Thursday, July 24, 2025

AI is real and so is the bubble

Nowadays you can't avoid seeing the word "AI". AI is everywhere. Every company is rapidly adopting AI. Google today just released its earnings results which saw the top and bottom line beaten by a significant margin. But what worries some investors is the capex for 2025 is projected to grow to US$85bn. That's right. The Mag7 is spending billions and billions in AI infra buildout, front-running the anticipated accelerating AI demand. 

AI is data and power hungry

Remember that AI infra is expensive due to the high-cost GPUs and sky-high energy demands (nuclear energy is even needed). Training an LLM or AI deep-neural model requires massive datasets and energy , not to mention the associated R&D and talent costs. Data centers need large-scale cooling systems which require massive energy. Companies are poaching AI talents with multi-million-dollar salaries. They are also building nuclear power plants or leasing energy from them.
 

Who will be the winner?

Like all tech race, eventually there will only be one or two winners. In the 1980s personal computing era, Apple and Microsoft emerged as the winners and also amongst the largest companies as of today. In the internet era, Google emerged as the sole search engine winner. In social networking, Facebook emerged as the sole winner. You get the idea. What will happen to the losers? They will have to write off/divest their AI infra investments. Startups shutdown or consolidate one by one. The cloud revenues of Mag7 slowly taper off/decline as demand from startups fade away. The entire ecosystem collapses like a domino. There will be an oversupply of GPUs and you know what happens next. And then... the stocks will plummet and the bubble will eventually burst when investors become disappointed and disillusioned. 

Is it so simple?

Don't get me wrong. AI is real and does produce real productivity gains. Our lives and work will be completely transformed by AI forever. You can't put the genie back into the bottle. AI is here to stay. However, the bubbly part of this whole story is there are too many players -- OpenAI, Mag7, Perplexity, Anthropic, Mistral just to name a few. In any gold rush, most will be burned miserably at the end.  Any new tech will take at least a few decades to play out. Think of the personal computing and internet. After the first wave of AI hype cycle dies down, the winners will survive and continue to grow and dominate from there.

Parallels with the dotcom bubble

History never repeats but it often rhymes. There are many similarities with the dotcom bubble -- investors chasing after AI-related stocks, the Mag7 propping up the entire market, companies spending billions on infra anticipating accelerating demands that may never materialize so soon, sky-high valuations (all-time highs), most companies mentioning "AI" and related buzzwords during their earnings calls, absurd claims like AI will replace 50% of all human jobs. Nvidia today is basically the Cisco and Sun microsystem of 2000s. So much capex but with potentially diminishing returns. 

A perfect storm?

With that said, the stocks may go even higher for longer. Nobody will know exactly what happens next. The bubble may burst in the next few years -- either violently or gradually deflate. A macro environment with high interest rate, slowing economy and tariffs may provide the catalysts.

Other risks

The other risk is the invention of more efficient algorithms that require less data, GPUs and energy. Recall the DeepSeek "flash crash" earlier this year. This is a very real risk. Any breakthrough in research might cause another "flash crash". At the core of this AI revolution is the LLM which is basically the transformer deep-learning architecture with reinforcement learning. Before transformers, we have Recurring Neural Network and Long-Short Term Neural Net. So it is expected that active research will produce newer models/algorithms. The main problem with AI is that you'd have to constantly spend billions in R&D and training the models and to replace the GPUs every few years, otherwise you risk losing out to competitors. There is a lot of peer pressure to increase capex each year. As a result, it is very difficult for startups like OpenAI to be profitable. Before startups can even recoup the capex, they'd have to spend billions to train new models/algorithms. It's almost a never-ending game. Most companies are trapped in this AI race where they have to keep spending and spending to chase after finite demand.

Adding fuel

The US government has recently announced billions of investments in AI infra and made deals with nations in Middle East to build out sovereign AI. This further adds fuel to the fire. 

Warren Buffet

Is that why he has a record high amount of cash? Is he anticipating a bubble burst anytime soon? People keep saying don't ever time the market and he also preaches the same. But the truth is, he does market timing sometimes. 

Conclusion

  1. There are many big-name startups and smaller startups that drive up demand for AI infra, energy and cloud computing. They are all loss-making and burning cash rapidly to play in this very expensive game. Perhaps the most expensive out of all tech revolutions so far in history. The scale is unprecedented. And Mag7 and other big-names are still increasing capex exponentially. 
  2. When most of these startups die off after a few years, the demand for AI infra / compute may taper off/decline. There might be an oversupply when that happens. Whether the demand for these resources can be absorbed by other businesses remain to be seen. 
  3. To be profitable, a big-name startup needs to be dominant in its market, or at least get a lion's share and control its costs well, otherwise it is hard to compensate for the high burn rate. But given the massive competition, it is difficult to do so. Alternatively, raise prices for its products/services, but that may decrease demand. 

Who would be the winners?

I predict the incumbents -- namely the Mag7 will ultimately be the winners, not exactly sure which one but one or two of them,  as they are cash-rich and have an existing moat and ecosystem around their businesses. They are better positioned to weather any bubble burst. AI will enhance their existing products which will further improve their unbreakable moats. Microsoft is still the software king that has proven itself capable of integrating AI across its products. Google is also well positioned with deep research, big data and compute resources. It has access to an existing huge amount of data (Internet) for training its models. 





Monday, June 30, 2025

Liberation 2.0?

 As mentioned in my earlier posts, July 9 will determine the "final outcome" as the 90-day pause ends. My prediction as I have mentioned in Apr, is that after deals/pause, most countries will still be hit with reciprocal tariffs on top of the 10% baseline. There's not much reason to extend the deadline for most countries since the admin has no resources to negotiate with that many countries. They have said they had sent out "take it or leave it" offers to most countries but it will be tough for those countries to just accept without any negotiations. Nevermind about the smaller trading partners, the key is to watch whether there will be any extension for the top 10-15 trading partners.

Just today, there is a news article saying he's not intending to extend the deadline for countries with no deal. 

No deals yet?

So far we only have a "deal" with UK (one of the smallest trading partners). As predicted before, the economy has shown signs of cracks before any real deal.  

The tariff situation can only get worse from here. The level currently we have, is likely the best case scenario in the near-term. 

I think he is likely to bring back some of the tariffs otherwise countries will think he is "TACO" and won't take it seriously. On the flip side, he may change his mind. There might also be last-minute deals. So nobody really knows what will happen. Countries that are highly dependent on exporting to US may feel the pressure to accept a deal even if it is not in their favor.

I will just do some hedging.

There's also a tax cut bill that is intended to boost the economy by resorting to debt-fuelled growth again, the intention is to outgrow the debt and counter the impact of tariffs. Whether that can be done remains to be seen. Again, with the tariffs and the bill, how the market would react is stil an unknown.

Thursday, June 5, 2025

Next few weeks will be critical

 Previously, I mentioned that there is some euphoria/FOMO (Fear of Missing Out) in the market. Look at meme stocks like PLTR, HOOD, etc. and other AI-related names like APLD, Broadcom, CoreWeave, you will see they have ran up very quickly, almost vertically / exponentially with some of them forming a double-top pattern.  Stocks in general are fully priced, the market is being very optimistic right now. AI-related stocks are being chased aggressively ever since META and MSFT reported good earnings back in April, and of course Nvidia in May. From GPT: Magnificent Seven—Apple, Amazon, Alphabet, Meta, Microsoft, Nvidia, and Tesla—hold a significant share of the S&P 500. As of May 2025, they account for 34.1% of the index. Partly, the S&P index is being held up by Mag7 stocks due to the continuing AI fever.  AI is still in the early stages. Every tech revolution takes at least 10+ years to fully play out. Think of the internet, mobile revolution. There is immense potential for self-driving, robotics, AI-enabled software, etc. and investors have gotten ahead of themselves currently. Investors will be richly rewarded but the path to wealth will be treacherous and volatile. 

Economy is showing cracks

While the market can potentially creep up gradually, I'd be very cautious here and not allocate too much capital at this time. I used meme stocks  as a gauge of bullishness. When some of them form a double-top pattern (Mar and June), it is difficult to imagine them going up any further, thus there is an increasing chance of a strong pullback in the near future. Previously I mentioned the impact of tariffs will slowly be felt over the next 2 months, and we're starting to see some of the impact. I also mentioned we will see worsening economic data before any major trade deal (UK one is not really counted as it not one of the largest trading partners). The latest ADP payroll jobs data showed fewer-than-expected job growth. Producer price index keeps creeping up which would eventually translate to inflation or hurt corporate earnings. Manufacturing sector has been contracting for the past few months. Businesses are still hesitating to hire aggressively or make any long-term investments due to uncertainty around tariffs. Large retailers like Walmart, etc. warned of price increases and other retailers warned of declining profitability. 

Financially stretched


Consumer credit card debt is at an all-time high, and deliquency rate is increasing since 2022. Any slowdown in consumer spending will hurt economic growth. In fact, thrift stores like Dollar General have seen an uptick in customers from higher-income households, seen as a sign of economic uncertainty. Half of Americans live paycheck to paycheck and inflation will hurt their wallets and spending power. 

From GPT: 
more Americans are cooking at home as economic concerns grow. According to Campbell's CEO, home-cooked meals have reached their highest levels since early 2020. 

At the same time, Procter & Gamble (P&G) has noticed that shoppers are cutting back on laundry to conserve detergent.

This shift suggests that households are becoming more cautious with their spending. Things don't look good when the consumers even have to ration their detergent, etc. Lower-income households account for a significant portion of the GDP and cutbacks in spending will hurt the growth.

Why isn't there a crash or strong pullback yet?

Despite the bad economic data, it is quite surprising there is no crash / correction yet. I believe there are a few reasons: 1) Data is still considered not that bad yet or it could be somewhat priced in. 2) In a wait-and-see mode for trade deals, tax cuts which may boost growth or reduce tariffs. 3) Hoping that tariffs will be rolled back or reduced when there are signs of economic trouble.  4) Recession may be unlikely although there is a significant risk of stagflation. 5) market is still hanging onto lagging data from Jan-Apr (cooling inflation, stable employment, etc.)

The man has requested for countries to send in their best offers and by the end of the month, we should have more clarity on the "final outcome". Also there will be 20/30 year treasury auctions which may cause bond yield to spike. That's why the next few weeks will be critical. 

Debt problem 

I believe there is no practical solution to the budget deficit/debt problem, at least not a solution that a politician/president will implement without committing political suicide. The projected deficit is $1.9 trillion for 2025, which has been increasing steadily over the past few years. The only solution is austerity, higher taxes and severe budget cuts which will crash the economy and market. 

With higher-for-longer Fed rate, and will remain so due to the incoming inflation, the cost of servicing the deficit/debt is around 3.2% rate.  It may even go higher as the older debt gets refinanced at current Fed rate. They have to borrow money (by issuing treasury securities), to pay the debt interest. So the debt can only snowball from here.  It does seem that the debt will grow faster than the real GDP which is concerning. In 10 years, the debt-to-GDP ratio may reach 150%, from the current ~120%. If there are any crisis in the meantime which requires them to print money, it may go up even more.  

Lost decade?


With a debt problem and a potential stagflation, will we see a lost decade similar to the one from 2000 to 2008 (dotcom to GFC) where the S&P remained largely flat? From 2008 to 2021, the bull market has been phenomenal, largely because of monetary stimulation (money printing) and zero interest rate. 

What I will invest in

Should there be another dip which I think there will be in the near future, I'd allocate more heavily into AI-related software names and avoid any companies impacted by tariffs. The rationale is  AI can only keep growing in the future due to : 1) the 2 largest economies are locked in an AI race, and they are investing very heavily into AI infra/tech. 2) Tech will always progress and improve. 3) Tech has been the primary driver of economic growth for the past 4 centuries and beyond, and hence tech companies have consistently been able to capture a larger share of the growth.   

We're still early in the revolution, and thus choosing the right companies is crucial. In tech, it is usually the winner that takes all so we want companies that can remain competitive and outlast everyone else. Looking back at the Search Engine 1.0 war around the dotcom era,  there were many search engines (AltaVista, Yahoo, etc.) but ultimately Google won the race. Likewise, you know who won the social network war (FB, Friendster, MySpace, etc.). Google's dominance is now being threatened by new challengers in a Search Engine 2.0 war. 

Inflation may remain sticky, and interest rate will remain high, so choose companies with strong balance sheets that have pricing power. However, with that said, I may still allocate a small amount of capital to emerging tech such as quantum computing, etc. 




Wednesday, May 14, 2025

Euphoria in the market

 After the announcement of the pause in the trade situation, the markets went all euphoric. The fear/greed index right now is extreme greed.  This index is one of my favorites as it guides me when to buy/sell stocks. Even if you're a long-term holder, it might still be useful. The market is also overbought. I'd say this is the time to be cautious. Remember this fact, based on years of historical data: no market goes up forever, especially when it goes up so high so quickly. You can look at every chart you can think of, when a stock/index goes up so much so fast, eventually it will come down. The converse also happens, the bigger the crash, the higher the eventual rebound, which is what we're witnessing right now. I'm not saying this is the time to sell everything, but personally I'd take some profits and start to do some hedging. Any bad news can cause the rally to stop or even drop. My belief is that even though the liberation day losses are erased, it may not go back to all-time high. There is still some risk of stagflation/stagnancy even though there might not be an outright recession.  

Difference from 2018-2020 Trade Situation

It is likely the 30% on China and 10% universal on major trading partners will be the 'best-case' baseline going forward. Not to mention there are also 25% on auto, steel, etc. Possibly more incoming. Also we have high interest rate, and a record high of consumer credit card debt. Note that 70% of U.S. economy is based on consumption. A weakness in consumer spending spells trouble, and at the beginning there is already some warnings from consumer staple companies/retailers regarding this.  No doubt some of these may be cushioned by other areas such as tax cuts, investment growth, AI capex, etc. But it remains to be seen how it will play out over the next 3-6 months. 

Monday, May 12, 2025

A toothless tiger, nobody will take this trade war seriously

My analysis for the past few posts are thrown out the window. The announcement to mutually reduce tariffs by over 100% shows their weakness. Originally it was thought they would at least put up a fight, but who knows their pain threshold is so low? They have shown themselves to be a toothless tiger, big roar but no bite. Totally just caved in front of China. China didn't even have to make any concessions or uphold previous deal from 2018-2019. If I were China, I wouldn't take it seriously and will even take a tougher stance. Expect the same from other trading partners, I wouldn't be surprised if most of them starts to play hardball. This on-off nature is exhausting to investors/traders/businesses. 30% won't stop them from importing China goods, so in the end they will end up on the losing side for paying the additional taxes. A totally meaningless situation that does more harm to themselves. Again expect inflation to creep up over the 2-3 months, though it will be less than forecasted, probably won't be high enough to cause any real pain. Fed might even consider raising rates, who knows? Q2 GDP numbers might be weaker than usual again as businesses might do some front-loading/stockpiling again before the 90-day pause ends in Aug. 

My conclusion is: no recession, no bear this year unless of course there are some drastic escalation again. I will just invest/trade normally. All these tariffs are just pure noise. Any dip is a buy. Even if he announces 100% pharm/semi tariffs, it is still a buy opportunity because we all know it will just be noise eventually. If there is any economic damage, Fed / government will step in to save the day. So no worries of any doom / gloom. 


China emerging as the leader of the new multipolar world order

They can't even win a trade war, or whatever war. To make weapons, they have to rely on China's materials. Therefore they can't even win a hot war. I will look to accumulate China equities when it dips a little, and diversify a little.


Is this 2022 all over again?


Update: some of the below points may no longer be valid due to the recent trade truce.

This is going to be a long post. Many things have happened for the past few weeks. To understand where the market is heading, we need to understand the reasons behind the trade situation as stocks are increasingly mixed with geopolitics regardless of whether we like it or not. This trade situation is much more serious and broader than the first in 2018-2019. Previously I have discussed the motivation behind the trade situation, based on my analysis of the many articles I have read. Let's recap, the objectives are:

1. Manufacturing self-reliant. They want to be able to manufacture their own critical products without relying on foreign countries. Example, semiconductors, rare earths, steel, automobiles, etc. That's the reason why they have wooed Taiwan Semi TSM to setup factories in the homeland. Also that's why they have auto tariffs to protect domestic auto industries, especially against the flood of cheap foreign EVs. What's the reason? In the event of conflicts, they must be able to manufacture their own products, supplies, etc. 

2. Reshoring jobs for the lower-income groups. To re-industrialise some critical sectors and buildup the know-how. 

3. Counter the second largest economy, similar to how they dealt with their main opponent during the first cold war. Also, how they dealt with the rise of Japan from 1960-1980s.

4. Boost government tax revenue for other initiatives like tax cuts, etc. Reduce debt load and the deficit. Boost GDP growth via increasing exports globally. 

However, it is not smooth-sailing to achieve the aforementioned objectives. There will be many hurdles ahead, and there will be economic/non-economic damage in the meantime. But they will still try to do it anyway, as they are very determined to solve the above problems that they perceive as critical. Some of these objectives are not entirely new, they are a continuation from the previous administration. Recall the CHIPS act which was setup to woo TSM to come to the homeland? Every administration has a different style of executing the objectives. But also each administration has some of their own unique objectives due to their governing philosophy (for example tax cuts).  

Final outcome

There are 3 camps here: (1) believe that the troubles will go away with the deals, and the economy will be back to normal, just like in 2019. (2) deals won't make much difference and there will still be substantial headwinds. (3) those in between. 

I am leaning towards (2). With the objectives in mind, I think the final outcome, even after all "trade deals" are done, will be 10% universal on most trading partners, 20-30% for "average" partners and 40-60% for the "worst", proportional to the trade deficits. This is inline with what many experts have mentioned. Think about this: if the second largest economy gets only 20-30%, there is no real need to negotiate or be too serious about it. So they have to come down hard with a much higher level.
 
Recently, there has been a small UK "deal" which many experts dismissed as more of theatrics, rather than any major win. It only affects $6bn worth of imports which is tiny considering the trillions of total imports. Of note the 10% baseline remains,  and during the presentation, they also highlighted the "external revenue" which means they really want to use tariffs as a source of revenue, so all the more the baseline won't go away even after the 90-day pause.  

Deal or no deal?

A real trade deal takes many months or even years to be negotiated. It is unrealistic to expect any substantial comprehensive deal to be signed during the pause. Experts have said that it is more likely for "frameworks" or "MOUs" to be signed instead which might be re-negotiated down the road. 

Fed rate cuts

Analysts have priced in 2-4 rate cuts in the second half of the year to boost economic growth and counter unemployment. The Fed will be in a tough spot since there may be higher inflation and unemployment. Their dual mandate is to maintain low inflation and strong employment. If they cut too soon, inflation may worsen; cut too late, may risk a recession. Indeed, the chair has said recently the risk of stagflation is getting higher. The Fed is not perfect. They have made mistakes before, example during the pandemic when they thought the inflation was transitory and in the end they raised rates too late and had to rush the raise in 2022. This time round,  I think it will be even more difficult not to make a mistake. They also prefer to adopt a 'wait-and-see' approach which means there might be a chance they act too late. 


More coming soon?

There might be more coming: pharma, semi, minerals and others. "All-rounded" to cover almost every critical sector to achieve the earlier objectives.

Market reactions

Meanwhile, every piece of good news, no matter big or small, will get the market excited. To me, the deals don't really matter since they simply point the way to the final outcome which might have been priced in already to a large extent. The market may slowly inch up while bad economic data pulls it down. We have already seen this on Apr 30 where worse-than-expected data caused the market to drop 2% on open.  This coming week, experts have already forecasted inflation to creep up gradually. So bulls and bears are fighting over the next 2 months. I think ultimately, there may be another leg down, unless the trade situation is scaled down significantly very soon. Eventually the market may realise the reality -- higher inflation, increasing unemployment, lower growth, etc. And then there may be headlines running fears of stagflation/recession.  This year may be similar to 2022, a sideways market, but whether it will trend down or up, depends on how the trade and economic situation evolve. 

Update: some of the above points may no longer be valid due to the recent trade truce.



 

Monday, May 5, 2025

Is a bear market coming?

Major indices have staged a remarkable comeback in the past 10 days and recovered most losses since Apr 2. However, whether or not this rally will continue as per normal into a bull market depends on several factors. It is difficult to know whether you're in a bear until you see one / strong signs of lower highs and lower lows, but by that time, it will be too late. Nobody would know for sure and it's always in hindsight that you will realise this is a bull or bear. Even in a bear, there will be strong rallies like this one, example in 2022, where the market is whipsawing while trending downwards. 

Motivation behind this trade situation

To understand where the economy and market is heading, let's examine the movitation behind the current trade situation. I think their main objectives are (1) reshore manufacturing of critical products so as to reduce reliance on foreign countries. (2) bring some jobs back for the lower-income groups (3) encircle/isolate the second largest economy. (1) is especially important as they are currently overly reliant on foreign imports for critical materials and goods.  There are many articles talking about (1).  (3) is just repeating the playbook for how they successfully isolated that "well-known" country during the first cold war which led to its collapse. Now they are partnering with the largest populated country to counter the second largest economy.  Overall, this is a very dire and serious situation for them, as (1) and (3) concern their survival. There are also secondary objectives like increasing their exports to boost GDP and to reinforce their current dominant position in the global economy, thus ensuring their continuous relevancy. With this context in mind, it is likely that they won't give up so easily, that is, this trade situation will last for several months at least, although there might be some de-escalation which may not come soon enough to fend off any economic fallout. 

Any deals soon?

There have been a lot of news regarding talks, etc. Also Australia and Canada have just concluded elections recently and their new leaders are not so friendly towards "them". Hence it is likely talks with these countries will take time and possibly met with obstacles. Even traditional friends like Japan and South Korea are expecting a deal only until July, and there are still hurdles to be ironed out. So deals may not come so soon in the next few weeks, except for India which could be one of the first to sign. Overall I don't think there is much incentive for these countries to conclude a deal so quickly before July. 

Economic data

The April jobs report was released in May and appeared better than expected. Of note, the "nonfarm payroll" of 177k was better than the expected 130k. I'd take this with a pinch of salt as this number is based on surveys which could be inaccurate. In fact, the numbers for previous months have been revised downwards subsequently after being released.  From GPT:

Nonfarm payroll data is a valuable economic indicator, but it does undergo revisions that can sometimes be significant. The Bureau of Labor Statistics (BLS) releases initial estimates based on survey data, which are later adjusted as more comprehensive information becomes available. These revisions can reflect seasonal adjustments, updated employer reports, or benchmarking against more complete datasets


Anyways, data for Jan - Apr are backwards looking and largely untainted by the trade situation. There is also distortion due to front-running by stockpiling. Going forward, experts have forecasted higher prices/inflation, supply chain disruptions, shortages and even mass layoffs due to slowing trade between the 2 largest economies. I think the full impact will be felt during the next 2 months. Such a scenario will be tricky for Fed to cut rates.  If inflation goes up but labor is still OK, the Fed may not cut rates. 

De-escalation

Time is of essence here to minimise ecnomic damage. I think the longer  this drags, the bigger the damage. Previously, Walmart, Target and Home Depot have sought reliefs but did not get any. There are some exemptions already but going forward, I don't think either side will de-escalate (in a meaningful way) unilaterally without any mutual agreement/deal, as neither side wants to be seen as "weak", and they are determined to achieve the aforementioned objectives. I think both sides will need to talk for at least a few weeks before any mutual de-escalation happens. Any comprehensive/serious talk will take several months at least. One side wants to exert max pressure and would not lower voluntarily unless there is some serious situation (example, bond markets, economy crash). Even if there is de-escalation, how much lower can it go to? 50%? 100%? 100%  is still very high and may not reverse any damage. 50% would likely increase inflation and drive lower growth without causing a recession. 

Conclusion

Since the current rally is based on (1) hopes of de-escalation (2) backward-looking economic data and earnings, going forward, without any meaningful de-escalation / rate cuts, the market may likely drop as new economic data worsens over the next several weeks.  The speed and magnitude of de-escalation are both important in determining the extent of future damage. Also, given how furious this rally ocurred, it will need continuous positive catalysts to sustain, otherwise it's likely the current positive sentiment gradually wanes off. Further, deals and de-escalation are unlikely to materialise so soon to support the sentiment. 

Various outcomes are possible: outright recession, high inflation with low growth, etc. I can't imagine any positive outcome, unless the trade situation is entirely resolved, which itself is unlikely. Whether the market has truly bottomed depends on how the trade situation evolves. 




 

AI is real and so is the bubble

Nowadays you can't avoid seeing the word "AI". AI is everywhere. Every company is rapidly adopting AI. Google today just relea...