Echoes & Alpha — The Signal in the Noise

The Roaring '20s

Sun, 26 Apr 2026 20:23:00 GMT

If you had told me 8 weeks ago at the start of the Iran war that the Strait of Hormuz had been, and effectively continues to be, completely shut down, that the latest core inflation numbers came in high at 3.3% and that was before the oil price increases really started to hit, that jobs numbers for March came in well below what they were in 2025, and that there still has been no official peace treaty signed for that war, I would have said that all wasn't very surprising (in fact all of my previous posts predicted pretty much exactly that). Then I would have asked how bad the markets were and when you told me they were trading at all time highs I'd have called you a liar. Yet here we are…

What's Driving This Exuberance

Insanity? Extreme optimism? A cabal of shadowy forces? The US Leading Economic Index is well below the lowest it hit during the COVID pandemic and is continuing a downward trend. Oil prices remain significantly elevated compared to their pre-war prices as Brent Crude still sits above $100 / barrel. The Fed has held rates steady and signaled they are unlikely to make any cuts in the foreseeable future, when at the start of the year the market was supposedly pricing in at least 2-3 rate cuts. The Buffett Indicator (one of my favorite metrics) is currently over 200% signaling that we are in a market that is significantly overvalued.

The current reasoning the big Wall Street analysts and prognosticators are giving is that it's because companies continue to crush earnings and there's no slow down of that in sight. There's something to that line of reasoning. We're seeing insane numbers coming in from the big tech companies, in terms of revenue, profits and what they're spending on AI. It does seem like the AI gravy train is storming down the tracks and there's no let up in sight. Every week there are new models dropping with increasingly powerful capabilities, there are new start-ups showing up on the scene promising to disrupt every sector of the economy while garnering eye-popping valuations on nothing more than a dream, and there's even talk of a nearly $2 trillion IPO coming up.

It's a roaring party out there! Why shouldn't markets be at all time highs! With AI we're almost certainly going to continue this trend and pricing that in now is the only rational thing to do. Maybe we should start calling this time period the roaring 20s!

While big tech is popping champagne and smelling their own farts about how great of a future they're building, in the background they are shedding thousands of jobs to keep those good times rolling. The public backlash to data center construction is rising and the US debt is continuing its march to 200% of GDP with no political will or impetus to stop it. The circular deals between these very same companies continue, growing ever larger all while investors continue to cheer them on.

And the logic that we're at all time high because earnings reports continue to be stellar is so flawed it's laughable. Earnings reports look backwards. They capture how things went. And sure, sometimes companies will give guidance out into the future that can be a spark for optimism and continued investment but fewer and fewer companies are doing that and the ones that are, are typically only providing it for the next quarter or two.

The Future That's Coming

We've seen this happen before. The market becomes so exuberant that it acts irrationally to everything that's going on in reality. People start to believe that there really is no stopping this market and that they would be foolish to miss out on all the growth that's going to happen so they continue to pile in. We've been trained to do this from the recovery after the 2008 financial crisis, to the V-shaped COVID market, to the 2025 TACO trade after the tariff liberation day sell-offs.

I highly encourage reading this New York Times article that was written in 2021 and ask yourself how things have played out since then and how they compare to the trends in the 1920s. How often do you find yourself thinking “Wow, that's eerily similar”?

History Doesn't Repeat Itself, but It Often Rhymes

Mark Twain

I'm not saying things are going to go exactly how they did back then and that we're going to have a 1929-esque meltdown in the not-so-distant future. We have built some better guardrails on the whole system and have a lot more tools, technology, and policies in place that can help to mitigate a complete collapse. But what I am saying is that we are in the middle of the party, the music is bumping, the drinks are flowing, but the world outside that party is very murky and things aren't trending in a good direction. Eventually the music will stop, the party will end, and everyone will need to go outside and take a look at what they were ignoring while they were having their good time.

Rapid Fire Observations

Some quick top level thoughts on other topics:

Claude Mythos: Lots of talk recently about Anthropic's Mythos model and how it's a game changer in terms of capabilities. I don't doubt the claims that are being made about its abilities. I am glad that Anthropic is being somewhat cautious about releasing this model and giving some companies the opportunity to harden their cybersecurity defenses with it. But I'm not convinced that this model will be the reckoning that some are worried about. First, I expect this model will be extremely costly to use and that will limit the amount of organizations, nations, etc. that will be able to use it. That limit will make it easier to monitor its usage and spot bad actors and help to prevent major fallout. Second, I'm not fully convinced the model will even see the light of day. If Anthropic is as compute constrained as they claim to be why would they utilize their capacity to host a model that is more a liability to them than a revenue and profit generator? I'm not going to sweat this one yet.
AI's Leading Regulator is…the Pope?: File this under the amusing but also scary when you think about it bucket. There's been a few reports recently that the Vatican has been moving quickly to put in place guidelines and rules to prepare the Catholic Church for the AI era. It makes sense that the Catholic Church would want to get ahead of AI and what they are calling a “crisis of truth” especially given their mission. It's hard to imagine building faith in what a spiritual leader is preaching if one were to find out that it was actually a robot that was coming up with their teachings. Pretty dystopian actually if you think about it. It's a sad state of affairs however when one of the most ancient organizations on the planet, one that is notorious for being extremely slow on everything, is acting faster than most governments across the world.
Kevin Warsh's Fed: Kevin Warsh, Trump's nominee to be the next Fed Chairman when JPow's tenure ends in May, had his confirmation hearing this past week. It's looking very likely that he is going to be confirmed so what can we look forward to during his tenure? One of the most consequential things that came out of his hearing was his thinking and approach to the way the Fed communicates. He believes that the Fed communicates too much, and that it should avoid providing forward guidance (aka the dot-plot) as he believes that hems the Fed into future actions. Not sure I agree with him on that one, but he also did say he believed the way the Fed calculated core PCE needs changing, which I do agree with. But I don't agree with the way he wants to change it. He wants to strip out “price shocks” to try and get to a core underlying inflation rate which to me is just academic nonsense. Inflation is what the actual consumer or business faces when they have to pay for it, regardless of if it's a “shock” or a long term underlying trend. His approach will most likely understate inflation so that he can use it as a justification to lower interest rates to appease the president which will do nothing to actually curb the inflation that the average person sees, further resulting in a divide between Main Street and Wall Street.

The Last Word

We're living in interesting times indeed. Most indicators seem to be showing that things are dark and stormy but the stock market is all sunshine and rainbows. There's so many adages for this type of situation it's nauseating but the one that holds the most weight in my opinion is:

Markets can remain irrational longer than you can remain solvent.

John Maynard Keynes

Things are glaringly irrational in my opinion, but it's a fool's errand to try and guess when that irrationality is going to end and sanity will come back and take hold. It's likely it won't be for some time. Large economic changes, much like geological changes like those that produce volcanoes, take a long time to show up in ways we can observe. But when they do, much like those volcanoes, they often erupt with extreme consequences.

Sinking Growth and the Great AI Cover-Up

Sun, 15 Mar 2026 20:32:00 GMT

Two weeks into the war with Iran and the fog isn't lifting, it's thickening. Last week was defined by miscommunication, confusion, and escalation. Conflicting reports on the status of the Strait of Hormuz whipsawed markets for days, only to be followed by US strikes on critical oil infrastructure at Kharg Island (Iran's primary crude export terminal). Oil has been hovering around $100 a barrel for most of the week and the VIX is up nearly 37% since the start of the conflict. If the geopolitical chaos wasn't enough, Friday's GDP print came in at a glacial 0.7% and the latest CPI report showed inflation ticking back up. We are staring down the barrel of a stagflationary environment and the market knows it. Meanwhile, the tech sector is dealing with its own storm as a wave of heavy layoffs rolls through the industry under the thinly veiled banner of AI efficiency. There's plenty to unpack this week, so let's get into it.

The Macro Picture is Getting Ugly

Friday's second estimate for Q4 2025 GDP came in at an annualized rate of 0.7 percent, a revision downward from the previous estimate (which was at 1.4%). For context, Q3 posted at 4.4 percent. Consumer spending and investment were the only things keeping the number in positive territory at all, while government spending and exports both contracted. Imports (which subtract from GDP) also fell, which normally would help the headline number but clearly did not do much heavy lifting here.

The natural instinct is to look at 0.7 percent and start to consider a potential recession looming. I am not quite there yet but the trajectory is not encouraging. The economy went from running hot to barely moving in a single quarter and that kind of deceleration tends to carry momentum in the wrong direction.

Now layer on Wednesday's CPI print. The Consumer Price Index rose 0.3 percent in February (up from 0.2 percent in January). Shelter was the largest contributor to the monthly increase, which is the same story we have been hearing for over a year now. Food prices jumped 0.4 percent on the month. Energy rose 0.6 percent (and with oil sitting at $100 a barrel that number is only going one direction from here). The year over year all items index held steady at 2.4 percent and core CPI (all items less food and energy) came in at 2.5 percent year over year.

On the surface those annual numbers look almost boring. They are basically flat compared to January. But that is precisely the problem. Inflation is not coming down anymore. It is just sitting there, stubbornly refusing to cooperate with the Fed's timeline. And that was before oil became a triple digit commodity again.

Put the two prints together and the picture we're looking at becomes clearer. Growth is collapsing while prices refuse to budge. That is the textbook definition of stagflation. The Fed is now stuck in a position where cutting rates to stimulate growth would risk letting inflation run hotter, and holding rates steady (or raising them) would put even more pressure on an economy that is already losing altitude fast. There is no clean move here.

We're hearing more alarms being raised from more outlets and economists. Rate cuts are almost certainly off the table for the foreseeable future (although with the new Fed Chair coming in a few months who knows what they are going to do). The CME FedWatch tool has pushed rate cut expectations all the way out to December. That is a significant shift in expectations and one that I don't believe has fully worked its way through asset prices yet.

For those who are not familiar with why stagflation is such a dreaded word in economics, here is the short version. In a normal recession the Fed cuts rates, borrowing gets cheaper, businesses invest, consumers spend, and the economy recovers. In a normal inflationary environment the Fed raises rates, things cool off, and prices stabilize. Stagflation breaks both of those playbooks simultaneously. You have an economy that is slowing down and needs stimulus, but you also have prices rising in a way that makes stimulus dangerous. Every tool in the toolbox either fixes one problem while making the other one worse or does nothing at all. The 1970s are the most cited example and for good reason. It took Paul Volcker deliberately engineering a deep recession (pushing rates to 20 percent) to finally break the cycle. That is the nuclear option and nobody wants to go there.

So what is the right move in an environment like this? Historically, stagflationary periods are tough on the traditional portfolio. Both stocks and bonds can lose value at the same time, which means the classic 60/40 split offers very little protection. The assets that tend to hold up are the boring ones. Companies with real pricing power (the kind that can pass cost increases to customers without losing them), energy producers who benefit directly from elevated oil prices, and hard assets like commodities. Cash also becomes genuinely attractive, not as a long term strategy but as dry powder. When the market eventually finds a bottom (and it will), having capital ready to deploy into quality names at distressed prices is where the real opportunity lies. I am continuing to hold cash and will look to deploy it in tranches toward companies that are insulated from oil pricing risk (particularly in software) once there is some clarity on direction.

The Convenient Scapegoat

While the macro picture deteriorates, the tech sector has found itself a neat little narrative to hide behind. AI is coming for your job. Or at least that is what every earnings call and press release would have you believe.

The latest domino is Meta

. Reports are coming out that the company is planning layoffs that could affect 20% or more of its workforce because, allegedly, they want to offset the cost of its AI infrastructure buildout and prepare for efficiency gains from AI assisted workers.

If this feels familiar it should. This would be Meta's largest restructuring since the back to back rounds in late 2022 and early 2023 that they branded the Year of Efficiency. Those moves turned out to be the starting gun for a wave of layoffs that swept across the entire tech industry. Every major player followed suit within months. There is no reason to think this time will be different. Meta tends to move first in these situations and the rest of the Mag 7 tends to follow once they see the market reward the move.

And the market does reward it. Block's shares have risen ~17% since announcing it was cutting 40 percent of its workforce, with Dorsey attributing the move to AI. The S&P 500 fell over the same period. That is the incentive structure at work. Tell Wall Street you are replacing humans with AI and you get a bump. Tell them your business is struggling and you get punished. Executives are not stupid. They know which story to tell.

When the Block news broke I wrote about how AI is genuinely displacing white collar workers and how the people downplaying it are not looking at the data honestly. I stand by all of that. The job growth numbers are propped up almost entirely by healthcare. New grad unemployment is climbing. Software engineering grads are sitting at 7.8 percent unemployment. The pain is real and it is accelerating faster than our economic systems can adapt to. None of that has changed in the last few weeks.

What has changed is the volume of companies now rushing to wrap themselves in the same AI flag, and a growing body of evidence that suggests a lot of them are full of it. Forrester published a report in January that put it plainly: many companies announcing AI related layoffs do not have mature AI applications ready to fill those roles. They called it AI washing. The firm went further and predicted that over half of layoffs attributed to AI will be quietly reversed as companies realize the operational challenges of replacing human talent prematurely.

According to this HR Executive post, 59% of companies admitted they emphasize AI's role when explaining layoffs because it resonates better with stakeholders than citing financial constraints with 9% saying AI had fully replaced certain roles. Even Sam Altman (a man who has every incentive to hype AI's capabilities) acknowledged the phenomenon, saying that some companies are blaming AI for layoffs that they would otherwise do.

So here is where I land on this. Dorsey was being honest (or at least more honest than most). AI is changing the economics of certain roles and that is not going away. But what we are now watching is every other company in the industry using that kernel of truth as cover for a decade of bloated hiring and bad capital allocation. Blaming tariffs risks political backlash. Blaming slowing revenue is an admission of weakness. AI is the one explanation that makes leadership look forward thinking instead of incompetent. A lot of bad decisions are getting laundered through a narrative that conveniently absolves the people who made them.

I said when the Block news dropped that I expected layoffs to continue rolling across the tech sector. That prediction is playing out and I expect we'll see a drumbeat of additional layoffs from the rest of the big players in the coming weeks.

Rapid Fire Observations

Some quick top level thoughts on other big things that happened last week:

Anthropic v. The DoW, Continued: The Anthropic saga escalated some more this week with Anthropic filing two federal lawsuits against the Trump administration, alleging the Pentagon illegally retaliated against the company for its position on AI safety. As I wrote last week, this is fundamentally about the precedent being set when a government attempts to bankrupt a company over a policy disagreement. What makes this even more absurd is the backdrop. The Pentagon has continued using Claude in the ongoing conflict with Iran to process intelligence and targeting data, even as it labels the company a supply chain risk. Read that again. They are actively using the product in a war while simultaneously trying to destroy the company that makes it. I said last week I was deeply pessimistic about the direction this was heading. Nothing that happened this week changed that. The first hearing is March 24. Watch this closely.
Gaming Is Getting Squeezed From Every Direction: This one hits close to home for me. GDC wrapped up last week (it is hard to believe it has been seven years since we unveiled Stadia there) and the state of the industry is grim. The AI boom is squeezing gaming from multiple angles at once. Data centers are expected to consume around 70 percent of global RAM production in 2026, which has created a genuine hardware crisis. Multiple panels at GDC this year were dedicated to how studios can optimize their games to work with less memory because the supply simply is not there. On the jobs front, roughly 45,000 gaming employees lost their jobs between 2022 and the end of 2025, with up to 10,000 more layoffs predicted for 2026. Studios are gutting junior roles and backfilling with AI tools, and the developers I have seen commenting on this are candid about the fact that people are going along with it because they are terrified of losing their own positions. The gaming industry is becoming one of the clearest examples of AI's collateral damage: it is not just taking jobs inside the industry, it is physically consuming the resources the industry needs to make its products.

What's Happening Next Week

The biggest event on the calendar is the FOMC meeting on Tuesday and Wednesday. This is one of the meetings that comes with a Summary of Economic Projections and a press conference, which means we will get the Fed's updated forecasts on GDP, inflation, and unemployment. The market is expecting them to hold rates steady. The real question is what the dot plot looks like and how Powell frames the path forward. If the projections show fewer rate cuts than the market is hoping for (or none at all) expect a selloff. If the Fed acknowledges the stagflationary dynamics without offering a clear plan, expect volatility to spike further. There is no version of this meeting where the Fed says something that calms everyone down.

Micron reports earnings on Wednesday after the bell. This one is worth paying attention to given the RAM shortage discussion mentioned above. Analysts are expecting revenue growth of 138% and EPS growth exceeding 450%, driven by AI related memory demand. Those are absurd numbers and they tell you everything about where the demand for memory is going right now (hint: it is not going to your next gaming PC). I will be watching their guidance closely for any commentary on how they are allocating supply between AI customers and the consumer market. If Micron signals they are leaning even further into datacenter at the expense of consumer products, the RAM shortage narrative gets a lot worse.

The Iran situation will continue to dominate. Any escalation (or de-escalation) at the Strait of Hormuz will move oil prices and, by extension, everything else. At this point I am not expecting any positive developments but I would love to be wrong.

The Last Word

The thread connecting everything this week is that the consequences are arriving. The GDP print is the consequence of an economy that was running on borrowed momentum. The CPI print is stubborn even before the full weight of $100 oil works its way through the system. The layoffs are the consequence of reckless hiring and a weakening economy dressed up in a shiny AI narrative. And the Fed meeting this week will be the consequence of all of it landing on Powell's desk at the same time. There are no easy answers right now. Keep holding cash, keep watching the data, and do not let anyone tell you this is priced in.

$100 Oil & Oracle Layoffs

Mon, 09 Mar 2026 18:08:00 GMT

The broader market took a 2% haircut after a full week of Middle East attacks spilled over borders to include Lebanon. It is looking increasingly likely that this conflict will spread further as we have heard of at least 2 situations where rockets/drones were intercepted or landed in Turkey (a NATO member). I'm going to focus on macro level things this week, since that's going to be the main driver for a while here.

Oil, oil, and more oil

Oil prices are spiking and went over $100/barrel over the weekend. Let's talk about why that's happening and what it means.

As mentioned in last week's post, the Strait of Hormuz is shut down (effectively cutting off 20% of global oil consumption). The meager increase in output from OPEC+ is not nearly enough to cover the lost supply. We also saw strikes on multiple oil depots in Tehran which prompted Iran to threaten retaliatory strikes on oil facilities around the Gulf. If Tehran follows through they will effectively kneecap the supply capacity of the entire region. This turns a short term logistics problem (getting ships moving through a strait) into a long term production problem (spending months repairing facilities to reach pre war levels).

The US government is attempting to staunch the bleeding by promising they will provide insurance and escorts for ships that transit the strait. JP Morgan reports there are real questions on whether they even have the capacity to provide the amount of insurance needed because the providing agency lacks the necessary funding from Congress.

All of this fear is pushing oil prices higher. What does that mean? In the short term it means higher prices at the pump. We are already seeing gas prices rise up 17% from when the war started. Given the political climate in the US around affordability and inflation this is a terrible sign. We are already hearing chatter about tapping into the strategic oil reserve to ease the pain on the average consumer. In the medium to long term this means the cost to produce goods will rise (pushing the price of nearly everything up with it). That stubborn inflation the Fed has been battling will likely continue. Rate cuts are probably off the table (although a new Fed chair coming in May could shake things up).

New Supreme Leader

On Sunday Iran announced they selected a new Supreme Leader, Mojtaba Khamenei. He is the son of the former supreme leader and early indications suggest he is even more of an extremist than his father. This is a clear signal that Tehran has exactly zero intention of backing down.

With a formal leader named Iran once again has a single person directing their armed forces. It's currently unclear what direction he will take but I am fairly confident that a man whose nearly entire family has been killed is not going to suddenly sue for peace.

What does this mean for us? Every reliable indicator suggests this conflict is going to drag on for a while. As I mentioned last week (and history readily confirms) a prolonged war will most likely lead to continued selling pressure across the broader market. Big banks are actively repricing their risk. They now have to account for a Federal Reserve that will likely abandon the rate actions everyone anticipated earlier this year. This reality will force a rotation out of speculative bets and into safe haven assets. Finding a safe haven that is not already wildly overpriced and painfully volatile (like gold) is going to be exceptionally difficult.

I am going to continue holding cash until there are actual signs of market direction. When the time comes I will deploy capital in tranches toward distressed assets that are insulated from oil pricing risk (specifically software).

Rapid Fire Observations

Some quick top level thoughts on other big things that happened last week:

Oracle's AI Conundrum: It was announced that Oracle is ending its plans to expand its largest AI datacenter (the primary customer of which was OpenAI). There is also reporting that Oracle is planning a colossal layoff to fund their continued datacenter buildout spend (with some outlets estimating nearly 30,000 employees will be impacted). Nothing signals a healthy company quite like halting a primary data center project while simultaneously laying off employees to supposedly fund other data center projects.
Anthropic and the DoW: The saga over Claude and its military application continued this week with Anthropic vowing to fight against the DoW's supply chain risk designation. It is entirely apparent to me that this is politically motivated and that egos at both the DoW and Anthropic are adding fuel to the fire. The primary concern here is the precedent being set by a government attempting to bankrupt a company simply over a disagreement on product usage. In a functioning free market the correct move is to cancel the contract and purchase from a competitor. Let the market decide their fate. I am deeply pessimistic about the direction this is heading and will actively monitor the outcome of this spat.
CrowdStrike Earnings: Crowdstrike reported their 4th quarter earnings and they crushed it. I mentioned last week that this would be a bellwether for how the SaaS industry is trending and they guided to increased revenues for their fiscal 2027. That's a positive sign, however I still expect that macro level impacts of the Iran conflict will override here but if there's a sell-off this is a stock I wouldn't mind holding.

What's Happening Next Week

We get our next CPI print for February on Wednesday. After the atrocious February jobs report last week any sign of increasing inflation in this report will inevitably exacerbate jitters and accelerate a broader selloff. We also get a second estimate for the 25Q4 GDP report on Friday. I will be looking to see what direction the revision goes. An upward revision could help calm the market. A downward revision means things will get exceptionally volatile.

Oracle reports earnings on Tuesday, where I expect they'll officially announce their layoff and datacenter buildout plans. Investors are on edge for any signs the AI bubble might be popping. If there is a place for the pop to start, Oracle is my most likely candidate.

Dollar General also reports this week on Thursday. I follow the intuitive logic that discount chains provide interesting insights into the state of the economy and the consumer. If they see an increase in consumer spending it is a sign that the consumer is starting to feel the pinch and trading down (which is a clear warning sign for the overall economy). If things are flat or declining the consumer remains strong and the market still has momentum. I am guessing that things are generally flat here. If the Middle East conflict continues and we see sustained job loss then next quarter will tell a completely different story.

The Last Word

All of these macro level signals are clear warning signs right now. Things do not look positive. I am starting to stockpile cash in hopes of deploying it to purchase solid companies (specifically those in areas that are not directly tied to the price of oil). It is going to be a bumpy ride for a while.

Geopolitics Eclipses AI

Sun, 01 Mar 2026 19:48:00 GMT

The only news that actually matters this week is the outbreak of war with Iran. It is a developing situation with a trajectory that is currently anyone's guess. While I hope for a swift and peaceful resolution, history suggests we should prepare for a protracted conflict. I'm going to avoid any geopolitical commentary on the situation and keep my focus here strictly on the investing side of the equation.

War with Iran

With the outbreak of hostilities, it appears that the Strait of Hormuz has been effectively shut down. This means that a fifth of the world's total oil consumption (and nearly a third of all seaborne crude oil trade) is no longer moving. OPEC+ has signaled only a slight increase of output of 206k barrels per day, but it is not enough to make up for the supply shock.

For perspective, it's estimated that roughly 20 million barrels of crude oil and refined products flow through the Strait every single day. If the conflict does last for a long time, we are likely to see oil prices jump significantly, with some analysts already forecasting surges well over $100 per barrel.

To understand where markets might go from here, it is helpful to look back at the historical precedents set by the 1990-1991 Persian Gulf War and the 2003 Iraq War.

During the 1990-1991 conflict, the world experienced a massive drop in gross oil supply. Oil prices spiked leading up to the war but actually dropped to an average of $22 per barrel during the military conflict itself. That price stabilization was largely saved by Saudi Arabia rapidly ramping up production and the U.S. tapping into its Strategic Petroleum Reserve. The critical difference today, however, is that there appears to be far less excess oil production capacity available globally than there was in 1990. This means a similar production rescue by neighboring nations will be much harder to pull off.

We will likely see the trends found under the war puzzle phenomenon on the equities side. Equity markets were extremely jittery and weighed down by anticipation prior to the 2003 Iraq War. A large buildup of military assets in the area had been going on in the background as well. The stock market rebounded sharply the moment military action actually commenced back in 2003. The Dow Jones Industrial Average surged 8.4% following the invasion, because the start of the war resolved the uncertainty and investors bet on a quick victory.

The current situation with Iran is playing out quite differently. Uncertainty is increasing daily at the moment and clarity on the overall objective or trajectory is completely absent. Wars that break suddenly tend to cause stock prices to decrease as investors scramble to assess the newly introduced unforeseen risks according to the War Puzzle research. It is entirely unclear where things will go from here (especially with the de facto blockade of the Strait of Hormuz threatening to suffocate the global commodity chain). I do not expect we will see a relief rally. We will probably see a swing into the typical defense sector stocks. I am expecting a general overall selloff at least for the next few days.

Blue Owl & Private Credit

News broke earlier this week that Blue Owl was limiting withdrawals from one of their private credit funds. This sparked fears across Wall Street. Even Treasury Secretary Scott Bessent said he was concerned about the development. It feels like a call back to 2007 when a hidden and murky aspect of the banking industry starts to show cracks. I do not think we are at a point yet where this is a red alert. Similar red flags have flown before (namely when Silicon Valley Bank imploded in 2023) and we made it through that relatively fine. I will monitor for contagion before getting really worried. This is definitely a sector to watch closely.

Rapid Fire Observations

Some quick top-level thoughts on other big things that happened last week:

Block's Layoffs: I did a more in depth write up on this in Friday's Signal Brief. I think we are officially in an era where companies brag about AI taking jobs. The market rewarded Block ($SQ) by jumping 16% on the news. I expect we will see several other companies following suit. The divergence between Wall Street and Main Street widens.
AMD & Meta Make a Deal: The circular AI dealings will continue until morale improves (or something like that). Meta () and AMD () announced they struck a deal in which Meta will buy 6 gigawatts of compute from and take a 10% stake in AMD. All of the hyperscaler balance sheets will soon contain some significant stake in the chip manufacturers. They are building a knot that will be exceptionally gnarly to untie.
Nvidia earnings: Nvidia reported substantial revenue growth of 73% YoY but saw their stock drop 7%. This continues the trend of a jittery market punishing anything less than perfection. It is about time some air is let out of the balloon. The reasoning for why it is happening now is dubious though.

What's Happening Next Week

The major catalyst is the conflict with Iran and how things develop there. That will ultimately drive how the market reacts. Expect some selling if it looks like we are barreling toward a long and drawn out conflict. Any signs of optimism will move the market upward. I do not expect we will see any of that until the middle to the end of the week (if we see it at all).

CrowdStrike () reports earnings on Tuesday. This report would have been interesting to see if the SaaS-pocalypse feelings continued or if things trended opposite the current trend. It remains an important bellwether earnings report for the sector. It will just be nearly impossible to glean a meaningful signal with everything else going on.

The Last Word

Geopolitics has officially shoved AI off the front page. We have spent the last year obsessing over compute capacity and software margins. Now we are back to counting barrels of oil and assessing military blockades. The market hates uncertainty above all else and right now uncertainty is the only thing in abundant supply.

The Tariff Fallout, Meta's Court Case, and Claude's Continued Rampage

Sun, 22 Feb 2026 13:00:00 GMT

You know it’s a big week when every news outlet, podcast, and blog drops an emergency or off-cycle update. That’s exactly what we saw this week with the Supreme Court’s tariff decision and the administration's immediate reaction. This wasn’t exactly a surprise, prediction markets and most prognosticators had been pointing in this direction for a while, but now that it’s here we get to navigate the fallout.

Supreme Court Tariff Decision

For anyone who’s been living under a rock, the reason this decision is such big news is that tariffs have been the primary leverage (cudgel?) the administration has been using in what seems like every negotiation with foreign countries. The original legal basis the admin used to impose those tariffs was the International Emergency Economic Powers Act (IEEPA). The admin argued that the country was in a crisis and that under IEEPA, the president had the authority to impose tariffs. The court’s decision ruled that the president actually doesn’t have that authority and that the tariffs were effectively illegal.

Now, where does that lead? Well, first, all those tariffs that got paid potentially need to be refunded to the corporations that paid them (and for those corporations that already passed the cost on to consumers, that could mean a double payday). Unfortunately, the Supreme Court failed to weigh in on this specific topic for…reasons? So we’re in for a bit of chaos on this front as lawsuits get filed and the issue winds its way through the courts. This doesn’t exactly seem like the most efficient way to handle things, but I guess that’s just America at the moment. Anyone looking to make a bet on how this goes is taking a massive gamble. It will simply take too long to get a definitive answer, and there are too many opportunities for things to go in completely unexpected directions to make the risk worth it.

Second, and probably most salient, is what the admin is going to do next. They have already said they will impose a blanket 15% tariff under a different statute (one that limits the time these tariffs can be in place to 150 days). Long term, they seem poised to continue pulling the tariff lever in any way, shape, or form they can. It’s already being tossed around, and I hate to admit that I agree, but the infamous quote attributed to Andrew Jackson seems to echo loudly today:

John Marshall Has Made His Decision, Now Let Him Enforce It.

Andrew Jackson

The markets jumped on the news, with the major indices ending up on the day, although the move was rather tepid given how the market reacted when the tariffs were initially imposed. I, and many others, believe this outcome was largely priced in. What’s most likely holding things down is that no one really believes this is the end of the line for tariff. The future is just going to get messier and murkier. Keep an eye on the VIX this week. I expect it’s going to jump.

Section 230 and Meta Court Case

A hugely consequential court case for the future of the internet and big social media platforms is underway in LA, and this week we saw Mark Zuckerberg take the stand to defend Meta. The (extremely boiled down and oversimplified) crux of the case is that Meta made deliberate design choices that negatively impacted children’s mental health and wellbeing in order to boost engagement and retention. In the past, Section 230, the law that says platforms are not liable for the content users post as long as they take action when notified about illegal content, has been a shield protecting them from all sorts of liability. However, a while back, Clarence Thomas noted that design choices could be the avenue by which plaintiffs go after the platforms, and this case is the first true bellwether argument in line with that reasoning.

If the jury finds in favor of the plaintiffs, it will open up all platforms to intense scrutiny over their design decisions, potentially reshaping how they operate. Fundamentally, this would rewrite the entire premise of these businesses, and platforms like Reddit, Meta, X, YouTube, and TikTok would be forced to change. This, coupled with efforts across the globe to limit or block youth access to social media entirely, will almost certainly lead to top-line growth compression. It will be much harder for pure-play social media companies and stocks to justify their current multiples, and we’ll most likely see a sharp sell-off if this comes to pass.

Claude Continues its SaaS Rampage

Another week, another Claude feature causing a market panic sell-off. This week, the cybersecurity industry was the latest to feel Anthropic’s wrath as they announced the ability for Claude to scan codebases for potential security vulnerabilities. Similar to the other sell-offs we’ve seen, this is overblown, especially in light of what this feature actually does. Crowdstrike and Cloudflare both fell nearly 8% on the news, which is ridiculous since they provide bedrock infrastructure and tools, while the Anthropic feature is just a codebase review agent. The correlation between the two is minimal at best, but to non-technical folks, I guess they seem related? I expect this sell-off will be short-lived and these companies will claw back their losses once this wave of panic stops. The real questions are: how many more Claude feature releases are coming, and what’s the next sector that will feel the pinch?

Rapid Fire Observations

Some quick top-level thoughts on other big things that happened last week:

Figma Earnings: Figma released their Q4 and full-year reports. As I mentioned in my deep dive, I was looking here for signals on how their AI implementation plans were going and if they were continuing to see growth in NDR. I’ll be doing a more in-depth update, but the answers to both were very positive. They also announced their Claude Code integration, which sent the stock up 16% on the week.
Klarna Earnings: I mentioned last week I was looking to Klarna earnings to get a signal on consumer strength. They reported Gross Merchandise Volume (GMV) surged 32% year-over-year to hit $38.7 billion, pushing their quarterly revenue past the billion-dollar mark for the first time at $1.082 billion (up 38% YoY). What’s really interesting is that their provision for credit losses actually dropped to 0.65% of GMV, and the company noted that 30-plus days past due delinquency rates are showing a consistently improving trend. It seems that consumers are still making their debt payments, which is a comforting macro signal.
Robinhood Letting Retail Buy Pre-IPO: Robinhood announced this week that they are launching a $1B fund to let retail investors buy pre-IPO shares. I won’t be getting into this one myself, but I’m interested in seeing how this fund fares and what the overall return profile will look like.
Ads in AI: This is becoming the hottest battleground topic, seemingly supplanting benchmark scores, as Perplexity drops their advertising plans. I've firmly held the belief that Ads in AI is a bad idea. It will completely erode trust in an environment where user trust is paramount to continued usage. Burn the user once with bad info, and they're less likely to come back while the drumbeat of negative news continues. I don't like the UX of ads in AI, and I don't like the slippery slope it will inevitably open in the generation of answers. It's a bad idea all around. There's a reason OpenAI viewed it originally as a last resort, and now that they're going full bore on it, all the other players are virtue signaling that they're not going in that direction. I say "virtue signaling" because I expect they all will eventually incorporate ads; they're just waiting to see how OpenAI does it so they can learn from those mistakes before diving in themselves.

What’s Happening Next Week

Volatility. A lot of big things happening all at once.

Nvidia earnings: Nvidia , the largest company in the world by market cap, reports on Wednesday. I expect we’ll continue to see jaw-dropping revenue growth as the AI players continue to increase CAPEX. Let’s see how Jensen guides, but I bet it will be rosy no matter what. Will they continue to meet Wall Street’s lofty expectations? Things are more jittery and on a knife’s edge than in the past, so if they miss, things could get ugly fast.
Workday & Salesforce Earnings: Workday has had a rough 2026 so far, down 33% YTD. Salesforce hasn’t fared much better at -27%. Both report earnings this week with their entire business model (and as I’ve argued in the past, probably their entire business identity) under threat from AI. I’m looking to see how they guide and what they’re seeing in terms of retention. Any sign of weakness here, and they’re going to get clobbered.
Tariff Updates: I expect we’ll get a deluge of news and updates on the tariff situation, although I don’t expect the picture will get much clearer in the next few days. I’m looking to see how the administration reacts and if it follows through on the blanket tariffs. The market open on Monday will be telling.

The Last Word

If there's a throughline this week, it's that the old playbooks are being thrown out the window. Whether it's the Supreme Court upending the administration's primary economic weapon, the LA courts opening the door to fundamentally alter how social media operates, or AI clobbering SaaS, structural change is happening fast. We are staring down the barrel of a highly volatile week. Keep your head on a swivel, don't get shaken out by headline panic, and as always, watch what companies actually do, not just what the market thinks they'll do.

AI Fears Bleed into Finance

Tue, 17 Feb 2026 18:00:00 GMT

A little late this week on posting. Trying to get into the swing of things and it being a holiday weekend didn’t help with building the habit. Last week was a continuation of jittery trading, with the AI fears breaking containment a bit and bleeding into the financial sector. A lot of big macro-focused news came out as well at the end of the week. Let’s dive in.

The SaaS-pocalypse Infects Finance

I wrote about this last week, but the narrative that AI is going to fundamentally change our relationship with software has been localized mostly to pure-play tech companies. It appears that last week was the start of that narrative breaking containment a bit and it hit a sector that I wasn’t expecting, financial services. It looks like investors are starting to question business models that have exposure to the powers that AI companies are starting to build out in their models.

Let's look at the damage:

Charles Schwab dropped ~12%.
Morgan Stanley shed ~5%.
Raymond James finished Tuesday down ~8.7%—its worst drop since the March 2020 pandemic panic.

This sell off was allegedly triggered when Altruist announced an AI tool that creates personalized tax strategies by interpreting financial documents without manual entry. Suddenly, the market is panicking that we are on the verge of "vibe financial planning." I’m skeptical. When it comes to retirement, people are inherently risk-averse. Handing the keys to your life savings to an AI agent seems a bridge too far for most. However, commoditized tax services like H&R Block and Intuit are a different beast. They are walking into an existential buzzsaw.

The tax code is effectively just a massive, convoluted rules engine, exactly the kind of environment where AI thrives. We are rapidly approaching a reality where you can simply upload your W-2s to a secure model and ask, "How do I fill this out?"

At that point, the value proposition collapses.

Why would anyone continue paying hundreds of dollars to TurboTax for a "guided experience" when the reasoning cost to process that data trends toward zero? These companies have built their empires on the friction of complexity. But if AI solves the complexity, the justification for their fees evaporates.

Maybe then, if their lobbyist war chests finally dry up, we can get the tax reform that simplifies this whole process and the entire sector can finally go away.

The Surveillance Creep: Privacy vs. Name Tag

We are witnessing a massive tension between the commercial success of consumer hardware and the steady erosion of public anonymity. Recent news from Meta and Ring suggests a strategic pivot in the Surveillance-as-a-Service economy. Companies are betting that convenience will eventually provide enough cover for features that were unthinkable a decade ago.

Meta’s smart glasses are an unexpected commercial juggernaut. The company sold over seven million pairs last year and now they are reported to be deploying their facial recognition software internally called Name Tag. This would allow wearers to identify strangers and pull up their information via an AI assistant. Zuckerberg has allegedly even internally debated whether to keep the recording light on (or if the glasses should use a different signal) when the super sensing feature is keeping a record of a person’s entire day.

Sounds like a dystopian nightmare to me. Thankfully, we’re seeing some resistance here. Ring faced a massive uproar following a Super Bowl ad for a Search Party feature that uses AI to find lost dogs. The public was quick to recognize that if you can find a dog you can use the tech to find a person. That backlash seems to have resulted in Ring killing its deal with Flock Safety which sells license plate reader tech to police. I really don’t want to live in even more of a Black Mirror episode than we already do so I hope that the public and the market continue to push back on these types of features and product developments.

Rapid Fire Observations

Some quick top level thoughts on other big things that happened last week:

The Jobs Revision Shell Game

The January jobs report showed a gain of 130,000 jobs, which looked like a win on the surface. However, the benchmark revisions for 2025 were a disaster. They wiped out 403,000 jobs that we previously thought existed, leaving 2025 with a pathetic total of only 181,000 jobs added. This creates a massive trust gap for investors. When data is this susceptible to being erased after the fact, the only pragmatic move is to assume current numbers are a fantasy.

Unity’s Crisis

Unity announced their earnings last week and took an absolute beating, dropping another 34% bringing their YTD returns to ~-58%. As I mentioned last week, this was their change to assuage concerns about AI’s impact on their business and they failed to do so.

I do however LOVE this quote from CEO Matt Bromberg regarding AI:

The impact [of AI] is gonna be much more about..the time to innovation than the time to market.

I wholeheartedly agree with that sentiment. We’ll see if Unity can help be the bridge to that future.

Discord’s IPO Cleanup

Discord is in the classic cleaning the house phase before its planned IPO. They are rolling out age verification and global teen-by-default settings to satisfy regulators. This is the standard social media lifecycle: skirt the rules to get big, then lock the doors once you need institutional buyers. While users are reportedly surging toward TeamSpeak in protest, I expect that will be shortlived. Users are already figuring out how to get around the checks and the networking effects Discord has are real.

What’s Happening Next Week

We have a few more interesting earnings report coming this week ahead of the big Nvidia earnings next week.

Figma Reporting

As I mentioned in my deep dive I am looking closely at Figma's AI guidance and spend. The lock-up share release schedule hits on Feb 18th (putting roughly 44.4 million shares on the market). If management cannot prove their AI spend is driving real margin expansion, that supply glut will keep a lid on any price appreciation.

Klarna and the Debt Wall

Klarna's upcoming report is the canary in the coal mine for the K-shaped economy. Consumer debt is rising. I want to see if the buy-now-pay-later model is finally hitting a wall. If the consumer is tapped out, Klarna will be the first to show the cracks in the household balance sheet.

GDP Expectations

There is a widening gap between the model and the market. Prediction markets are currently pegging Q4 GDP at 3.3 percent. However, the Atlanta Fed’s GDPNow model is forecasting 3.7 percent. If we print closer to what the prediction markets think vs. the Fed’s own predictions, I’m curious to see how the markets react. Are the big players starting to trust prediction market data more than the governments? It will be interesting to see how things play out here.

The Last Word

We’re going to continue to be on a rollercoaster this week. Sentiment isn’t shifting yet and we have a few more big earnings reports to get through and digest.

We’ll see if the AI fear continues to spread even further. I’m sure there will be another big AI advancement or new tool announcement that will give everyone an opportunity to overreact about.

The SaaS-pocalypse & AI Cognitive Dissonance

Sun, 08 Feb 2026 22:17:00 GMT

Welcome to the debut of my weekly market thoughts posts. My plan is to use this as an open-air brain dump where I attempt to connect the dots between macro trends, earnings, and what’s going on in the world of tech. My goal for these posts, at least initially, is to stress-test my thoughts and predictions to keep myself honest and try to spot the opportunities before they hit the headlines.

If that ends up being interesting and helpful to others, then I’ll consider that my own personal alpha.

The Great Tech Divergence

This week was a whirlwind, to put it mildly. The technology sector is currently trapped in a fascinating paradox.

On one hand, investors are starting to get squeamish towards the hyperscalers for their astronomical CAPEX spend on AI. In a vacuum, it’s scary to see how much money they’re spending for the datacenters, chips, and other infrastructure needed to build/support AI. Just this past week, Google

raised its 2026 CapEx guidance to $185B and Amazon raised theirs to $200B, evoking ghosts of the early 2000s fiber bubble.

On the other hand, the market is battering traditional software companies under the fear that their business models face imminent obsolescence because of AI. It forces the question, how can the SaaS sector shed $300 billion in value because AI is too capable, while the market simultaneously signals that the hyperscaler's investments building that capability won’t generate sufficient ROI?

In my opinion, we are likely are overbuilding AI capacity in the short term. But I believe this capacity will find a purpose much faster than those dark fiber lines did two decades ago. The adoption curve for AI compared to the internet's adoption curve is really eye opening. It took the Internet >30 years to reach 68% of the population while the Fed reports that AI adoption is already at 54%.

In my opinion, the market is reacting irrationally on the build-out.

The reaction towards software however… a lot of that SaaS value isn't coming back.

The SaaS-pocalypse

Last week Anthropic released a new legal plugin for its Claude Cowork agent which is supposedly the catalyst that set off this week’s SaaS-pocalypse.

The damage was widespread. Workday dropped nearly 8% and Salesforce

fell 10%. ServiceNow shed nearly 15% and Atlassian plummeted almost 20%. Google’s release of Genie 3 added fuel to the fire. Video game stocks saw sharp declines with Unity dropping 17% and Take-Two falling 13%.

As I noted in my deep dive on Figma, margin compression in SaaS will continue. This meltdown is not unexpected. Investors are waking up to the fact that AI can do what it’s been hyped to do, at least when it comes to coding. I don’t necessarily believe ALL the AI hype, but I do believe in its coding abilities. I still hold firm that many big SaaS companies are in trouble unless they pivot. They must switch from selling software to selling AI agents that replace that software. This includes enabling customers to build the very platforms they currently sell.

On the other hand the reaction to the Genie 3 drop is overblown. Genie 3 is a cool proof of concept. But I don’t think it will completely change how video games are built and enjoyed. It will help designers imagine and prototype different worlds. If anything I hope it can herald an explosion of new game concepts.

Was this week’s drop overblown? Probably. Markets rarely move in straight lines. We will likely see a dead-cat bounce in the short term. But I expect more red in the future. We are watching a fundamental repricing of software utility. Companies that can't transition from tools to agents are going to be left behind.

Rapid Fire Observations

Some quick top level thoughts on other big things that happened last week:

The Triple-X Merger

Elon Musk is combining SpaceX, xAI, and X into a $1.25 trillion monster with an IPO expected later this year. I love the sci-fi ambition of space-based datacenters, but I’m skeptical about saddling a what has been a successful rocket company with the toxic volatility of social media and a capital incinerator in the thick of the AI race. As always with Elon, it will be interesting to watch.

The Crypto Crater

Bitcoin is hovering at $70,000, which is ugly when you consider it costs ~$90,000 to mine a single coin right now. Lots of interest centered on Michael Saylor’s Strategy, which announced an operating loss for the fourth quarter of 2025 at $17.4B as a result of their digital asset strategy. Yikes!

OpenAI is Chasing, Not Leading

Did you catch that OpenAI launched GPT-5.3-Codex this week? If you didn't, you aren't alone. The launch was incredibly muted, especially compared to the shockwaves Anthropic sent through the market. It feels like OpenAI is starting to take the back seat, especially when you couple it with the ad news storyline that came up this week. Anthropic is gaining steam.

Oracle and Debt

I recently read (and really enjoyed) Andrew Ross Sorkin’s book 1929. One of the things he noted in his piece in The Atlantic was that “Debt is the almost singular through line behind every major financial crisis.” That context makes Oracle’s kicking-off of a massive bond sale, part of a plan to raise up to $50 billion this year to buy more GPUs and build more data centers almost entirely for OpenAI, pretty worrisome. Given what I mentioned above, and the PR debacle of Oracle's own post this week, I don’t see this playing out well.

Prediction Markets vs. The House

DraftKings stock is down ~40% over the past 6 months, primarily driven by the rise in prediction markets. Kalshi and Polymarket are eating into the traditional sports books. I’m very skeptical that prediction markets will be allowed to operate this freely forever.

My prediction? The thin veneer of prediction markets acting as information aggregators and that being a basis of their CFTC approval will wear away eventually. Regulators are already circling reports of insider betting. That will force a regulatory hammer drop, though likely not until there’s an administrative change in Washington.

What’s Happening Next Week

Next week is shaping up to be a macro-heavy gauntlet that will test whether this correction has legs.

The Delayed Data Dump: Usually, we get a breather between jobs and inflation data, but not this time. We get the Employment Situation report on Wednesday, Feb 11, followed almost immediately by the CPI print on Friday, Feb 13. I’m expecting that we’re going to see the unemployment rate continue its upward trend, so if those inflation numbers come in hot while hiring remains cold…Hello Stagflation, is that you?
DraftKings: They report this week and I’m curious to see how things are going with their recent release of DraftKings Predictions, but I don’t expect it’s going to be very positive. Management needs to explain how they plan to stop the bleeding here.
Unity: This earnings call is their chance to recoup some of that beat-down they took this week. Can they assuage investor’s concerns and prove that generative AI will actually be a boon to their business?

The Verdict

I think what we saw this week with the big tech companies is a temporary correction driven by panic over CAPEX. We are talking about the richest companies in history with massive cash reserves and reasonable P/E ratios. Fundamental mechanics are being tested, but the actual revenue from AI cloud services is real.

I’m keeping my eye on Amazon this week. It’s looking like a good entry point and I expect they’ll bounce back a bit this week as fears subside.

Figma in the AI Era

Mon, 02 Feb 2026 00:00:00 GMT

Before we dive deep into Figma, let's get into the tech industry as a whole, specifically SaaS businesses. As everyone who has used the internet in the past three years knows, AI is the only thing the industry cares about right now. Predictions on where this leads have been varied, to say the least. However, one narrative I believe is real is the impact AI will have on the B2B SaaS business model.

The problem for these products is that AI makes it exponentially easier to build software. If a company sees its costs for SaaS tools rise, there will be a tipping point where it becomes cheaper (and I’d argue eventually better, due to customization) to build an in-house solution using tools like Claude Code, Cursor, and the Gemini CLI, or to utilize open source software. I’ve seen this firsthand while building this very website. Things I could never have built on my own in the past now take a weekend and a few well-phrased prompts. Because AI is inherently rules-based, it is only going to get better at coding.

So, what does this mean for Figma? Is it screwed because it’s a SaaS business?

Well, yes and no, but more no than yes. Its current business model, charging per seat/license, will likely not be tenable in the long term as AI agents start doing the heavy lifting. In the future, Figma will be constrained on pricing power as that "build vs. buy" tipping point continually shifts downward. As AI costs and agentic coding abilities follow their own version of Moore's Law, the rent for software becomes harder to justify.

However, a big advantage for Figma is the Jevons paradox. For the non-econ nerds, Jevons paradox occurs when a resource becomes more efficient to use, leading to more total consumption of that resource, not less. We'll likely see significantly more people doing design work in the future. As the time needed to actually write code shrinks, the premium shifts entirely to the creative work and the architecture. If you can vibe code an entire app in a weekend, you're going to spend a lot more time in Figma making sure that vibe is actually good.

Another buoy to Figma is that its users seem to genuinely love the product. Love for a productivity tool is rare, but when it happens, it builds great businesses. Think back to the early days of Salesforce or Excel. These weren't just tools, they became the bedrock of their respective industries because they solved a fundamental problem so elegantly that the users became the marketing department.

Compare that to Adobe , where there's less love and more lock-in and inertia. I believe Adobe initially grew because their products were loved, but enshittification seems to be the law of entropy for most software. Today, people use Creative Cloud because they have to, not always because they want to. Figma is still in that magical phase where growth is driven by genuine utility and a superior workflow rather than hostage-taking. This gives them a massive head start in the AI transition where users most likely prefer Figma to win because they actually like using it.

The Business

Let’s take a look at the actual business. Today, Figma is almost purely a seat-based subscription play. They have a free tier to hook the hobbyists, but the real money comes from the Professional, Organization, and Enterprise tiers. Like most SaaS businesses prior to the AI mania, it was a high-margin, predictable cash machine growing like wildfire. Now that we’re in the middle of the AI transformation however, things are looking a bit…different. Let’s dive into some numbers.

Revenue and Growth: Figma is putting up solid revenue numbers, having just crossed the $1 billion ARR mark per their latest Q3 2025 filing. Their annual growth rate from 2024 to 2025 was roughly 40%. This is a deceleration from their historical hyper-growth standards, but not alarmingly so given the law of large numbers. Their P/S ratio, based on these revenues and today's market price of ~$26, sits at approximately 12.8x (we'll touch on this more later).
Net Dollar Retention (NDR): One very interesting metric is their NDR for large customers (>$10k spend), which sits at 131%. This confirms that once a company adopts Figma, they deepen their usage over time. However, this metric will be the one to watch as they hit the "Saturation Wall." Figma has effectively conquered the pro designer market. To maintain this retention and expansion, they must successfully cross the chasm to non-design roles (e.g., product managers, developers).
New Product Offerings: To fuel that next leg of growth, Figma is aggressively expanding its portfolio. In 2025 alone, they launched four new products: Figma Sites (publishing), Figma Make (AI prototyping), Figma Buzz (marketing assets), and Figma Draw (vector editing). These standalone bets are designed to widen the funnel beyond the core UX designer. Success here depends on their ability to continue making products that their users love (which I think they can do).
Operating Margins: Historically, Figma operated with 90%+ gross margins, fairly standard for SaaS. However, in the first nine months of 2025, that number compressed to the 83-86% range. We’re seeing the real impact of AI here. Running generative models for tools like Figma Make requires heavy compute and inference spend, fundamentally altering the unit economics of the business compared to the pre-AI era.
The Dilution Dilemma: The most immediate pressure on the stock price is supply. The IPO lock-up period expired on January 27, 2026, flooding the market with roughly 37 million shares. But we aren't done yet. Per their filings, another 20% of lock-up shares will be released after the 2025 annual earnings report, followed by another 27.5% after Q1 2026 earnings. This scheduled supply glut will likely keep a lid on price appreciation in the short to medium term (more on this later).

The Opportunity

Now that we’ve established the baselines, let’s talk about where I think Figma is going.

As mentioned earlier, Figma currently trades at a P/S ratio of roughly ~13x. To put that in perspective, let’s look at other big-name SaaS companies in their moderate-to-high growth phase. Historically, Datadog has traded in the ~14-16x range, Snowflake

hovered around ~14-18x, and Cloudflare , typically a premium asset, often commands multiples north of ~20x.

At ~13x, Figma is entering territory that looks enticing for a company growing at 40% with best-in-class retention. However, it’s not trading at a valuation discount by accident. Everyone in the SaaS space is seeing their margins compress as they scramble to integrate AI. Either they are spending more on third-party AI services or shelling out for the raw infrastructure to build their own. Figma falls squarely in the latter bucket, and that capital intensity is acting as a drag on its multiple; investors fear that their recent margins will be their forever margins.

Beyond the balance sheet, there is the existential fear that AI-native tools like Flora, V0, or even ChatGPT will make the platform moot. If you can simply prompt a design into existence, you don't need a Figma seat.

But I think the opposite is true. To understand why, let’s look at another company that went through a similar fear cycle, Google .

Remember when ChatGPT launched, and the narrative shifted overnight to "Google is dead"? The logic was that LLMs would replace search engines, and Google’s ad business would crumble. But then the narrative shifted again. As Google rolled out its own AI offerings while growing its core Ads business, the market realized that Google was actually in a prime position. Why? Data and distribution. Google sits on top of what is probably the richest trove of data needed to build super-intelligent AI, from across its product suite. It is THE key differentiator between it and the other big AI players and why so many people now think Google is on top in the AI race. Coupled with their distribution reach given their billions of users, Google has seen its stock jump from ~$88/share when ChatGPT was announced to over $338/share today (~284% growth).

Well, you know what Figma has? Data and distribution.

We’ve already established Figma is the #1 tool for designers globally. This gives them a huge distribution moat. Just like Google can instantly put AI in front of billions, Figma can deploy AI agents directly into the workflows of millions of professional designers without requiring them to switch tools.

They also sit on a goldmine of proprietary design data that no other company can access. Here is the key nuance: it’s not just about the final pixels. Any crawler can scrape a live website and see the final result. Figma can see the whole process that led to the creation of those designs.

Figma sees the 50 rejected iterations. They see the comments between the developer and the designer explaining why a button should be blue and not red. They see the diffs and the zany ideas that didn’t make it to market. This is process data, and it is exponentially more valuable for training an AI agent that is actually useful. If you want to train a model to think like a designer, you need to show it how a designer works, not just what they finished.

If Figma can execute on training models using this process data, they won't just stay relevant; they can become the operating system for the entire creative process. That should entrench their dominance, but the market is currently acting like they’re a legacy tool. I suspect we won't see AI replacing Figma, but we'll see people using Figma as the default AI engine for design.

The Threats

Of course, Figma isn’t the only player making moves in this space. It’s worth looking at what the competitive landscape looks like.

Flora: They just announced a $42M Series A, pitching themselves as a sort of AI-Native Figma. The argument is simple: Figma was built for vector manipulation while Flora was built for generative orchestration. Investors are betting that you can't retrofit an old tool to be AI-first and that you have to rebuild it from the ground up. This is the Innovator’s Dilemma in action. Can Figma navigate the new paradigm without breaking the workflow that millions of users rely on? If not, Flora is positioned to step in.
Adobe: They’re the 800lb gorilla in the room. Their Firefly models are integrated into everything from Photoshop to Premiere. If they decide to bundle a good-enough UI tool with the Creative Cloud subscription, Figma’s pricing power gets tested. However, Adobe’s attention is stretched across a wide array of products, and I don’t have faith that they’ll be able to build a product users genuinely love.
V0 & Cursor: While not direct competitors, tools like Vercel's V0 and Cursor allow developers to prompt UI directly into code. If the handoff between UX and Dev disappears because the code is generated instantly, does the design phase shrink? Figma's seat count depends on design being a distinct, heavy phase of the lifecycle. If we move to a world where developers just vibe code the frontend without a mock-up, Figma could find itself bypassed entirely for simple applications.

The Interesting

Before we wrap up, there are two distinct factors that make Figma a fascinating stock to watch.

First, Adobe's $20 Billion offer. We can’t forget that Adobe was willing to pay $20 billion for this company back in 2022. Yes, regulators killed the deal, but it served as a moment of price discovery. It proved the strategic value of Figma’s platform. If the stock price drifts too low, Figma becomes an acquisition target again—perhaps not for Adobe, but for other AI giants. If you are a model builder wanting to own the "creative process layer," Figma becomes an interesting data acquisition play.

Second, the "Extended Lock-Up" Calendar. As mentioned earlier, the short-term supply dynamics are brutal. Approximately 54% of Figma’s Class A common stock is held by Extended Lock-Up Holders, subject to a tiered release schedule that will keep pressure on the stock through 2026.

Here are the key dates:

Post-2025 Annual Earnings (Feb 18th): ~44.4 million shares (20% of the locked-up total) will become eligible for sale on the second trading day after Figma announces earnings for the year ending December 31, 2025.
Post-Q1 2026 Earnings (some time in May probably): Another ~61.1 million shares (27.5%) will hit the market on the second trading day after earnings are announced for the quarter ending March 31, 2026.
Post-Q2 2026 Earnings (or Aug 31): The Remainder (~77.7 million shares) will be released on the earlier of (i) the second trading day after earnings for the quarter ending June 30, 2026, or (ii) August 31, 2026.

As millions of shares become eligible for sale for the first time, we are likely to see sustained downward pressure.

Final Verdict

So, where do we land?

Fundamentally, I believe Figma is a rock-solid product. They have that rare combination of utility and genuine user love. I believe in their ability to build products people actually want to use, and I think they are uniquely positioned to thrive in the AI era by leveraging their twin engines of process data and massive distribution.

From a valuation standpoint, we are getting close. A P/S ratio of ~13x for a company growing at ~40% is starting to flash "Buy." The fundamentals are green, the product is green, and the growth is green.

But I’m still holding at Yellow for now.

There is too much downward pressure on the horizon. The entire SaaS sector is likely to see further multiple compression as weaker companies feel the AI pinch, dragging the sector down with them. More importantly, the supply glut isn't over. We just absorbed the IPO lock-up expiry, but the next tranche of stock releases is right around the corner.

I’m staying on the sidelines for another quarter. I want to see how the market absorbs that supply and, more critically, what guidance management gives in the Q4 2025 earnings report. If they can show that AI monetization is real and not just a margin-eater, and if the stock price survives the next dilution event, that yellow light might turn green very quickly.