👉👉👉Reminder to sign up for the Weekly Recap only, if daily emails is too much. Find me on twitter, for more fun. 

Subscribe now

👋👋👋Random Walk has been piloting some other initiatives and now would like to hear from broader universe of you:

(1) 🛎️ Schedule a time to chat with me. I want to know what would be valuable to you. 

(2) 💡 Find out more about Random Walk Idea Dinners. High-Signal Serendipity.

5 Idea Friday

Some quick-ish hitters for your weekend reading pleasure.

1. Durable Goods, Ain’t What They Used to Be

One big, but still under-appreciated change to the economy is that it’s much less dependent on big durable goods, like houses, than it was before.

It used to be that a downturn in residential housing investment was a sure-fire predictor of recessions:

Residential investment historically drags in the lead-up to recessions, while non-residential investment actually tends to increase its relative GDP contribution.

The exception was Dotcom, a tech-driven cycle.

Well, now that everything is a tech-driven cycle, it should come as no surprise that this hasn’t been a bigger deal:

Housing’s share of GDP has been in steady decline, and has now retreated slightly below the pre-pandemic level.

An economy no-longer primed to people-growth, is also no-longer primed to homebuyer growth.

Back out the series farther, and you can see the straw that now stirs the stuff-making drink is of a different variety:

Private Capex as a % of GDP has climbed, while residential investment has really been flat-to-down for a while.

And, of course, as of this moment, it’s just any private capex, it’s one particular kind of capex that makes the world go round:

Tech-related Capex is the Captain Now.

That manufacturing boom-let? It’s definitely real (contra what some pundits would have you believe), but once again, AI is doing all the work:

Manufacturing production has surged, but it’s all AI & Data Centers, all the time.

Durable goods orders, blowing out wide?

Sure. But again, it’s really just one thing:

New orders for computers & electronic products went vertical.

But, unless a hot new iphone dropped that I’m not aware of, it can mean only one thing: AI is the Capex Cycle, Now (ad infinitum).

Let’s keep it up, and hope reinforcements arrive soon.

ICYMI

2. Price Check on Terminal Value (reprise)

Earlier today, Random Walk made the following assertion about stock prices:

Bigger picture, though, if you want to be skeptical about stock prices, that’s fine, but your skepticism isn’t about present-day fundamentals. Present-day fundamentals are all screaming the same story: this train is becoming a full-bore rocketship. There is no dark fiber, and there are no empty houses rotting in the sun belt.¹

Any skepticism has to be rooted in some claim about the future. The train is hurtling forward as quickly as market prices are saying it is—perhaps even more quickly. If you want to be bearish, you have to believe that something will stop (or at least slow) this train.

In other words, the current rate n’ pace of fundamentals support current valuations (and, arguably, a bit more). If you want to make an argument that prices are too high, you need to make an argument that the current rate n’ pace of fundamentals won’t last.

One can make that argument, and Random Walk has variously gestured at it in the past, but that’s not really my point. This is really just a somewhat clunky segue to a related, but different, point.

One way to think about the fraction of a company’s share price that is not priced-to-fundamentals is the perceived “terminal value.” If some part of the valuation is the net-present-value of, say, 10-years’ worth of future earnings, dividends and buybacks, then what’s left is the “terminal value” beyond that period.

Recently, Goldman took a stab at estimating what percent of the index is attributable to that “terminal” tail. It turns out, a lot:

Terminal value comprises ~75% of the S&P’s total price.

At 75%, the terminal share is nearing the Pandemania Peak, but is still well-below Dotcom.

Of course, part of why the terminal share is quite so high is that the index is rife with techcos, which tend to be high-growth companies, which tend to be the sort of companies that are assigned a higher terminal value:

Tech stands alone in the “high-growth” upper-right quadrant of rapid sales and high margins, and “high growth” stocks have an ~84% terminal value share.

While it’s always a bit hazardous to extrapolate future growth-rates from present ones, one of the great investing mistakes of the last ~20 years or so was to discount that extrapolation. Tech grew bigger and faster than any sector ever had before, and the more you doubted that possibility-cum-reality, the more you missed out on the upside.

Of course, that doesn’t mean that this time will be like that last time—it could be worse or it could be better—but it does mean, once again, that whether you are bullish or bearish about current asset-values depends on whether you think techcos can keep this up (and for how long).

Fundamentals support the high-premium to terminal value, but if growth were to slow a bit, the discount would be substantial:

A 1pp decline in long-term growth rates would correspond to a 29% decline in high-growth share prices, in Goldman’s Terminal Value model.

Considering that the market is rife with speculation about AI’s potential negative impact on the terminal value of high growth companies, then this might give you pause. Of course, considering that the market is also rife with speculation about AI’s potential positive impact on pretty much all companies, then you might conclude terminal value isn’t high enough.

Investing is hard.

3. Economies of Scale in Energy

All things being equal, you’d think that declining demand would correspond to declining prices, but when it comes to energy, that’s not always the case.

In fact, in many parts of the world, the opposite is true: more demand leads to lower prices:

Across Europe and the US, electricity prices have tended to increase with a collapse in demand (and vice versa). Ireland is a weird outlier that has more expensive energy and more demand. Too bad for Ireland, I guess.

The reason price and demand are inversely correlated is a function of grid economies of scale: more demand spreads the costs of fixed assets across a wider universe of consumption.

It’s part of why the “a solar panel in every pot” has failed as an energy policy. The grid was built for a few large sources of relatively invariable energy, which efficiently distribute distribution costs across a massive grid. Many, small and variable power sources do the opposite of that. The transmission costs alone often make the juice-not-worth-the-squeeze.

Speaking of transmission, don’t expect it anytime soon:

The bottleneck for new transmission lines has blown all the way out past 2030.

That too could put a damper on AI-related growth, I suppose, i.e. demand overwhelms supply. Of course, that would drive prices even higher and the balance is self-regulating, but it’s a thought.

4. Uncharted Waters, Previously Charted

Random Walk writes and thinks a lot about Entering Uncharted Waters, Blissfully Unaware.

The very short of it is that we’re losing a century-old demographic tailwind that’s so powerful, but also so constant, that we’ve forgotten it’s there, and we’re surely going to miss it when it’s gone.

It’s not entirely true that the waters are uncharted, though. They’re uncharted for us, but perhaps not for some other countries like China, and Japan.

At the risk of drawing spurious correlations from chart-peeping, please proceed to draw some correlations by peeping at the following charts:

• Japan’s working age population:

  

• Japan’s corporate profits as a share of GDP:

  

So, you’re saying that Japanese corporate profits started a relentless upward climb the moment that Japan’s working age population rolled-over?

Yes, glad you noticed, too.

And hang on, US corporate profits have more recently started a similar structural climb (and “labor’s share” of those profits a structural descent)?

That’s the gist of it.

I guess when we run out of people to hire . . . we just stop hiring, but keep finding ways to sell (maybe)? Idk. Could just be a coincidence, but it’s hard to unsee.

5. Hidden Healthcare Taxes

Last week, Random Walk wondered what non-fraud explanation could possibly explain the massive disparity in home health workers per senior in New York, but also CA, DC, and MN.

So far, I’m yet to come across anything that isn’t “well, it may not be deliberate fraud, so much as those states made it really easy to commit fraud, and pretended not to notice.”

Fine. Judge that however you see fit.

This week, I would suggest that perhaps there are some very tangible consequences of that fraud-not-fraud. And it’s actually something Random Walk has observed many times before: ‘Someone else pay for my healthcare’ is probably the worst idea ever.

Once again, the cost of healthcare is rising, and while people are paying for it, they probably have no idea because the bill is handled by their employer:

The employer-cost of healthcare is growing much more quickly than wages.

What that means is that healthcare costs are “crowding out” wage growth, and the impact is regressive, in that it doesn’t scale with incomes—the less money you make, the higher the “healthcare tax” is, as a percent of your income.

OK, but what does that have to do with home healthcare aids in New York?

I’m so glad you asked.

Well, part of why employer-based healthcare costs are rising is because of the terminally higher cost of services. Another part of why they’re “rising,” is that the Obamacare subsidies have expired, which is to say that costs aren’t rising, but now the bill is somewhat less-attenuated.

Yet another reason costs are rising, and this is the one that matters for today, is cross-subsidies.

You see, our “free” national healthcare system doesn’t fully reimburse providers for what they charge. Hospital systems et al eat some of the shortfall (which can be quite large), but the rest they foist on the private market insurers . . . who then pass those costs on to employees (unbeknownst to them) via higher employer-paid premiums, as per above.

In other words, when public health consumption increases, taxes go up to pay for it, but we don’t call them “taxes,” we call them “[some fraction of] employer sponsored health insurance premiums.”

That being the case, if, for example, you were a state with a high influx of illegal immigrants, and you wanted to create “jobs” and “growth,” by making it easy for them to get access to public healthcare . . . then you could do that, and at least some large portion of the bill would get redirected to “private” market employees. And not only would they likely not even notice—hooray for you, elected officials—those private citizens would probably blame the insurance companies for your largesse.

It’s the perfect crime.

Is that definitely what’s going on? Idk. It’s a theory though, and my guess is that it’s got legs.

Previously, on Random Walk

Random Walk is an idea company dedicated to the discovery of idea alpha. Find differentiated data, perspectives and people, and keep your information mix lively. A foolish consistency is the hobgoblin of small minds. Fight the Great Idea Stagnation. Join Random Walk. Follow me on twitter. Follow me on substack:

Get more from Moses Sternstein in the Substack app

Available for iOS and Android

Get the app

👉👉👉Reminder to sign up for the Weekly Recap only, if daily emails is too much. Find me on twitter, for more fun. 

Subscribe now

👋👋👋Random Walk has been piloting some other initiatives and now would like to hear from broader universe of you:

(1) 🛎️ Schedule a time to chat with me. I want to know what would be valuable to you. 

(2) 💡 Find out more about Random Walk Idea Dinners. High-Signal Serendipity.

Price Check on Terminal Value

Last week, Random Walk mused about the weirdness in public equity valuations.

Yes, the stock market is clipping all-time-highs, and yet, you could argue that it’s not high enough . . . profits are accelerating, while the price of those profits compress. It just doesn’t happen very often (and nor would one expect it to).

High-growth companies are still not being rewarded commensurate with their growth:

Price performance for even high-growth companies has substantially lagged forecast profit-growth since late last year.

A great wall of profits

But, it’s not just high-growth companies putting on a show.

I mean, there’s a great wall of profits, and it keeps on climbing:

Every single sector in the S&P is appears to be prepping for a Walter Payton summer training program.

Since then, the market melt-up has mostly continued, and even then, it’s still weird . . . because on the other hand, whatever you think of profit-growth, isn’t there some kind of massive oil shortage that’s going to crush demand and supply chains the world over?

There’s some evidence that demand-destruction has already started:

Air-Freight prices go up, Air-Freight volumes go down.

That, of course, could be the discount weighing on those impressive earnings—the bullishness of the AI trade has mostly outweighed the bearishness of Hormuz, but “mostly” is not the same as “entirely.”

Fair enough.

ICYMI

AI Keeps Cooking

What we do know, however, is that the AI trade is showing no signs of slowing down. Whatever oil, fertilizer, resin, helium, etc. shortage we’ve got coming, it’s got nothing on the apparent shortage of available compute.

Why do the hyperscalers keep pouring historic amounts of money into data centers and chips? Well, earlier this week, they gave us a clue:

Read more

👉👉👉Reminder to sign up for the Weekly Recap only, if daily emails is too much. Find me on twitter, for more fun. 

Subscribe now

👋👋👋Random Walk has been piloting some other initiatives and now would like to hear from broader universe of you:

(1) 🛎️ Schedule a time to chat with me. I want to know what would be valuable to you. 

(2) 💡 Find out more about Random Walk Idea Dinners. High-Signal Serendipity.

5 Idea Friday

Some quick-ish hitters for your weekend reading pleasure.

1. Uncharted Waters, Still Unawares

Just briefly revisiting Uncharted Waters, Blissfully Unawares.

First, a nice illustration of just how unusual it is to have such a yawning gap between GDP growth and employment growth:

GDP and employment growth tend to run together, except (1) now; and (2) in the lead-up to the Dotcom Bubble.

Interestingly, the Dotcom bubble was followed by a “jobless recovery,” until a rate-driven real estate cycle kicked-in, which helped drive employment up (for a bit). Draw your own conclusions.

Here too, is another look at current weirdness of “growth without growing:”

Private sector payrolls showing steadily declining growth, accompanied by a huge jump in profits.

It’s a very odd thing, and without the AI Capex boom, it’s really hard to imagine where the economy would sit right now. All that cash would have to go somewhere, perhaps in the form of even more buybacks, but man alive, all chips are on . . . chips.

Second, while I stand by the descriptor “blissfully unawares,” Nick Eberstadt of AEI published a 63 page paper Can a Depopulating America Still Flourish? and I thought perhaps “unawares,” was not entirely fair.

I did, however, settle-in to read the whole thing, and while it’s not bad, per se, it’s mostly . . . besides the point?

Read it yourself, but the gist of it is:

• we defeated Malthusian predictions before because see, e.g. the extraordinary productivity gains of industrial farming, in which fewer people produced much more food;

• the problem though is that too many people are out of the workforce, we’re too old, we spend too much on healthcare, and the deficit is unsustainable.

I mean, sure, but . . . that still doesn’t answer the question of whether or how we grow without growing?

At the outset, it’s very odd to begin with a heart-warming resolution to a Malthusian doomscape for an inverse-Malthusian problem. Yes, we have lots of precedent of producing more with less . . . which precipitated a massive population boom, which led to massive growth. What we’re now wondering is if we can produce ‘more with less’ to keep economic growth humming along without a population boom.1

Eberstadt doesn’t even address the core relationship between people-growth and economic growth, in any meaningful way.2 Nor does he look at any of the attributes of the current cycle, e.g. supercore inflation, tech-driven earnings, healthcare’s share of the labor market, etc. It’s mostly just, “well, not to worry, we’ll be able to create enough stuff to feed and supply everyone, but mind the deficit” and it’s like, “thanks, dude, no one was really worried about that, but glad to hear it.”

Idk. Maybe I’m being unfair, but “blissfully unawares,” remains apt and correct, imo.

ICYMI

2. Almighty Consumer Doing Gas and E-Comm

On the Almighty Consumer, it’s been a gas and ecomm story of late:

Transit, Gas, and e-comm (and I suppose lodging), are the only categories comfortably in the green.

Higher gas prices are clearly effecting other categories of consumption.

Unsurprisingly, discounters have been having a run, lately, as well:

Average transaction values at discount stores has been growing at a healthy clip since the middle of 2025.

The implication is (perhaps) that people are doing more of their shopping at the discounters.

In general, on an inflation-adjusted basis, there’s been something of a slowdown:

Real retail and food-services sales are flat-to-down since October.

Income growth has lagged inflation for a bit longer than is comfortable, as well:

Real wage growth has been negative for ~half a year, although I suspect the gap has closed somewhat recently (especially, accounting for fuel costs).

Anyways, it’s not a great look, and I’m not really sure what to make of it all, but things have held up, and a temporary energy shock isn’t the worst thing in the world. Plus, there’s been recent evidence of an uptick in hiring and wage growth, so some of this might be a bit lagged.

Probably tax refunds have helped somewhat:

Tax refunds are running ahead of years prior, and other stimulants may be helping the Almighty Consumer along.

But, bigger picture, as long as everyone stays employed (and they have), I don’t worry too much about the Almighty Consumer.

3. Healthcare[-Fraud] Domestic Product?

While on the subject of consumption, and particularly the growing importance of healthcare to consumption (and thereby, job-growth), this too was a striking chart:

Non-healthcare consumption as a % of GDP is actually lower than it was in the 1960s, and it’s been getting steadily lower (with the exception of Pandemania and Dotcom).

Put another way, without healthcare, we’ve been doing less and less personal consuming (as a share of GDP) for almost 70 years!

That’s pretty wild. It’s not an understatement to say we’ve been taking our wealth and plowing it straight into healthcare, and we’ve been doing it for an awfully long time. And, of course, given how unsustainable all that debt-financed spending actually is, it presents quite the pickle.

Random Walk has written at-length—and well-before it cycled into the mainstream—about how Healthcare Makes All the Jobs. Home health aids, specifically, have been the “private sector” jobs engine, especially in places like New York (but really everywhere). It’s part of what I’ve called the Rotation to the National Nursing Home and/or Healthcare Domestic Product.

That dynamic has helped disguise some underlying structural weakness in the economy because (a) healthcare isn’t really “growth” (let alone a pro-cycle);3 and (b) we’re paying for it with borrowed money that we can’t afford.

Even Random Walk, however, may have under-estimated the extent to which the home health hiring bonanza was disguising structural weaknesses.

I mean, how does one explain this chart with reasons that don’t sound in “fraud”:

New York has ~3X the number of Home Health Aids per senior than the US average (which includes New York)!

I’m taking the math at face-value, but assuming it’s right (and I haven’t seen anything to suggest otherwise), what innocent explanation could there possibly be for such an incredible discrepancy?

I mean, yes, New York made it dramatically easier to sign-up for “Consumer Directed Personal Assistance,” (i.e. relatives getting paid to provide elder care), which perhaps makes fraud extremely easy, but opening-up a self-checkout grocery store and pretending not to notice that the shelves are bare and the till is empty, isn’t really an excuse.

This is just staggering:

New York spends 85% more per resident in Medicaid than the national average.

And for an already-cash-strapped state (and city), the economic picture would have been even grimmer without letting healthcare go brrrr:

New York’s Medicaid spending per resident is already the highest in the country at $4,942 as of 2024, nearly $1,000 more than the next highest state. We estimate that if New York’s per resident Medicaid spending was the same as the U.S. average, they would have spent nearly $43 billion less in 2024. While it is a state program, it is a matter of federal concern as the federal government bears 57 percent of the cost.

New York unleashed a ~$43B stimulus in 2024 alone by lavishly spending taxpayer money on healthcare.

I know it’s partisan-coded, but it doesn’t make it any less-true: chase-out your productive, tax-paying residents and cover the shortfall with federally-funded medicaid fraud just isn’t a sustainable model.4

4. AI Lawyers for me, but not for thee

On the subject of AI-lawyering, this too, was hilarious.

AI-drafted pleadings have apparently exploded in recent years:

The share of filed complaints with “detected AI text,” is approaching 20% (and looks to be doubling every year).

Lawyers are very concerned about non-lawyers using AI for the “unauthorized practice of law,” but there is very little hesitation about super-charging their own outputs with LLMs.

Truthfully, this is probably a good thing. Well, not this specifically, because this finding is focused on complaints, and the last thing we need is to lower the cost of filing complaints below what it already is.

But let’s assume that this is true of the broader universe of pleadings (because I suspect that it is). In that case, my actual view is that AI isn’t going to displace lawyers (with or without regs), but it will increase their throughput (and perhaps automate and/or simplify some of the less-fun stuff). If AI is drafting a lot of the boilerplate recitations that lawyers require (by their own conventions) then there’s an element on ongoing absurdity, but at least it’s less strenuous and expensive.

This, however, is a less-promising (but unsurprising) development on the AI-generated text front:

“AI-Assisted” text on the internet is (supposedly) going parabolic.

So, it’s not exactly Slop-to-Table output, but it’s Slop-to-Table with a human-in-the-loop.

Let’s hope this is all some collection of necessary-but-boring text summarization that wasn’t all that high value in the first place, but I’m ever-reminded of this quip:

Obviously a lot has changed since March ‘22, but “the more summaries you consume in your life the stupider you become,” still rings true.

5. Long but not short

And, to wrap this up, an observation that probably means something, but what, I can’t really say.

Long-term interest rates have separated from short-term rates:

Short-term and long-term rates are still moving together, but they’re just dancing farther apart . . . and it all seemed to start with Trump’s election, or thereabouts.

I’m not really sure how to interpret this. Really, feel free to let me know what you think, because I’ve got only bad guesses.

I’ve long wondered whether higher long-term rates are just a thing, and the Fed can’t do anything about it (because long-term rates are driven by supply-and-demand, and the marginal buyer of US debt is looking at the deficit and demanding yield). In other words, the Fed can lower short-term rates, but long-term rates are whatever the market demands.

Idk if that’s what’s happening, or why it should specifically have started in Dec. ‘24 . . . maybe the market just decided that Trump was going to run the deficit structurally higher for ever? Again, idk. I just thought it was an odd thing.

Previously, on Random Walk

Random Walk is an idea company dedicated to the discovery of idea alpha. Find differentiated data, perspectives and people, and keep your information mix lively. A foolish consistency is the hobgoblin of small minds. Fight the Great Idea Stagnation. Join Random Walk. Follow me on twitter. Follow me on substack:

Get more from Moses Sternstein in the Substack app

Available for iOS and Android

Get the app

2

As an aside (to an aside), I found this result on AI productivity and hiring to be somewhat surprising:

Mining is a standout for AI adoption (with above-trend productivity), but has done little in the way of extra hiring (in fact, hiring is below-trend). Mostly, the point of the research is to show no-correlation between AI Adoption and Hiring (for better or for worse) because AI is a Job Maker, Not a Job Taker, but if you want to be gloomy, well the miners are going to be SOL.

👉👉👉Reminder to sign up for the Weekly Recap only, if daily emails is too much. Find me on twitter, for more fun. 

Subscribe now

👋👋👋Random Walk has been piloting some other initiatives and now would like to hear from broader universe of you:

(1) 🛎️ Schedule a time to chat with me. I want to know what would be valuable to you. 

(2) 💡 Find out more about Random Walk Idea Dinners. High-Signal Serendipity.

The marks aren’t high enough?

Ruminating briefly on equity values, public and private.

Tech is cheap?

Random Walk has alluded to this before, but the stock market has been a little odd lately.

It’s somewhat less-true recently after the big ceasefire(ish) melt-up, but it’s pretty rare to see earnings expectations skyrocket, while share performance scuffles. The only way that happens is if valuations collapse, given that share price is simply the product of earnings x the price of those earnings.

Unsurprisingly, valuations and earnings expectations tend to rise (and fall) together: if earnings go up, valuations rise (and vice versa). What’s happened recently, however, is the opposite of that—earnings are going up, but valuations are plummeting.

And really, it’s a tech story:

Forward earnings multiples for software have lost almost the entirety of their premium (relative to the rest of market), just as earnings expectations continue to grow—and not just grow, but grow at 2x the pace of the field.

Robust profits, with declining multiples, leads to weird observations about market performance, like this one:

Multiples on tech have dropped ~28%, while tech share performance fell only ~2%, leading to an unprecedented divergence in the series.

In other words, when the light blue line falls (multiples), the dark blue line (performance) usually falls too: multiples contract on relatively poor earnings results, and so share price falls accordingly . . . but if only multiples contract (while profits stay strong), then the blue lines separate.

And, as you can see, multiples and performance generally don’t diverge like that—post-GFC is the only period that comes roughly close.

Which is to say that someone is probably wrong. Either earnings will not, in fact, continue to grow (and share prices are too high), or earnings will grow (and share prices are too low).1

As it happens, post-GFC, the bears were definitely wrong:

If you were chastened to the sidelines by the stock market collapse in 2008, you would have to admit, with the benefit of hindsight, that there was no “collapse” at all.

So, the lesson here is “buy, buy, buy,” right? Perhaps.

ICYMI

Software terminally less-impressive(?)

It’s of course not necessarily the case that “someone is wrong.”

Read more

👉👉👉Reminder to sign up for the Weekly Recap only, if daily emails is too much. Find me on twitter, for more fun. 

Subscribe now

👋👋👋Random Walk has been piloting some other initiatives and now would like to hear from broader universe of you:

(1) 🛎️ Schedule a time to chat with me. I want to know what would be valuable to you. 

(2) 💡 Find out more about Random Walk Idea Dinners. High-Signal Serendipity.

5 Idea Friday

Some quick-ish hitters for your weekend reading pleasure.

1. The Smashing Success of DOGE

The consensus treats it as a foregone conclusion that “DOGE Failed.” This was shortly after consensus was sure DOGE was causing a recession, lol.

They’re generally quite happy about that declaration, partly because of negative polarization around the guy(s) behind it (and the sometimes icky and braggadocios comms that came with it), partly because they are champions (or participants) of the largesse that DOGE undertook to unwind, and partly because it’s fun to be mean.

It’s obviously true to some extent that DOGE did “fail.” It did not cut trillions of dollars from the budget, it did not (but also did) uncover lots of fraud, and most of all, federal spending continues to grow.

Well, sort of:

If you net out interest expenses, federal outlays actually did flatten out a bit, for what it’s worth.

Well, that’s something, but it’s too bad DOGE never got the chance to do more.1

This too is something:

DOGE really did dramatically reduce the size of the federal government, in at least one sense.

Total federal employment dropped pretty dramatically. Fewer people to gum up the works, and that’s not nothing.

And hey, since output/headcount is considered “productivity,” then I guess the Federal Gov’t is more productive now, too (lol).

ICYMI

2. Luxury Woes

A few weeks ago, Random Walk laid-out the definitive take on the current state of cyclical affairs. Read the whole thing, it’s great, the best even. No need to read anything else.

But, even so, one part of my theory had some holes in it, especially with respect to luxury:

Random Walk is pretty bearish on luxury, but at the time, markets had put some wind in luxury’s sails.

Whatever. Their mistake.

Behold, luxury is doomed:

Luxury revenues have been trending flat-to-down off their peak, since ~2023.

OK, so not exactly doom (yet), but hardly an impressive picture.

My understanding is that luxury is following the airline and hospitality playbook: wring even more out of its price-insensitive wealthiest customers (while everyone else pulls back).

3. Manufacturing Boom-let

Lots of smart (and not-smart) people scoffed at the notion of “reindustrializing,” America.

We’d never bring back blue collar jobs (and who’d want ‘em anyways). Tariffs won’t make a difference—if anything, tariffs will hurt domestic manufacturing. This is all a pipedream romanticization of a bygone hard hat-era that wasn’t all that great to begin with.

There’s a lot of truth to all that, and I too share some of the skepticism.

And yet, despite all the recent turmoil (or perhaps because of it), Random Walk has some bad news for the cynics: there is, in fact, mounting evidence of something resembling a manufacturing and industrial revival.

In no particular order:

• Philly PMI showing more business activity, including new orders (and more hiring and higher prices):

  

• KC Fed Manufacturing Composite Index up-and-to-the-right:

  

• Chicago Fed Survey of Economic Conditions, with yay! for manufacturing, and meh for everything else:

  

• Chicago Business Barometer (which tracks the ISM fairly closely) with a big jump:

  

• Empire State manufacturing Survey, showing a big jump in new orders, employment, and prices paid:

  

• Logistics Manager Index onwards and upwards:

  

• ISM Survey, with rising deliveries, production, and again, prices:

  

• Core capital goods orders have been pretty steadily positive (while Durable Goods tread water)

  

I could go on, but I think you get the point.

The soft stuff and even some the semi-hard stuff all point the same direction: manufacturing and industrial activity is trending upwards.

The rising demand for blue collar stuff flowed through to the freight market, as well (via Freight Alley):

• Truckload volume for 2026 is trending well-above prior years:

  

  

• Tender rejection rates (where a high rejection rate is usually evidence of tight supply) has climbed significantly:

  

• Trucking rates are unsurprisingly up all over the country:

  

Again, you get the point: there is mounting evidence that domestic industrial activity is buzzing.

It’s flowing through to employment, as well.

• Outside of healthcare (of course), manufacturing and loading & docking are the only sectors with above-trend job postings:

  

• Construction and Manufacturing small businesses were the only small businesses paying more to hiring firms than they were in 2023:

  

• Data centers have created a massive run on the skilled trades (for the umpteenth time):

  

Even Jensen Huang recently noted that the real AI infra bottleneck isn’t chips, power or memory . . . it’s plumbers and electricians.

And, of course, with great hiring demand, comes great wage growth (sort of):

The Dallas Fed survey shows steady upward inflection of wage growth for manufacturing since the end of ‘25.

So, what’s going on?

Well, obviously the AI buildout surely has something to do with it. Tariffs might have something to do with it, too. There’s also probably some Hormuzian front-running, going on as well, plus some Hormuzian end-running for domestic energy.

But, also, as Random Walk has observed before, the decrease in manufacturing investment that began at the turn of the century is truly staggering, and even the slightest mean reversion—driven by labor shortages, automation, overdue glow-ups, the electrification of everything, whatever—would constitute an meaningful upward inflection in industrial activity.

In a recent earnings call, JB Hunt’s EVP of Sales (for one of the largest freight cos) took the real or temporary question head on, and he was pretty unequivocal that this isn’t temporary:

“All proof points of structural change.”

Now, JB Hunt is talking his book. Plus, a lot of the commentary is specific to the structural changes in freight—yes, there is an uptick in demand, but there are other reasons that freight supply has decreased.

But either way, sorry haters, it seems pretty clear that industrial activity is definitely bumping.

4. Exactly the right amount of housing

Just another installment in one of the longest running Random Walk series: there is no housing shortage.

Rather than some secular shortage of housing, it turns out that housing supply is remarkably responsive to demand.

I mean, how else would housing units per capita stay so darn stable since the late 80’s (and higher than ever before):

The total stock of housing per 1,000 people peaked around 430 units in 1989 and has stayed pretty much exactly that high until the present.

Definitely no shortage there. Collapsing household size? Yeah, we got that. But housing shortage? Not a chance.

Indeed, new housing gets completed as new households are formed:

Outside of a little pre-GFC overbuilding, followed by some under-building, completed housing units have tracked household formation pretty darn closely.

. . . well, except recently, where new housing supply ran ahead of household formation, but that’s probably fine.

But yes, builders tend to build where the bodies are adding up, and when they stop adding up, they stop building . . . until the bodies add up, and then they start. Again, to wit:

Single-family permitting is down everywhere (after a big run of building, and now soft demand), while multifamily perks up (after a lull in building, precipitated by stagnant rent-growth).

They’re even building MF in the Northeast, where supposedly it can’t be done!

I mean, Random Walk told you the Palmetto State was the place to be:

Single-family permitting goes to where the people growth is, and if you think causation is running the other way, well, idk what to say.

All hail Myrtle Beach!

Anyways, more evidence of well-functioning supply-demand in housing, perhaps a moderate housing glut, and nothing at all like a shortage.

5. The real threat of AI

Thomson Reuters asked a bunch of lawyers about “GenAI’s threat potential to the [legal] industry” and you’ll never believe what the lawyers said:

73% said the “unauthorized practice of law,” was somewhere between “major threat” and “somewhat,” but, in all events, the biggest threat.

The lawyers are civic-minded folks, you see, and they just don’t want the normies hurting themselves by practicing law without permission. Because it’s the bar association that keeps us safe from unscrupulous lawyers, and umm, without all those horn books and law school exams, how could anyone be a lawyer in real life?

As an ex-lawyer myself, I can say with a high degree of certainty: experience definitely helps, but law school (let alone the bar or CLE) most definitely do not. Not one whit. I mean, law school is fun and interesting, but it has basically zero-to-negative bearing on real life lawyering.

A long time ago Random Walk quipped that AI would never replace lawyers because lawyers would make it illegal to do so and . . . yeah. People really don’t hate regulation enough, and professional cartels are only the tip of the iceberg.

Previously, on Random Walk

Random Walk is an idea company dedicated to the discovery of idea alpha. Find differentiated data, perspectives and people, and keep your information mix lively. A foolish consistency is the hobgoblin of small minds. Fight the Great Idea Stagnation. Join Random Walk. Follow me on twitter. Follow me on substack:

Get more from Moses Sternstein in the Substack app

Available for iOS and Android

Get the app

👉👉👉Reminder to sign up for the Weekly Recap only, if daily emails is too much. Find me on twitter, for more fun. 

Subscribe now

👋👋👋Random Walk has been piloting some other initiatives and now would like to hear from broader universe of you:

(1) 🛎️ Schedule a time to chat with me. I want to know what would be valuable to you. 

(2) 💡 Find out more about Random Walk Idea Dinners. High-Signal Serendipity.

Uncharted waters, blissfully unaware

Random Walk has demonstrated ad nauseam that the single most important theme going forward is “growth without growing.”

For the uninitiated, some context:

People-growth is hugely important to economic growth, but the ~80 year tailwind is coming to an end. For nearly a century, we’ve been able to drive growth by adding productive people to the mix: more output, more demand, more workers, more shoppers, etc. And when the domestic people-growth engine slowed, we fell back on immigration and globalization (and cheap credit), but still, more people, at-home and abroad.

But, people-growth is no-longer. The failure to babymake has aged the native born population, immigration has reached unsustainable (and zero-sum) levels, and international markets aren’t growing either.

From this point forward, growth is productivity-driven or bust. It’s uncharted waters, and no one really knows whether it can be done, but we’re sure as hell gonna find out.

And it’s funny, because the fact that we’re now growing slowly-to-not-all for the foreseeable future is the biggest economic change in over a century, and most people are barely aware it’s happening, let alone grok the significance.

I mean, sure, they’re starting to notice the side-effects:

• no job growth, but no unemployment? check;

• an AI “atoms-driven” industrial growth story? sure, that too—can’t have a durable goods, real estate, or retail story without more worker/shoppers/tenants;

• healthcare uber alles? yes, of course—old folks consume lots of healthcare;

• higher “supercore ex-housing” inflation? with labor in relatively short supply, the high-touch services will have to pay a premium . . . which makes services some combination of more expensive and/or worse (aka “service shrinkflation”).

Simply put, the economy is not getting larger in a very literal sense. Soon there may even be fewer mouths to feed, but especially, wealthy and productive mouths to feed. And that’s a huge deal!

While everyone was banging around about tariffs and the AI-took-my-job nonsense, Random Walk told you the actual biggest thing was the closed border, and Random Walk was right.

The open border was a deflationary growth surge, and unsustainable as it was, once over, it meant that job-and-shopper growth would grind to a halt.

And so it has.

The breakeven is lower now

A lower employment “breakeven” is the new normal now, and people (aka Fed Researchers) are starting to notice (all at once, apparently).1

First, the Dallas Fed says, “uhh, without immigration, the employment breakeven just isn’t a thing anymore.”

Comparing our break-even estimates with actual payroll growth (Chart 2, blue bars) suggests that job growth over the recent three months—December 2025–February 2026—has slightly exceeded the break-even rate on average, consistent with the unemployment rate remaining stable despite softer headline payroll numbers.

Real-time data point to an important change in the U.S. labor market: The benchmark for evaluating payroll growth has moved significantly. As net outflows of unauthorized immigrants reduced employment growth in late 2025, payroll gains that might historically have signaled economic slack are now consistent with a balanced labor market.

At this point, slow job growth says very little about the economy, other than the fact that the economy is adding very little in the way of potential workers.

The “benchmark” for a ‘good’ jobs print has changed. It’s much, much, lower now.

Along similar lines, and slightly more recently, researchers from the Fed BoG are likewise estimating a labor market breakeven converging on close to zero:

As population growth stalls, and participation rates already near post-GFC highs, there simply aren’t (many) new bodies to add to the labor force, such that it needed them.

The employment breakeven is now lower than its been in the entire series (going back to the 1960s).

Yes, this time it really is different!

When someone sounds alarm bells about a “bad” jobs print because tariffs or AI or whatever, grab ‘em by the collar and holler “well, what the hell else would you expect?!”

Or, more politely:

One important implication of a near-zero breakeven pace is that, even if the output of the U.S. economy (GDP) is growing at the same pace as potential output (potential GDP), employment growth in any given month is almost as likely to be negative as it is to be positive.

Yes, exactly that. Regardless of what GDP does, job-growth will be little ups and little downs, because there really aren’t too many alternatives.

It will look like this:

“A record 11 months in a row of alternating jobs-up-jobs-down.”

That’s just how it’s going to be, unless old folks and the youngs start coming off the bench, en masse (or the border is opened again). There just isn’t any other way.

ICYMI

Growth is a productivity story, from here on out

And will it matter?

Read more

First, the Burry prelude.

And then, the post from the archives:

It’s the interest rates, stupid

• decline in home values is obvious, if you know where to look

• CRE values are down, rents are down . . . but owner-occupied single-family homes just go up? Puh-lease.

• St. Louis Fed confirms: Random Walk is right, it’s the interest rates, stupid

• Decline in existing home values is showing up in clearing prices

👉👉👉Reminder to sign up for the Weekly Recap only, if daily emails is too much. Find me on twitter, for more fun. 

Subscribe now

👋👋👋Random Walk has been piloting some other initiatives and now would like to hear from broader universe of you:

(1) 🛎️ Schedule a time to chat with me. I want to know what would be valuable to you. 

(2) 💡 Find out more about Random Walk Idea Dinners. High-Signal Serendipity.

A Big Short Prelude

As longer-term Random Walk readers know, the non-existent “housing shortage,” is (and remarkably continues to be) the wrongest thing that most people believe to be true.

The proof has been in the pudding, as RW correctly took the under on all home price forecasts, while homebuilders consistently print GFC-level margins, reflecting the modest housing surplus, the nation currently enjoys. What the commentariat casually refers to as an “affordability crisis because NIMBYism” or whatever, is in reality a bid-ask “crisis” driven by the only actual “broken,” thing in the housing market, i.e. the 30-year fixed rate mortgage.

We either kill the 30-year, or . . . we wait. Those are the solutions. Sorry, I don’t make the rules.

• A nice places to live shortage? Yes, we’ve got one of those.

• A misallocation of large empty nests? Yes, perhaps something along those lines, as well.

But a housing shortage? Not in this universe.

Shortage-Bros will likely never fully capitulate, but every now and again, there is a defection. The latest is of Big Short fame who has now chimed in to say “Random Walk is correct—the problem in housing is a dislocation of the mortgage subsidy.”

Not quite in so many words, but you get the idea.

Burry then goes off the rails a bit when he concludes “and that’s why we should privatize the GSEs,” but he can be forgiven, I suppose, because while “privatizing” the GSEs probably wouldn’t do much to help the housing market, it would be quite beneficial to the equity-holders of the GSEs. I mean, who wouldn’t want to own a piece of the nation’s biggest mortgage lender with a free risk-transfer courtesy of the taxpayer?1

Anyways, let’s hear from the Big Shorter himself:

We studied housing square footage per capita adequacy, and found that there is no problem there. The US in fact has more residential square footage per capita than any other country in the world. This is not a housing shortage, despite what so many say.

I mean, weird way of putting it, but yes, no housing shortage. Check.

What else?

The problem is that bigger houses are inefficiently housing fewer people. The post-COVID low rate environment locked people into a lifecycle real estate position. Empty nesters can't sell, first time home buyers cannot buy. Second-hand home inventory is near all-time lows due to record low supply, not record demand. Prices are high due to the same reason.

Home equity is now a record $35 trillion, nearly doubling pre-COVID levels. 40% of homeowners own their homes free and clear - a record. And about 30% of all home buyers pay for homes without borrowing. Older homes were upgraded at a record pace during COVID, extending and refreshing the usefulness of residential real estate.

Again, with the word salad, but, true enough, what we have is (a) something of a misallocation of housing; and (b) rather than a “too-much-demand” problem, or a “net-new supply” problem, we have a “won’t sell” problem.

The phrase “lock in” is a misdirection, because Burry correctly points out that ~40% of homeowners have no mortgage (and the “mortgage lock-in effect” is not actually a thing), but the expression he is looking for is “sticky prices,” (and “unmotivated sellers”). He prefers, instead, the phrase “lifecycle real estate position,” (lol wut?).

Not a wordsmith. That’s ok.

But, here’s where he gets closer to the heart of the matter:

Artificially low interest rates, ~$6-7 trillion in helicopter cash and forgivable loans helped drive both the home updates and high housing prices.

Work from home moved the office into the home, often expensed or deductible. People with white collar jobs and means chose to live/work in exotic or remote locations.

All of this together does not speak of a housing shortage, or a housing problem.

Instead it is a problem of current residential space allocation and mobility, and this problem was created by government manipulation of interest rates, cash money supply, and COVID lockups that went on too long and changed work/home behavior.

Allow me to parse this in a way that, well, actually makes sense.

• Thing one: the specific pandemania “affordability crisis” was a massive pull-forward of demand caused by temporary-ish lifestyle changes and a boatload of subsidies (and NIMBYism ain’t got nothing to do with it).

• Thing two: at a deeper, more structural level, heavily subsidized mortgage rates—not shelter shortages—are the straw that stirs the home values drink.

Correct, again, Mr. Burry. You’re on a roll.

Please, continue:

Government created the problem and now maintains policies that prevent free markets from reaching a solution, not the least of which is keeping the GSEs inefficiently run while in conservatorship. Recall Pulte's video upon arriving at Fannie Mae - no one was in the office buildings. The companies have become atherosclerotic, inefficient government programs, while a decade of financial engineering optimized for homeowner wealth accumulation rather than housing market velocity/mobility/fluidity.

Government must fix this problem by facilitating efficient re-allocation of housing stock with higher housing velocity/mobility through the release of the GSEs into free markets.

This is a problem made for the GSEs. Through well-targeted programs, the GSE can help the free market find spaces to intelligently reallocate , and help US citizens with housing mobility.

Building more new overpriced, poorly built homes in increasingly dangerous flood zones and other hazardous fringe areas is not the solution. It adds to the problem through high maintenance burdens on new homeowners with little equity in their homes.

Rather, to build mobility/velocity of homeowners and housing space, the GSEs need to be recapitalized and retain easy access to capital markets. They also need to be run by real mortgage executives, not government functionaries.

To achieve this they need to exit conservatorship in a manner that excites markets to fund these companies, now with guidelines to prevent risk-taking outside of their purpose, and grow their purchases of mortgages of well-targeted specification.

Here again, one has to read between the lines a bit to tease out the gravamen of Mr. Big Short’s argument, but the key is to first rotate the page at a 90 degree angle, and then . . .

Nah. This last part just doesn’t make much sense, as best as I can tell.

This is what I referred to as the “going off the rails” part, where Burry just wants to make a boatload of money by owning part of a boondoggle and, in his defense, it’s pretty hard to make a coherent argument for why socialized risk ought to be privately profitable, so instead he throws a lot of incoherent ‘isht at the wall about “flood zones,” “mobility/velocity,” “real mortgage executives,” and “well-targeted specification,” and I suppose that’s as good an approach as any.

I too want “housing space,” and not “hazardous fringe areas,” so pay me, damn it.

Look, let’s be honest, the point here is Random Walk likes being right about things, and a famous investor personality said at least one thing that Random Walk really likes being right about, i.e. “there is no housing shortage.”

The rest is, I suppose, open to interpretation, but fun nonetheless.

Burry prelude complete.

Now, on to the republished oldie, but goodie, from Aug. 8, 2025:

When it comes to housing, it’s the interest rates, stupid

A brief reminder that there is no “housing affordability-shortage crisis,” but there is a 30-year fixed rate mortgage crisis.

Home values are coming down (because it’s the interest rates, stupid)

There is, of course, no housing shortage (and here), no affordability crisis, no lock-in effect, and no perma-upward-ratchet in home prices.

All that there is are sticky prices and unmotivated sellers. But just because transactions aren’t clearing, doesn’t mean that values haven’t declined (because they have).

When rates go up, asset prices go down, especially when the asset is heavily financed, like a car, or real estate. With higher borrowing costs, purchasing power goes down, therefore prices go down. It’s that simple.

The reason it doesn’t look like existing homes have gotten cheaper (as opposed to new homes where price depreciation is indisputable)2 is solely because existing homeowners are less inclined to sell in a downmarket, and they don’t really have to.

Banks bear the decline in home values, for the most part

All those increased borrowing costs are born by their banks, who are locked-in to low-rate 30 year mortgages—the loss in value shows up as “unrealized losses” on their balance sheets:

Low-interest mortgages have lost hundreds of billions in value, but the underlying collateral just gets more valuable . . . sure sure.

If you wanted to “fix” the housing market, the correct and effective solution would be to convert 30-year mortgages into floating rate or shorter-term fixed mortgages. If you did that, homeowners would be forced to internalize higher borrowing costs, and they would be much more inclined to sell—at prices that would actually reflect prevailing borrowing costs.

CRE values are down (because their mortgages aren’t 30 years long)

In other words, homeowners would become more like institutional property owners, and you’ll never guess what’s happened to the values of institutionally-owned properties:

Green Street’s commercial property price index shows property values lower than before the pandemic, with residential assets specifically, a bare 10% more pricey than they were seven years ago (and 20% lower than their pandemania peak).

10% appreciation over 7 years, is not very impressive, at all. Little wonder that RE-bros cannot raise money:

Real Estate fundraising is the absolute pits.

Why would no one want to invest in the asset that only goes up?

Rents too, are going down.

CBRE just revised all their rent estimates to the downside:

Every major market (except SF, LA, and Orange County) just got a downward revision in estimated rent growth.

Elsewhere, we know that median asking rents for new apartments has been in decline for ~2 years.

If you’re a shortage-affordability bro, how do you explain that divergence between CRE and owner-occupied? The answer is ‘you can’t.’

If, however, you accept reality, then the divergence makes perfect sense: CRE investors do not get 30-year fixed rate mortgages. They might get 5-year fixed rate mortgages, but then they have to face the music (and so asset-values fall).

St. Louis Fed confirms “Random Walk is correct—the affordability crisis is actually just a 30-year fixed rate mortgage crisis”

Not only is it intuitively obvious that the 30-year fixy is causing the distortion, the St. Louis Fed just did some research to prove it.

Sweden and Canada, where shorter-term and adjustable rate mortgages predominate, real home values are either breakeven or lower.

US home values, by contrast, remain curiously stable.

It’s not curious, though. It’s still just this:

Sticky prices, unmotivated sellers.

No shortage. No affordability crisis. None of that.

Ability to pay is an outer limit:

Monthly home payments, as a % of disposable income, are remarkably stable since 1990 (“housing crisis,” notwithstanding).

People can’t just manifest more money to throw at homes—they’re constrained by their incomes, and their borrowing costs. If the latter goes way up, while the former only goes up a little, then home values come down.

In any event, home values are coming down

Again, that home values would come down—even for existing homeowners—in a higher rate environment has always just been a matter of time.

Eventually, slowly, but surely, homeowners would have reason to sell, and they will have to find buyers where they are.

In the Southwest and West Coast, that’s already happening:

About 1/3 of major US markets are experiencing resale price declines.

There are still some primo locations where when the occasional house actually trades, it’s trading for more than before, but on a volume basis, they are the exception, not the rule.

And price declines aren’t just happening in YIMBY-paradises like the Southwest. CA, Colorado, and the pacific northwest are experiencing the same thing.

Random Walk’s favorite indicator (and recent meme-fancy) Opendoor , shows the same thing:

Opendoor shows negative price appreciation, trending lower than last-year, and much lower than the post-GFC average.

Home values are coming down.

I don’t make the rules. I just watch the rules playout as expected.

Previously, on Random Walk

Random Walk is an idea company dedicated to the discovery of idea alpha. Find differentiated data, perspectives and people, and keep your information mix lively. A foolish consistency is the hobgoblin of small minds. Fight the Great Idea Stagnation. Join Random Walk. Follow me on twitter. Follow me on substack:

Get more from Moses Sternstein in the Substack app

Available for iOS and Android

Get the app

2

New home values, which is where all the price discovery lives, have come down, while existing homes “go up”:

For the umptheenth time, it’s not because people suddenly love “that used home smell.”

• a preface: the game done-changed for everyone

• show me the money (pretty please)

• when to cut bait and/or how do you know when a fund is what it is

👉👉👉Reminder to sign up for the Weekly Recap only, if daily emails is too much. Find me on twitter, for more fun. 

Subscribe now

👋👋👋Random Walk has been piloting some other initiatives and now would like to hear from broader universe of you:

(1) 🛎️ Schedule a time to chat with me. I want to know what would be valuable to you. 

(2) 💡 Find out more about Random Walk Idea Dinners. High-Signal Serendipity.

When to cut bait?

Let me preface with a caveat that this is not a dunk on Venture.

Most, if not all, of the “doomer” observations about VC are applicable to the capital markets, more generally.

• the game-done-changed for everyone—a dramatic shift in the cost of capital will have that sort of effect;

• the end of an 80-year demographic tailwind to growth, applies everywhere;

• a buy-high-sell-not-as-high bid-ask overhang, causing a pile-up of unsold assets, a deeper “j-curve,” and thereby a dry spell for liquidity, is every manager’s and allocator’s problem;

• that that dynamic has lead to increasingly concentrated capital in the largest of funds (who are making increasingly concentrated investments in the largest of companies), and correspondingly tough-sledding for smaller managers, is true in PE, Real Estate, Credit . . . the works.

There’s obviously some nuance there, but if you were to observe the underlying challenges in VC that are downstream of a major cyclical regime-change, and conclude “something is wrong specifically with VC,” then you’d be more wrong than right.

If anything, VC with its uniquely pareto-like distributions, and bias towards technology (which will certainly define the next cycle), makes it a relatively attractive buy-fear-sell-greed opportunity, in Random Walk’s humble opinion.

Show me the money (pretty please)

Throat-clearing out of the way, this is a very striking chart, as it pertains to the deepening of the j-curve, for VC in particular:

The only two categories of VC funds to have DPI > 1 (i.e. returned more cold-hard-cash than received) are top-decile funds from the 2017 and 2018 vintages.

That’s it. Every other LP is still waiting for their payday. The point being that allocators to even the very best of funds have gone an awfully long time without seeing their money back.

Read more

👉👉👉Reminder to sign up for the Weekly Recap only, if daily emails is too much. Find me on twitter, for more fun. 

Subscribe now

👋👋👋Random Walk has been piloting some other initiatives and now would like to hear from broader universe of you:

(1) 🛎️ Schedule a time to chat with me. I want to know what would be valuable to you. 

(2) 💡 Find out more about Random Walk Idea Dinners. High-Signal Serendipity.

Continuing to ignore the main story in the timeline to surface other interesting things, of note.

AI-Beneficiaries-in-Training

One of the many interesting puzzles about today’s market is that despite the near trillion dollars spent on AI infrastructure, investors are pricing-in almost no downstream benefits of AI.

All the returns are flowing to the picks n’ shovels making the transformational technology possible—a trillion dollars’ worth of invoices will have that effect—but for the transformees, the net-present-value of their future cashflows is either flat, or worth less than before.

The Goldman “AI Basket” is stumbling along with the rest of the Nasdaq100, while infrastructure runs away from the field.

And that’s strange.

Random Walk has made this observation before, but it bears repeating, because the zeitgeist doesn’t seem to get it yet: investors are making a massive bet on AI, while making basically no bets on AI.

Why might that be the case?

• One possibility of why that may be is some combination of uncertainty and opportunity cost. We don’t know who the transformed winners will be, but we do know who the transformers are, so investors are placing their chips accordingly. Who has time to speculate about two AI adopters in the bush, when you’ve got plenty of semis and turbines in the hand?

• Another possibility is that, in fact, the transformers are, and forever will be, the only winner, either because (a) the AI buildout is like an expensive meal that no one will eat, but still must be paid-for; or (b) the buildout is the new tollbooth for what will simply be a ubiquitous higher cost of doing business, going forward. Companies will be transformed, but the relative value-capture lives at the infra layer.

I’m sure one could come up with some other possibilities, but for present purposes, here’s the point: of these three possibilities, there’s really only alpha in one of them: option #1—there will be transformed winners, we just don’t know who they are yet.1

ICYMI

First you must believe

Markets have already priced-in accelerated growth for the buildout (although perhaps not enough, or perhaps too much), so, at this point, to be early in the AI trade, one has to anticipate some other acceleration, ideally before it shows up in earnings (or at least before those earnings are reported to the street).

It sounds obvious (because it surely is), and yet, for the most part, there really aren’t any companies trading on the promise of growth-to-come-but-not-yet-started.

And yes, one has to believe that AI/Tech will, in fact, lead to acceleration . . . because otherwise, why all the Capex, but also, because it does not seem that hard to believe:

Ramp’s spending data shows a very striking divergence between those spending on AI and those that are not: the former are growing much more quickly.

• There is definitely sample-bias here, and it’s not obvious which direction causation runs.

• It’s also possible that high-growth companies are simply passing through AI costs at negative gross-margin, too, which is a neat trick to generate topline revenue growth, but obviously not a signal of sustainable AI-driven expansion.

• Even more probably, it’s almost certainly the case that “AI companies” (who necessarily spend a lot on AI) are the high-growth companies because AI is the thing everyone wants to buy, but again, that doesn’t mean that “investing in AI makes you grow.”

Those are valid concerns, but for now, I’m going to accept the premise that AI does, in fact, have some properties of a wonder-drug.

So, it got me thinking: who is investing in AI today for benefits that will materialize sometime tomorrow. It’s an obvious question, that nonetheless, markets have left entirely unanswered.

Follow the R&D

That question too, lends itself to, at least, two categories of companies that might fit the bill:

Read more

Read more

• oil prices split into three

• no one knows what’s happening

• …except maybe this one guy, who has some astute observations about what’s happened before, and why this time may be different

• reasons for optimism, but also, there’s almost no way there’s at least some pain

👉👉👉Reminder to sign up for the Weekly Recap only, if daily emails is too much. Find me on twitter, for more fun. 

Subscribe now

👋👋👋Random Walk has been piloting some other initiatives and now would like to hear from broader universe of you:

(1) 🛎️ Schedule a time to chat with me. I want to know what would be valuable to you. 

(2) 💡 Find out more about Random Walk Idea Dinners. High-Signal Serendipity.

A tale of three prices (for oil)

While the general timeline toggles between “the war is total victory,” and “the war is total disaster,” Random Walk keeps a steady grip on “it’s too soon to tell, and we lack the information, anyway,” which is perhaps less viral, but nonetheless the most correct.

Cost of oil depends on who’s asking

That said, what’s going on in commodities right now, is pretty wild. I mean, the price of oil is . . . well, it depends on who you’re buying from:

Oman is selling crude at ~20% higher prices than Dubai (and now even higher) . . . while in the US, oil is wiggling around a lowly $100.

It seems hard to expect that a commodity would continue to trade like this, since it’s perhaps the core attribute of a commodity that it’s generally governed by global supply-and-demand. But, lol, wouldn’t that be a BRICS realignment backfire, if it turns out that the oil market fractures into two . . . Hormuzian oil:Expensive—US Oil:Cheap.

Not that this is a laughing matter, in the main.

Apparently, demand-destruction is already underway for Asian markets (where ~80% of Strait traffic is typically bound).

But, at the same time, short interest on oil is apparently quite high, as well?

Short interest on the is near its pandemic high.

From of all of this I conclude: no one knows what’s going on, but hold on to your butts!

ICYMI

Structural realignment, accelerate!

For a much smarter conclusion, however, I find inspiration elsewhere.

Read more

Read more

👉👉👉Reminder to sign up for the Weekly Recap only, if daily emails is too much. Find me on twitter, for more fun. 

Subscribe now

👋👋👋Random Walk has been piloting some other initiatives and now would like to hear from broader universe of you:

(1) 🛎️ Schedule a time to chat with me. I want to know what would be valuable to you. 

(2) 💡 Find out more about Random Walk Idea Dinners. High-Signal Serendipity.

Let’s see what the cycle dragged in

This is a long-ish one, but Random Walk invites you to take a break from the show in Hormuz to ponder the new cycle. Consider it something of a reformulation of some longer-running themes, with some fresh data to back it up.

The two bottlenecks

One way to think about the next cycle (or really the current cycle) is that there are two bottlenecks that are two sides of the same coin: (1) workers; and (2) worker-enhancer/substitutes (i.e. compute, energy and robotics, and yes, software too).

• Without a steadily increasing supply of people to do things, there will be an increasing premium on the people who can do things, and/or technological alternatives to getting the work done. That premium, of course, assumes we’ll keep finding new things to do, even though the slower growth of new people will challenge net-new demand, which it surely will.1

• On the other hand, what we won’t really need more of is the stuff that people buy, especially big durable stuff, like cars, houses and refrigerators. Likewise, we’ll see a (continued) shift away from big expensive goods and services, and towards cheaper stuff/experiences, and of course, healthcare . . . lots and lots of healthcare.

• Finally, the stuff/service providers who operate most efficiently, will likely accelerate their share-gains relative to the ones that don’t—if the pie isn’t growing, then the inevitable move is to take a larger piece of the pie. Winners will win, but even more so than before.

Random Walk can’t/won’t noodle it all out in one post alone, but I’m planting this flag as a noodling-point for future noodling on these themes.

For now, we’re already seeing the cycle play out in a few ways, for better and for worse, so let’s dig in.

The great industrial rotation

The first sign is the great market rotation to industry, i.e. the infrastructure (and people) making all the worker-enhancement/substitutes possible. If machines are going to carry even more of the load going forward, then the mechanized substrate is in desperate need of a glow-up, and the market is all over it.

The cycle’s new winning sectors

Consider the shift in cyclical multiples and where they stand in relation to the broader market:

Of the big cyclical sectors trading at a premium to the index, we have:

• Semis;

• Capital Goods; and

• Consumer Discretionary (which we’ll return to at a moment).

Materials aren’t trading at much of a premium relative to the index, but the relative price of their earnings is somewhat higher than its been.

In all events, the rotation leaders make sense. The premium is on the industrial glow-up.

In terms of the laggards, those would be cars, consumer services, insurance, durables, and fin-services. That too generally makes sense, insofar as those businesses are linked to the consumer economy (in the main), and that’s not the growy part.

If you double-click into consumer discretionary and industrials it becomes even more obvious:

Of the winners:

• it’s all the commercially oriented industrial sub-sectors, like electrical equipment, construction, etc.;

• e-commerce adjacent;

• aerospace & defense; and

• luxury and specialty retail (again, more on those in a bit)

Of the losers:

• planes, hotels, leisure;

• cars and household durables; and

• professional services;

That too mostly makes sense.

All the money flowing to the big industrial buildout, and the scarce labor to make it happen, is cooking up some big things for the gridiron heroes (and not the football kind).

You want to invest in the US industrial supply chain of goods n’ services? There’s an ETF for that ($PAVE). Oh, you manufacture gas turbines? Put on your party hats because company has arrived it comes bearing gifts:

No one can get enough of the big expensive stuff necessary to build data centers and the electrification of everything.

That’s quite the run for some historically pretty sleepy cyclical businesses (that have mostly been around for a long time). For a market that’s been dominated by high-flying tech names for so long, it’s a pretty remarkable thing.

On the other hand, for those businesses selling anything but cheap stuff to a slow-growing consumer market (ideally online), you’re not feeling the love. There are some exceptions—like luxury and the non-cheap stuff parts of retail are surprising out-performers—but personally, I’m fading those, as driven by some part misbegotten “tariff relief,” “flight to safety” and/or “lack of better ideas.”2 To be clear, the Almighty Consumer is fine, just not overly growy.

Bigger picture, I think the market has got this mostly right, but not entirely right. Money is flowing to cap-intensive builders and sellers of cheap tchotchkes, and away from high-end consumer services and big expensive goods. Smart.

ICYMI

Capital Intensity is back (and a Capital-Light era fades)

Let’s ruminate for a moment on the market’s newest high-fliers (that are actually pretty old) because the gravity of the thing shouldn’t be underestimated.

Read more

Read more

👉👉👉Reminder to sign up for the Weekly Recap only, if daily emails is too much. Find me on twitter, for more fun. 

Subscribe now

👋👋👋Random Walk has been piloting some other initiatives and now would like to hear from broader universe of you:

(1) 🛎️ Schedule a time to chat with me. I want to know what would be valuable to you. 

(2) 💡 Find out more about Random Walk Idea Dinners. High-Signal Serendipity.

Of Secondaries and IPOs

Secondary markets are getting deeper and more liquid. Rather than complain about it, adapt and survive.

IPOs still bigger than secondaries, but barely

2025 was nearly a watershed moment for secondary sales.

Secondary liquidity nearly matched IPO liquidity for VC this past year:

At a 31% share of exit value, private secondary sales generated nearly as much liquidity for VC as IPOs.

Secondaries went from basically a rounding error in the liquidity picture, to an equal participant alongside acquisitions and public-market exits, in just 5 years.

Now, in the big scheme of things, secondaries are still tiny relative (~4%) to the stash of asset value currently managed by VC. As foretold, secondaries are an increasingly big deal, but they’re not an off-ramp or cure-all for the mistakes of the prior regime.

Going for alpha

Pretty much the opposite of a cure-all, really.

Secondary shares are increasingly trading at a premium, and not a discount.

Rather than go bargain-hunting (for the most part), secondary investors have trained their fire on the same 15-20 companies that everyone wants to invest in. Secondary investors are hunting whales, not charity cases.

This isn’t a bag-passing exercise. It reflects an evolution of the private side to compete with publics.

Public markets aren’t closed. Private markets have simply gotten larger to meet the capital needs of the best companies, who, all things being equal, would rather not have to deal with the headache of going public, if they don’t have to. And so they don’t.

Oh, you need some secondary liquidity to keep your employees happy and fed (tired, as they are, of waiting 15 years for an IPO)? We gotchu:

2025 tender offer volume nearly recaptured peak-ZIRP highs.

By getting bigger, private investors get privileged access to the hottest action, and private companies get capital, without all the public market strings attached. Supply meets demand. Adapt or die, public markets.

Lots of people can complain about companies not going public (and want to blame VC or the companies themselves), but capital is capital, and if private capital comes with less unwanted baggage, then public capital needs to up its game.

ICYMI

Deeper and more liquid (and therefore deeper, and soon, even more liquid)

The other interesting thing is that private capital’s relative advantages are self-reinforcing (and therefore will continue to compound, with time).

Read more

Read more

👉👉👉Reminder to sign up for the Weekly Recap only, if daily emails is too much. Find me on twitter, for more fun. 

Subscribe now

👋👋👋Random Walk has been piloting some other initiatives and now would like to hear from broader universe of you:

(1) 🛎️ Schedule a time to chat with me. I want to know what would be valuable to you. 

(2) 💡 Find out more about Random Walk Idea Dinners. High-Signal Serendipity.

1. Citrini asks all the wrong questions, mostly

The current thing moves so quickly these days that this is a bit stale, but between snowdays and other commitments, I couldn’t wrap this up until today.

Anyways, I don’t really follow Citrini closely, but I’ve always generally liked what I’ve read, when I’ve read it. In case you didn’t already know, Citrini published a doomsday “scenario analysis” of “abundant intelligence,” or whatever.

Maybe people were stuck at home with too much internet on their hands, but Citrini unleashed havoc:

Companies mentioned in the post sold off en masse, well beyond the usual SaaS-Attack.

• Visa? Mastercard? Who needs ‘em? AI agents are just going to use stablecoins and skip the fees.

• Doordash? Well, AI agents don’t need a loyalty program, so that one is toast too.

• Private Credit? Well, we know the story there…it’s a house of cards built on loans to software, and if software goes down, so too does Blackstone and Apollo.

Of course, it’s not just some companies that are toast, in this scenario.

AI Agents are going to do everything, dramatically lowering costs (yay!), but also deprecating the white collar workforce to paperweight status (boo!). Layoffs everywhere, spending and housing collapse, and everything goes to hell.

OK. Maybe I’ve missed some key parts, but that’s basically the gist of it.

Kudos to Citrini for a fun thought experiment that also happen to set the internet and the investment community on fire, at least for day. And honestly, the most bearish thing about the whole event is simply how flighty and mercurial the investment community can actually be, especially when the boat gets arockin’. (Or maybe we should just appreciate how seemingly stable and coordinated markets generally appear to be?)

As for the actual substance of Citrini’s post? Lots of people have already chimed in along similar lines, but the very short version is that I’m generally not sure what all the fuss was about.

I mean, none of it is that original. “AI will kill white collar jobs and incumbent software businesses too, and then everything built on that artifice implodes,” is pretty basic. I too have wondered whether the increasing onus on profitability leads to more headcount cutting, so there’s that.

On the plus side, I suppose it is interesting to think about how “agentic shopping” will impact payment rails (in some way other than “I gave the AI my credit card,”) but the more interesting questions are around search, brand, marketing, etc. (which Citrini doesn’t really get into at all).1 What’s an Agent-friendly UI/UX anyways? Likewise, if the previous tech titans were built on network effects, is there any comparable high-margin moat for the AI era?

It’s food for thought, for sure, but nothing worth updating one’s priors over, imo.

In terms of the flaws, well there are a bunch, but let’s focus on the big ones.

ICYMI

Compute is the opposite of frictionless

The first major flaw is that Citrini rests his scenario on a demonstrably false factual premise: the notion that “friction,” (i.e. the cost of compute) “goes to zero.”

In other words, the entire scenario is premised on the exponential growth of AI doing everything by directing AI to do everything, including all the decisions that humans might otherwise direct AI to do. In Citrini’s telling, AI replicates and scales at basically zero marginal cost.

And that’s just . . . exactly wrong?

• Sure, inference is getting cheaper, but the whole reason that all this Capex is underway is to buy the GPUs (and generate the electricity) to run this whole shebang.

• Model providers are routinely throttling users, and almost certainly subsidizing all the silly image and video creation to further adoption.

• Every company and board member I’ve spoken with is actively thinking about token-cost as a rapidly rising expense item. My prediction is that it won’t be long before companies start to think in terms of “ROIT” (“return on invested token”).

If anything, rather than “friction goes to zero,” I’m more worried about a yawning gap between (a) the price of compute required for a sustainable OAI etc. business model; and (b) the price customers are willing to pay for the amount of compute required to routinely discover and deliver meaningful business outcomes.

OAI cannot go on like this, mostly subsidizing consumer adoption at a staggeringly high cost.

Remember that AI is really only good at code, right now:

Software engineering is by far the dominant arena for agents, in Claude’s admittedly biased sample.

As any entrepreneur will tell you, writing code is not the same as building a sustainable business, product, or feature. Don’t get me wrong—it’s a very big deal, but if building a business was as easy as modeling one in excel, the world would be a very different place.

And AI, left to its own devices, is still pretty unreliable:

Model capabilities are improving exponentially (by one metric), but “hallucination” rates are still quite high.

Agents making things up 28% of the time (at the low end) isn’t going to put the white collars on the breadline, just yet.

An imperfect analogy would be to think of AI as currently in its luxury product phase. A private jet and/or one of those first gen clunky car phones might be justifiable for certain high-end users, but there may not be a “mass marketable” product, just yet. If models started actually charging for all the “friction,” I’d be concerned that adoption would slow dramatically.

Keep in mind that another iteration of that scenario is still bearish, but not in the ways that Citrini suggests.

Perhaps AI is ready for the mass-market, but again, given the definitely-not-zero friction of deployment, all the extra-value generated by AI is accretive to the model layer, and not the enterprises that deploy AI. When SaaS incumbents (or Walmart) eventually integrate AI, it accelerates growth, but it turns out to be relatively low margin growth because compute becomes a big chunk of net-new COGS.

That’s bearish for SaaSCos, but not for the economy writ-large (and certainly not for the hyperscalers and/or frontier models).

Or, alternatively, AI is ready for the mass-market, but competition between the model-providers (and the rapid depreciation of state of the art models that don’t stay state of the art for long, not to mention the depreciation of the underlying hardware) means that model-makers spend enormous amounts of money for what’s ultimately a commodity.

That’s bearish hyperscalers and model labs, but bullish SaaSCos and the economy writ-large.

For what it’s worth, there’s plenty of evidence of companies routing to less-up-to-date models for the jobs that those models can do, which means outdated models do retain value.

Plus, pricing for older GPUs looks pretty solid, too:

Rental pricing for older H100s and A100 GPUs perked up recently.

Anyways, the broader point is that this stuff is expensive—perhaps too expensive—and not that it’s not expensive enough.

Open Weight Challengers

Now, do Open Weight models present a challenge?

Yes, sure. If frontier models come out with the next great thing and far cheaper Open Weight models follow on their heels, then that’s an unwinnable game.

Open Weight models lag proprietary models on performance, but not by much.

That said, my strong suspicion is that the (Chinese) Open Weight model makers are just stealing (and Anthropic thinks so, too), and while there may be some karma involved, I would expect the labs to get smarter about this, and perhaps a legal-political solution may follow, as well.

But either way, the idea that AI scales infinitely without any marginal cost just seems absurd.

To the contrary: what makes AI different from software of yore is that it doesn’t replicate at zero marginal cost—using AI costs money, every time. That raises interesting questions around how to value AI-driven growth (and is it more like the asset-heavy cycles of beforetimes), but ‘what if it’s manna-from-heaven?’ isn’t really where I’d spend much time.

AI is a productivity unlock without a surplus?

The other big error is a conceptual one, or really just a failure of imagination.

Read more

Read more

👉👉👉Reminder to sign up for the Weekly Recap only, if daily emails is too much. Find me on twitter, for more fun. 

Subscribe now

👋👋👋Random Walk has been piloting some other initiatives and now would like to hear from broader universe of you:

(1) 🛎️ Schedule a time to chat with me. I want to know what would be valuable to you. 

(2) 💡 Find out more about Random Walk Idea Dinners. High-Signal Serendipity.

The Jobless Productivity Boom

Bloomberg dropped a big story on the “unprecedented ‘jobless boom,’” and y’know, that’s a thing around these parts:

Forecasters expect Friday’s report on gross domestic product to show the economy expanded 2.7% in 2025, a solid pace by any standard for a developed country.

But employment barely grew, and the combination is drawing comparisons to the infamous “jobless recovery” of the early 2000s that followed the tech bubble and collapse.

Growing without growing, how will we do it? It’s the question all the Random Walk people want to know.

I’m glad Bloomberg is finally waking up to the Random Walkian view of the world, but sadly I wasn’t asked to comment. Of course, this is a story about demographics first and foremost, and yes, the parallel to the post-Dotcom “jobless recovery,” is interesting.

What Bloomberg does not report is that the jobless recovery wasn’t good enough, and so eventually the Fed began lowering rates to subsidize demand (especially for housing), and we all know how that ended. It’s a tricky thing when the actual growth of consumers begins to taper off (a problem that we “solved” by ramping up migration flows).

ICYMI

Step 1: Build Data Centers

For now though, the answer to growth without growing has been AI Capex (and healthcare).

The former is relatively people-light, compared with other big fixed investments (like residential real estate), which is probably a good thing, given the dwindling supply of labor:

Data centers have created a run on specialty contractors and engineers, but elsewhere, construction employment is pretty tepid.

Bigger picture though, a “jobless” boom is basically the only kind of boom we could plausibly expect.

It’s of course not entirely jobless—the workforce is still growing, albeit very slowly, and best-case scenario, the “boom” continues to create enough net-new opportunity to absorb those entrants, and keep unemployment low. That’s just what a lower “breakeven” means.

Step 2: Hire Lawyers and Architects

So far, so good.

There’s even been a moderate uptick in white collar prospects (to my surprise), once you clear out temps and business support services:

Professional Services employment is on net breakeven, but lawyers, architects, and consultants are getting hired a little more than before.

It’s the temps and support, i.e. lower down the foodchain, that are less in demand, but the white collar folks have a little pep in their step.

Again, that’s pretty good. My guess is that it will ebb and flow, but that’s better than just ebb.

Step 3: Keep it up, but mind the supercore

Where do we go from here?

Read more

👉👉👉Reminder to sign up for the Weekly Recap only, if daily emails is too much. Find me on twitter, for more fun. 

Subscribe now

👋👋👋Random Walk has been piloting some other initiatives and now would like to hear from broader universe of you:

(1) 🛎️ Schedule a time to chat with me. I want to know what would be valuable to you. 

(2) 💡 Find out more about Random Walk Idea Dinners. High-Signal Serendipity.

The hottest thing in employee comp, and other labor market missives

Buckle up.

Healthcare makes all the jobs (bc obviously)

Another day, another jobs release, and the story remains the same:

“If you exclude healthcare, the US has actually lost jobs since 2024.”

2024 is a bit of an arbitrary end-point, but the point stands, same as it ever was: Healthcare Makes All the Jobs.1

Expand the range to the past decade, and the story isn’t all the much different:

The Top10 Industries for job growth are old folks services, ecomm-stuff, and a smattering of pets, booze, and tech.

In other words, the jobs economy is fairly described as services for lonely, stay-at-home seniors and their pets.2

This is one of Random Walk’s longest running themes that was first mocked and scorned, and now, well, it’s just widely accepted reality. We’re getting older, and the customer has the national credit card on autopilot.

Unsurprisingly, job-growth was predominantly lady-driven:

All the employee gains were made by women, while total male employment dropped.

Healthcare as an industry skews female, so healthcare hiring does too.

And, in an economy with basically one pro-cycle firing (i.e. AI), and otherwise preoccupied with an acycle (i.e. aging), it’s unclear where additional demand for workers comes from (other than healthcare). And that’s been true for a while . . . job ‘growth’ was really job ‘catch up,’ and as the gap closed, so too did the number of ‘new’ workers.

ICYMI

It’s immigration, right?

Naturally, the other thing that’s going on here is the drawdown in new arrivals.

Or rather, it’s definitely the case that new arrivals were the driving force behind employment gains, mostly filling the gap left by retirees.

Whether new arrivals create their own demand for new labor, I suppose, is a separate question, but either way, there isn’t that much evidence that demand itself has increased (and, again, that’s been true for a while):

Employment gains have been driven almost entirely by additional supply, since the post-pandemic recovery began (according to this Fed model).

The way this model works is basically to look at job growth and wage growth at the same time.

• If jobs increase, and wage growth slows, that means job-growth is supply meeting demand.

• If they both increase, that’s expansion for both supply and demand.

• If wages grow ahead of job growth, then demand is in the driver seat.

In this case, increasing job growth combined with decreasing wage growth imply that new labor supply (by far) was the principal driver of employment growth. Demand though, has been mostly negative (because otherwise wages ought to have kept on rising, or at least, not have decelerating quite so dramatically).

In other words, the Open Border was deflationary, in that it lowered the cost of service workers by closing the retiree-aging-driven “worker shortage.”

So, perhaps the border was closed just-in-time, i.e. the labor market was hitting a saturation point, and had we kept adding migrants, they would have been increasingly unemployed.

Another possibility is that migrants themselves are a source of new demand (and by closing the border, that demand was knee-capped). That would make migration a sort of infinite money-glitch (a theory that has not (yet) been born out in Canada, or Europe, fwiw), but certainly adding people, is well correlated to adding GDP.

Interestingly, one possibility is that newcomers create their own demand, but not in the way that people conventionally think.

Read more