This will crash the market

Added: 2026-06-02

[00:00:01]Hey guys. So, let's go over the market for the past week. So, just like the previous week, not much has happened in the past week. Um, the S&P was up by around 1.8% and NASDAQ was also up by around 2.6%. And the markets were driven primarily by the tech and the semiconductor stocks. Just like the past weeks, the tech stocks were up by around 5% and semiconductors were up by around 8%. Now in terms of the key data, the two key data that we monitor on a weekly basis which are the most important data are oil and 10ear Treasury bond yield.

[00:00:34]The oil price came down a bit to around $91 as the market continuously expects the US Iran conflict to be resolved sometime soon and the 10ear Treasury bond is at around 4.45% and both metrics came down slightly from last week which is a little bit favorable to the market. Now, in terms of the macro data, one meaningful data that came out was the core PCE price index. On a month-over-month basis, it was at around 0.2% versus the 0.3% consensus. And on a year-on-year basis, it was at 3.3% versus 3.3% consensus.

[00:01:08]Now, the reason core PCE price index was relatively stable versus the CPI index is because the core PCE index, as you guys know, strips out food and energy.

[00:01:19]So basically the inflated oil price is not factored in into the core PCE price.

[00:01:24]So in terms of the core PCE price, the metric remained relatively stable. Now, one thing I'll say though is that if the oil price hovers around $100 for a longer period of time, for example, if the oil price hovers around $100 or somewhere around there for a few more months, for example, if it stays at around $100 moving into July or August, we will see core PCE index being impacted by the high oil price as well.

[00:01:52]As you guys all know the oil price impacts every industry possible given it impacts the factories, the logistics and etc. So if the oil price continues to be really high, it will eventually impact every single price that we encounter and that will be reflected on the core PCE price. But as of April that did not happen yet given that the war was still at an early phase of the stage. So um the April quarter PC price remained relatively stable. So we can say that both the oil price and the 10-year Treasury bond yield were relatively favorable to the market. On an absolute basis is still extremely high but versus the previous week it came down a little bit which means that it was relatively favorable to the market in terms of forming a favorable narrative. Now the other data that we need to monitor is the Fedatch interest rate predictions.

[00:02:46]When we look at Fed watch data, the interest rate hike probability by the year end slightly decreased. So I remember that the previous week it was at slightly over 50% meaning that by the end of the year the market was expecting that there will be one rate hike with a probability of 50% or more. But this week, because of the falling oil price and the stabilizing 10-year Treasury bond, the probability of a rate hike came down very slightly and now it's at around 48%. Now, lastly, let's look at the forward price to earnings multiples of the key semiconductor companies. Now, one thing I'll say though is because all these companies are increasing so much in terms of the share price. I used to look at one-year forward multiple meaning that the 2026 year end earnings based price to earnings multiple.

[00:03:36]However, because all these companies are hiking in price so much, I think it's fair to say that we also need to look at 2027 so two-year forward multiples as well. So on a 2026 year end multiple basis these companies including AMD, Intel, Marvel, Broadcom, Applied Materials, Nvidia, Lamb Research, ASML, Micron, all these companies on a median basis is trading still at around 44 times. Now because I monitor the medium multiple, it's fair to say that versus the previous week, the multiple stayed pretty much the same at around 44 to 45 times because I'm taking the medium multiple. But if you look at the average multiple it has come up a little bit um and it was mostly driven by micron. Now if you remember the previous week forward multiple from micron was at around 9 to 10 times but now it's at around 11 to 12 times meaning that micron has rallied a lot in terms of the price in the past week. Now if you look at the 2027 year end multiples the forward multiples for all these companies AMD is trading at around 38 times Intel is trading at 71 times.

[00:04:42]Marvel at 32 times, Broadcom at 21 times, Applied Materials at 26 times, Nvidia at 17 times, Lamb Research at 37 times, ASML at 34 times, and Micron is at 8 times. So in order for all these companies to make sense in terms of the trading multiples now we really need to monitor the two-year forward multiples as well given that if you look at the one-year multiples the market is expecting all these companies to perform so well in the following year as well i.e 2027 so all these companies are basically trading based off the forward forward multiples which the market is reflecting in the current share price.

[00:05:23]Basically, what this means is that the market is so excited about these companies earnings in the future years that they're basically grabbing the forward forward years earnings and pulling it back to the current year and applying that earnings to the current share price. So, for example, if you look at Intel on a 2025 trailing basis, it didn't generate any meaningful earnings. So, it's actually negative.

[00:05:46]But for 2026, Forward Multiple, which is a one-year forward multiple, is trading at extremely high multiple, which is over 100 times. However, the market is so excited about this company that they're actually looking at 2-year forward multiples, which is a still a high of a multiple at over 70 times. But in order for the market to justify the current valuation of these companies, they're basically looking at the forward forward multiples or even forward forward forward multiples to justify the current valuation. Also from Micron as well, historically speaking, this company was trading at a very low multiple. So even a 12 times forward multiple is really high from Micron. But if you look at forward forward multiple from Micron, it's only trading at around eight times. So it's fair to say that the market is extremely excited about this industry and these companies that they're actually looking at forward forward earnings and grabbing all those earnings to the current year to justify the current valuation and to justify that the current share price is not expensive. So from now on I think it's fair to say that we need to monitor not only the one-year forward multiples but also the two-year forward multiples as well. And if the share price continues to go up from here, we would even need to consider monitoring three-year forward multiples as well, which is based off 2028 earnings. I think on a relative basis, the Max 7 companies including Amazon, Meta, Microsoft, all these companies and Google as well. All these companies are trading at a pretty low multiple compared to their historical lever averages. Not because their earnings are really bad, but because they're spending so much cap tax on building data centers and those cap tax is flowing in into the semiconductor companies, which means that these semiconductor companies from the investors point of view and the market's point of view are the ultimate beneficiaries of the current AI hype.

[00:07:37]Okay, so that's enough of data. Um, just to reiterate my strategy again, my current strategy is hold. Now, I've been saying this for months, but I'll be holding my stocks in the tech and semiconductor sectors as well. And my recommendation to you guys as well, if you hold stocks in the tech and semiconductor sectors, is to hold those stocks as well and just keep the stocks for now. If you don't own any stocks within the tech and semiconductor industries, I would suggest that you very very carefully consider whether to buy into the market at this point of time. As I told you all the time, the market has rallied for such a long time that it is not a comfortable valuation to jump in at this stage. Now, jumping in with your entire savings or net worth, I think is hard to justify at this point of time. But if you really really want to open positions at this point of time, I would suggest that you just leverage the minute dips which happens in the market. As you guys all know, I gave you two buying calls. one in April 2025 and second one in March 2026. However, apart from those two big blind holes in between, there were many and numerous mini dips and plenty of mini dips in between which enable the investors to open their positions. So, if you really really want to buy into those sectors or stocks at this point of time, I really suggest that you take advantage of those mini dips which I believe will continue to happen in the near term. Also, as I say all the time, never use leverage or buy any weird options or anything like that at this point of time. Okay. Now, having that said, I just wanted to review some of the market crash scenarios. Um, I already told you guys what my sell signals are. Number one, any tech or semiconductor company saying at their quarterly earnings release that their revenue or earnings will plateau. And number two, Fed interest rate pivot towards an interest rate hike. And number three, crazy valuation of semiconductor stocks. Now in terms of the valuation I gave you around 50 to 60 times forward multiple as my personal threshold. So if any of those three conditions happen in the market I will consider liquidating my positions. And also depending on the weight of those signals I will either partially and gradually liquidate my positions or if it's a very strong signal I could even consider liquidating all my positions in one go. Now number two and number three which are the fed related point and the valuation related point are pretty clear and straightforward. So today I just wanted to talk about and expand a little bit on this first point which is the company saying at the quarterly earnings meeting that their revenue and earnings will plateau and I just want to go over some of the scenarios which this may happen. So the first scenario I wanted to talk about is the slowdown of Mac 7 companies capex spendings. Now, the semiconductor company's biggest customer at the moment are the Max 7 companies.

[00:10:32]I've been telling you for the past few weeks, but in order for the semiconductor companies to continue the bull run like the past months, the Max 7 companies need to spend around $2 trillion of capex. Now, I already told you this multiple times, but just to reiterate, 20% ROI on the $2 trillion is $400 billion. applying 20 times multiple on the $400 billion of earnings is $8 trillion of market cap expansion for the Mac 7 companies and as I told you all the time it's a huge stretch now the Mac 7 companies are spending on average around 80% of their operating cash flow on capex now this is not an exact number because we can't really get a exact data but that's roughly the amount of cash that they're spending just in buying trips and building data centers 80% of their operating cash flow, they're spending purely on chips and data centers. Now, in 2025, the combined capex of the Max 7 companies was around $400 billion. In 2026, it was over $700 billion. Now, this effectively means that they're spending their entire free cash flow on chips and capex. So, what are they doing to fund their capex now?

[00:11:45]They're basically taking on debt because capex is eating all of their free cash flow. The hyperscalers are shifted from being self-funded to tapping the credit markets. Now over the last track year, hyperscaler related debt issuance roughly doubled to hit $180 billion.

[00:12:02]They issued roughly 120 billion directly while the remaining $60 billion was raised via private equity and infrastructure lenders building data centers on their behalf. Now, while $180 billion in new debt sounds massive, it is actually a small portion compared to their equity valuation, the combined market cap of Max 7 companies is at around 15 to 16 trillion. And therefore, their total outstanding debt still represents only a low singledigit percentage on their market cap. And this is why credit markets are still eager to lend to them. their leverage ratios which is debt to equity and debt to assets have actually remained stable or actually trended downward over the past few quarters due to their massive equity appreciation.

[00:12:48]However, we do need to monitor this very carefully because a company and a set of companies spending their entire operating cash flow on capex which they're not certain on generating the appropriate ROIs is actually a huge risk to all these companies. They're basically spending cash on capex on a nation industry which they do not know at this point how much money that capex will generate. As I told you before, $2 trillion of capex requires $400 billion of earnings. If they don't generate $400 billion of earnings on the $2 trillion of capex which they spent, the market will reflect that and wipe that $2 trillion out from their market cap immediately. So we need to monitor that point very carefully and try to see whether the max 7 companies slowly trims down their capex spending because they do not believe the ROI can be generated from the capex which they're spending on a daily basis these days. Now the second point I wanted to talk about is the delay in data center buildouts. Now data center buildout is becoming more and more challenging even if you have the money and resources. Now, as you guys all know, the data center expansion is what is enabling the semiconductor companies to generate their revenue.

[00:14:03]More data centers means that there's more space for the trips of the semiconductor companies to go in. So, more data centers means more revenue for the semiconductor companies. Now, as data centers are built out, the rules in developing data centers and regulations are becoming more and more strict.

[00:14:19]Permits are getting more and more rigid.

[00:14:21]And even if the Mac 7s want to scale the data centers infinitely, the data center needs to obtain permits and it is getting more and more complex. Let me give you some examples. In March 2026, the federal lawmakers including the Senate and the Republicans introduced the artificial intelligence data center moratorium act which seeks a nationwide halt on constructing new data centers requiring 20 megawatts or more of power.

[00:14:46]Also at the state level, New York and at least 10 other states have introduced legislations to temporarily ban data center construction until environmental impacts are reassessed. Also at the local level, over 100 localities across the US have enacted moratoriums on data center construction over the past year.

[00:15:06]So as you can see the data set construction is not something which the max 7s can just scale out just because they have the money and power but it actually requires a lot of steps which needs to be negotiated with the congress and the local governments as well. And as the data center becomes more and more abundant these legislations and regulations may become a lot more complex which may delay the speed of the data center buildouts. Now if the data center scale out becomes slower, it would mean that the semiconductor stocks would face a plateau in their revenue as well. Now the third point is the tenants of the data centers capitulating. Now one of the main tenants of the data centers are LLM service providers like OpenAI, Anthropic and etc. Now these companies need to show financial strength in order for the market to have the confidence that the data centers which are already built out will have strong tenants. Now obviously these companies are rarely making any profits yet but in order for them to continue to raise debt and equity they need to show extremely strong revenue growth. Now from that sense Anthropic is actually demonstrating this very strongly. They just projected their first operating profit of around $550 million on $10.9 billion in revenue for the quarter ending June 2026. Now, this represents roughly 130% growth from the $4.8 billion they reported in the prior quarter. So, this is a phenomenal growth for Anthropic. However, OpenAI is bleeding cash due to the massive cost of serving free users. Now, while they hit an annualized run rate of $25 billion in early 2026, which is roughly $6 billion per quarter, they're projected to post an operating loss of exceeding $7 billion this quarter alone. Their cash burn is projected to hit 27 billion for the full year of 2026. Now, one thing I'll say on this point is that in order for the scale of the data centers to make sense, all LLM companies need to demonstrate strong growth. The market cannot afford to have a loser right now.

[00:17:14]Basically, the market is thinking that the entire LLM universe will have enough demand to serve the trips and the data centers. But if the market somehow starts to think that there's going to be losers within the service providers, that would basically mean that the data centers and the trips which are being projected at this stage may be an over supply. And once that mentality hits the market, the market may face a catastrophic fall in the share prices of the relevant companies. So we just need to keep in mind that the market is basically assuming that there's going to be sufficient demand from all of us for all players so that all data centers capacity can be utilized. Okay. So today we talked about some of the risks of a market crash. Now my view still holds the same. I don't think I see any immediate signs, but I am monitoring with my utmost focus on whether there are any slightest signs of things going the wrong way. So, let's monitor the market together very closely. I hope you enjoy the video and I'll be back with more videos very