation with Oak Harvest Financial Group. There’s no obligation. We’ll learn more about your goals, income needs, and concerns and help you understand whether there may be opportunities to improve your retirement plan. https://click2retire.com/lets-connect   00:00 Is the AI Boom Starting to Look Like 1999? 00:42 The New Yen Intervention Signal 01:26 This Is Bigger Than AI 02:20 The Market Framework We’ve Been Using 03:10 What We Expected for the Second Half of 2026 03:52 Earnings Are Still the Engine 05:21 Are Stocks Too Expensive? 06:49 AI’s Massive Physical Buildout 07:45 The Biggest AI Spending Risk 09:02 Why Liquidity Changes the Story 09:23 What Happened in 1999 10:25 The 2026 Yen Intervention 11:12 The Interest-Rate Risk 11:58 The Real Question Investors Should Ask 12:18 What 1999 Can—and Can’t—Tell Us 13:23 Confidence and Caution 14:16 Four Things to Watch Now 15:34 Free Retirement Plan Consultation   #StockMarket #ArtificialIntelligence #RetirementPlanning

Investors, something just happened this summer a few weeks ago that caught my attention. For more than 15 months, our team here at Oak Arvus has been asking one big question. Is today an AI boom starting to look a lot like the dot-com boom? I want to be clear, we haven't used the comparison as a reason to panic. In fact, our view has been that if a market is still early enough in a speculative cycle, high valuations alone don't necessarily mean the rally has to end. And now the comparison is getting even more interesting. earnings are very strong, AI spending is booming, semiconductor spending is rising, the stock

market's hitting new all-time highs, more stocks are joining the rally. Now something new has happened that we haven't seen in decades. The United States and Japan stepped in together to support the Japanese yen. The last time we saw something similar was June 17, 1998. That was during the Asian financial crisis right in the middle of the dot-com boom. So here's the question that hit me this weekend. Did Kevin Warsh and the global policymakers just create a new version of the late-1999 Y2K liquidity window? Because that period of added liquidity happened right before the final huge run, technology stocks that ended, the NASDAQ peaking in March of 2000. I'm not saying that has to happen again, and this historical comparison isn't a forecast, but that's set up important enough

investors should understand both the similarities and very important differences. Everyone, this isn't a story about just AI. It's about earnings, interest rates, liquidity, the dollar, and how those forces may affect the markets from here. The framework we've been using has helped us focus on the things that we believe matter most. That doesn't mean every future view will be right. Markets can change quickly, no framework can predict every turn. The goal isn't prediction. It's process. Instead of reacting to fear, headlines, or whatever financial TV is yelling about today, we want to keep watching the real drivers underneath the markets. Back in 2000, In 2020, our work kept coming back to three big things.

First, earnings. Second, interest rates including real rates and inflation expectations. And third, the U.S. dollar. Our published outlook remained constructive on the bull market while also recognizing risks from tariffs, war, inflation, and changing financial conditions. Then our outlook for the first half at 2006, we described a market that could have a rougher ride, possible weakness early in the year, followed by his stronger conditions in the summer. We also said that the dot-com investment cycle was worth watching as a historical comparison.

Before the sell-off late in the first quarter, we described the AA market as being in the seventh-ending stretch. Not necessarily early, not necessarily finished, more like a market taking a breath after a strong run. Then in June, our second half outlook discussed the possibility of a stronger third quarter, followed by a more difficult fourth quarter. So far, we've seen a strong summer rally on August 7th, the SP 500 close, at 7757, while the NASDAQ also posted it a strong week. But I want to stress something here. The historical outlook lining up with part of what later happened doesn't mean the next outlook will be right. We're showing you the process, not claiming the ability to predict markets in the future. And that process says watch earnings first, then interest rates, then liquidity, then the

dollar, then energy. That's much more useful than following the mood day to day. And this brings us to what I believe is the most important part of this story. Because earnings are still an engine. Fact set, August 7th report showed very strong results. 88% of the S&P 500 companies reporting, 86% had beaten earnings estimates, and 76% had beaten revenue estimates. Second quarter earnings growth was about 50.5% from one year earlier. The end of June, analysts had expected only 23.1% growth.

Now there's a catch here. Alphabet and Amazon had large one-time investment-related. gains. Those gains increase the overall earnings number. However, we remove those two companies, the S&P 500 earnings growth drops from about 505% to 32%. That's a big difference, but 32% would still represent very, very strong earnings growth. Revenue growth was around 15%. That's the strongest growth since 2021. And it's not just one quarter. Faxset currently expects earnings growth of about 27.5% in the third quarter and about 25.2% in the fourth quarter, or for a total, 2026 calendar year growth rate of about 30%. And then next year in 2007, earnings growth of around

13.5%. Those are estimates, of course. They can change sometimes very quickly. Current bottoms up estimate the SP 500 put that number at around $359 for this year and $405 for $2,000,000. Okay, everyone, so stocks aren't cheap. The SP 500 trades near around 20 times expected earnings. That's somewhat above the 10-year average of around 19 times, but valuation by itself has historically been a very poor short-term market timing tool. And here's one important difference worth watching compared to the late 1990s. For six straight quarters, the market's PE ratio has declined while earnings have risen.

In other words, earnings have been doing most of the work here. Here's another way to look at it. Using the old market rule of thumb of 10-year treasury yield, which is near around 4.65%, divided one by that, gives you a P.E of around 21 and a half times. Multiply that by the expected 2026 earnings of roughly $359, and you get an S&P 500 value near 7722. As of this recording last Friday, the index closed at 7757. That doesn't mean 7722. is the correct value of the market, this simply is an illustration of how interest rates and

earnings can interact in valuations. Markets can trade far above or below simple valuation models for long periods of time. For retirees, that distinction matters. A high valuation combined with falling earnings can create serious risk. A high valuation combined with rapidly growing earnings can behave very differently. Either situation removes the risk of market losses. So let's move on to AI. The dot-com boom required a huge physical build-out. It needed fiber. It needed routers, servers, switches, and computer chips. It also needs huge physical build-out. It needs GPUs, memories, advanced chip packaging, networking, equipment, optical links, power systems, cooling, and enormous amounts

of electricity for data centers. Factsets numbers show how important this build-out has become. Technology revenue grew at about 35.9% in the second quarter. Semiconductor and semiconductor equipment revenue grew about 77%. Semiconductor earnings grew about 135%. And if you remove semiconductors from the technology sector, tech earnings growth drops from about 70% to roughly 34%. That's why we describe semiconductors as the toll road underneath AI. AI needs compute and compute needs chips.

But this is also where one of the biggest risks is growing. AI spending has exploded. Large technology companies are spending enormous sums of money on data centers and equipment. So I think the question is changing. The old question was, is AI real? There's a growing evidence that AI demand is real. The more important investment question is, Now, how much money will companies eventually earn on the next dollar they spend building AI infrastructure?

Microsoft, Amazon, and Google have recently provided some encouraging signs. And investors still need to watch the return on that spending carefully. Demand for computing power is also growing as AI agents create more searches, more tasks, and more machine-to-machine activity. Cloud Fair recently discussed the rapid growth of AI-related traffic on its networks. There's another important difference from the dot-com era. Many of today's largest AI beneficiaries were already producing substantial revenues, profits, and cash flow. At the same time, Open AI and Anthropic remain private, which means some of the AI-related

speculation remains outside the public stock market. So yes, parts of the cycle could become too speculative. Yes, there are similarities to the late 1990s, but this isn't an exact copy of 1999. Everyone, this is where the story gets even more interesting, because strong earnings and strong AI spending are only one side of the market. The other side is liquidity. The events surrounding the yen may have changed that part of the story. Here's why. Dotcom bubble didn't end right after the financial stress of 1998.

After the Asian financial crisis and long-term capital management collapse, financial conditions change and policymakers changed. That's when 1999 happened. The Fed was raising interest rates, but officials were also worried about Y2K. Banks and businesses feared computer problems could disrupt their financial system and the calendar changed to year 2000. Most people find that hilarious nowadays, but it was a real concern back then. So, back then, the Federal Reserve created special liquidity tools.

designed to reduce that funding risk. One important facility opened on October 1st, 1999, and lasted until April 2000. That program didn't create the dot-com boom by itself. That's important to remember. There were many forces pushing technology stocks higher, but the additional liquidity helped reduce one major financial system concern at this time when investors were already willing to take more risk. From October 22nd, 1999, through the NASDAQ peak, March 10, 2000, then NASDAQ grows sharply. Again, that historical return isn't being shown to suggest today's market will do the same thing. It's being shown because the timing of liquidity changes can matter. So let's fast forward to 2006. I'm 60 now, not in my late 20s. The United States and Japan

just stepped in to support the yen after significant weakness in the currency, with the New York Fed carrying out transactions on behalf of the U.S. Treasury. That's unusual. But we need to be very careful with a comparison. This wasn't a new Federal Reserve quantitative easing program. It wasn't the same thing as the Y2K liquidity facility. And it doesn't mean broad monetary easing has begun. But here's the question worth watching. Could actions designed to reduce stress in the global currency markets

also affect financial conditions and global liquidity? It's very possible. And if policymakers take additional steps, financial conditions could change again. That is the more bullish side of the analogy. But there's a very different side investors can't ignore. Today's inflation and interest rate environment isn't the same as late 1999. Ten-year treasury yield ended August 7th around 4.64%. And markets continue to debate the future path of Federal Reserve Policy.

Remember how we think about the 10-year treasury yield. It's basically made up of two important pieces, real interest rates and inflation expectations. If inflation expectations rise, higher discount rates can put pressure on stock valuations. If real rates rise because financial conditions tighten, that can also put pressure on valuations. For growth companies whose biggest profits may be years into the future, higher rates can matter even more. And if both real rates and inflation expectations rise together, that could create much more tougher environment for growth stocks.

So here's the real question. It isn't. Did the Federal Reserve turn on the money printer? That's not what happened. The better question is this. Did policymakers reduce one source from global financial stress? the same time earnings and AI spending are accelerating. If they did, could easier financial conditions help extend the current market cycle?

Possibly. But it's only one possible outcome. History shows that some late-stage bull markets can become faster, more emotional, and more speculative before they eventually end. History also shows that markets don't follow a script. That's what makes the 1999 comparison useful. Not because 2006 has to end the same way. It doesn't.

is useful because it gives just a framework for understanding what might happen in strong technology investment, rising earnings, investor excitement, and changing liquidity all come together. At the same time, today's market appears stronger, some very important ways than much of the technology market was near the dot-com peak. Many of today's AI leaders have real earnings, real margins, real revenue, and real cash flow, but strong businesses can still become expensive investments. The key distinction. So the story isn't, this is 1999 all over again. The story is, we may be entering apart the cycle that shares some features with late 1999, while the businesses underneath today's market are very different. That's a difference maker. And I think we should give investors both

confidence and caution. Confidence because the earning story remains strong. The AI buildout is still happening. The semiconductor cycle remains important. And the market has continued to show strength. But caution? Because strong markets don't eliminate risk. Valuation matters, interest rate matters, liquidity matters, and investor behavior matters. Late cycle markets can sometimes feel their best when investors are becoming less careful. That doesn't mean investors should automatically run away from stocks. And it doesn't mean investors should automatically buy more. It means investment decisions to stay tied to your financial plan, your time horizon, your need for income, and the amount of risk you can eventually afford to take.

For retirees and people close to retirement, that matters even more. You don't want one exciting market story to replace your investment discipline. AI may change the world. And some AI-related investments can still be overpriced. Both things can be true at the same time. So here are the four things I'd keep watching. First, earnings. Earnings remain an important support underneath this market.

But estimates can change, so we'll keep watching whether those expectations continue to rise or begin to weaken. Second, AI spending cycle. The buildout continues, but the next question is return on investment. How much profit will companies ultimately generate from all this spending? Third, interest rates. Free will rates and inflation expectations rise together. That could create increasing pressure on gross stock valuations. And fourth and finally, liquidity. If the yen intervention turns out to be part of a wider change in global financial conditions, then the late-19-19-comparison may become more interesting. But for real rates, oil, and the dollar rise together, that would be a warning sign that

financial conditions are becoming more difficult. For retirees and near retirees, I think discipline is straightforward. Stay focused on earnings, respect valuation, watch the cost of money, watch liquidity, make investment decisions based on your own financial plan rather than one market forecast or one exciting technology. Because a great technology doesn't automatically make something a low-risk investment. History doesn't repeat exactly, but sometimes it gives us a rhyme worth paying attention to. Right now, that rhymes getting louder.

And if you're retired or getting close to retirement, you'd like a second opinion on your investment or retirement plan. Give O'Carvis Financial Group a call. We offer a free consultation where you can learn more about your goals, your income needs, and the risks you're concerned about to help you understand whether there may be ways to improve your plan. There's no obligation. Just give us a call or visit the link below to schedule your free consultation. We'd be happy to talk to you. All content contained with an Oak Harvest podcast expresses the views of the speaker and is for informational purposes only.

It is based on information believed to be reliable when created, but any cited data, indicators, statistics, or other sources are not guaranteed. The views and opinions expressed herein may change without notice. Strategies and ideas discussed may not be right for you, and nothing in this podcast should be considered as personalized investment, tax or legal advice, or an offer or solicitation to buy or sell securities. Indexes such as the S&P 500 are not available for direct investment, and your investment results may differ when compared to an index. Specific portfolio actions or strategies discussed will not apply to all client portfolios. Investing involves the risk of loss, and past performance is not indicative of future results.