Below is a chart of OpenAI FCF projections until 2030. Those dark red numbers are cash flow losses that need to be funded. They are big, rounded out; it adds up to circa US$80B. They are also projections; maybe they are optimistic. OpenAI, in many ways, is the poster child of the AI era. It launched the most popular chatbot and has over 800 million monthly active users, the quickest take-up in history. Whether it wins or not will, of course, depend on winning customers and offering what they want and ultimately funding itself. Secondly, it is pivotal to the AI ecosystem, which includes a big slice of the semiconductor industry and the DC/hyperscale industry. If OpenAI stumbles, it will be felt across the whole ecosystem, that is guaranteed. For example, NVDA GPUs find their way into OpenAI and its competitors big time. All these OpenAI losses are someone else’s profits. On top of this are all the competitors and collaborators (Musk, GOOG, AMZN, MSFT, Meta, ORCL) spending big time to match OpenAI, hyperventilating the ecosystem.
I've been around a long time and have seen many similar scenarios, ok, at a much smaller scale-lol, and they usually end badly. The market does not have the patience to fund losses for years into the future. We saw that with the tech wreck, where even those with half-decent ideas had funding cut and perished. That is what I want to focus on here, the funding, not the use cases. Both are important. My theory is that funding has changed big time since the GFC.
The big question is, can the funding continue, and is it different this time? I argue it is different this time and in a big way. Will it be enough to fund the losses over this time, that I do not know and would not own OpenAI if it were listed? But the risks are lower than in the past, imo, why?
The differences come about due to the changes the Western world went through during the GFC. Although it could be argued that they were in train from financial deregulation in the 1980s. The biggest catalyst was the destruction of the Western world's banking system in 2008/9. Ok, that’s huge, when every bank in the West (outside of the Australian and Canadian banks, where the regulators have to be congratulated, as well as the management team involved-that's another story) goes under, something has to give. What gave was the traditional banking-induced credit cycles. Through regulation and management caution the banks pulled their collective heads in. IMO, this is simply the biggest change we have seen in finance in the last 20 years and has had huge ramifications across the whole finance and investing world, although no one seems to talk about it. The second-order effects of the end of the banking-induced credit cycles have many implications, one being a lower chance of recessions, because banking liquidity does not matter as much. Another implication is that growth as a factor outperforms value, as secular growers become more precious; another is that start-ups and VC have a much higher chance of survival and will be more profitable. Another is that people using pre-GFC data may come to the wrong conclusions going forward. So, IMO, the changes have been and continue to be big!
Ed. I'm fully aware that those who started investing after the GFC are asking what the hell he's talking about, lol. Which opens up another story for another time.
But back to OpenAI and the chances of a crash. Liquidity is key, and the changes made have taken that liquidity away from the banking sector. But global liquidity has increased enormously since the GFC. COVID-19 was another huge boost. Where has it gone if not expanding the banking system as it should have? And that’s the rub and why this time is different. Liquidity has moved from banks to investment institutions, which have different funding constraints, that are less reliant on deposits and interbank loans, regulations and ratios. They are reliant on equity, which has been funded across the now massive private and public investing entities. These are global and huge, and their funding is different and I would argue much more stable than the banks.
The upshot of this is less economic volatility, which is, of course, what we have seen since the GFC. This is no coincidence; this is an outcome of the changes wrought. I was interested to read the US Treasury Scott Bessent, saying the issue with US growth and inequality (para) is due to the lesser role the banking system is playing and the funding of grassroots businesses in the US. He is right, but changing it is the issue. I am focused on any administrative rhetoric or efforts in this regard because a large part of my portfolio is positioned for no change in a dead business cycle.
Ok, finally, to OpenAI, will they make it and underwrite the prosperity of the AI trade? Well, as I said earlier, the funding is crucial and it is in the hands of those who are not as reliant on the banks and credit as they were in the decades pre-GFC. That is a big positive in my view and increases their chances of success, and many other firms we have seen succeed since the GFC that would have failed at the inevitable long liquidity crunch, which now has disappeared.
So the dangers we should focus on here are changes to the liquidity landscape, restrictions on the Fed to flood the market at any sign of trouble, restrictions on huge SWFs or the massive Blackrocks, Blackstones, etc and any other colour that funds these businesses. Liquidity is abundant and has found a way to goal outside the banking system is the message here. At the moment, the players look fine, even strong.
The problem with the naysayers are living in the wrong world; it died post-GFC, and so far it is not coming back. Using arguments based on structures that are not relevant now are in themselves not relevant, imo. Sorry, value guys, no mean reversion here.
Some quite complex notions here, hopefully I have done a half-decent job illustrating what I think is going on. Could be wrong, but I don’t think so in the general direction in this case. I'm attempting to highlight what is now important.
Very on topic is Alan Kohler's weekend brief, it's to poignant not to share (sorry Alan). I think he will have to get some notes of @Solvetheriddle regarding the inflection from value to growth, post GFC and the possible influence of moving from bank lending to private financing for business investment. He also talks to Viktor, as @Clio flags and @Solvetheriddle rates highly. As always with Alan, some very interesting charts for those who like pictures (aka me):
PS: if your not subscribed to Intelligent Investor, it's worth it just for Alan's work!
What If This Time IS Different?
Value investors like all of us at Intelligent Investor tend to live by the idea that the most dangerous four words in the language are "this time is different". It's a philosophy based on reversion to the mean, that fundamental - and perhaps most importantly, historical - principles of value will always be restored in the end.
At a personal level, this has kept me (mostly) out of the market this year, which has produced an opportunity cost of 7% year to date (sharemarket total return of 10.5% YTD vs bank deposit interest of 3.5%).
But what is now haunting me, and I assume other adherents of value versus growth, is: what if this time really is different? What if artificial intelligence and the energy transition result in a permanent shift in the way the economy, and therefore investing, works?
As readers of this Weekend Briefing will know, I have been thinking about these themes for a while now, along with many others, of course. But what sharpened the question for me this week was a comment from our friend Viktor Shvets at a conference in Australia. I wasn't there, but Livewire covered speeches from a live event, including Viktor's in this piece.
Viktor said: "Probably the most shocking prediction in the next couple of years could be that we might be reaching 5% productivity with zero inflation. Now, that will be truly shocking."
He went on to suggest that most current jobs will disappear and a universal basic income will become necessary for the people "to compensate them for their irrelevance." While he thinks this is probably more like a decade away, if it does happen sooner, traditional inflation hedges like gold and silver will become obsolete and equity markets will get repriced dramatically higher.
Five per cent productivity growth and zero inflation would mean that this time is very different, and as Viktor says, share prices would be repriced dramatically higher.
He didn't express this idea in terms of value versus growth, or mean reversion. Still, he addressed this more directly in a recent note, in which he pointed out that "growth investing" has dominated "value investing" for the past two decades.
This is a very long time to be waiting for things to go back to normal.
As Viktor said: "This is as far removed from a mean-reversionary world of Buffett, Graham and Lynch as one could be. The twin danger of exuberance and narrow returns makes investors justifiably nervous."
Why is it happening? Well, according to Viktor it's a combination of the suppression of economic cycles by central banks and governments and disruptive technologies.
Eleven years ago I wrote in an article for the ABC that "recessions have been abolished". I was only half joking, after all it's just another way of expressing the well-established "Fed put" idea, albeit at an extreme - you know, that one that says the Fed will always do what it can to prevent a crash and thereby bail out over-extended investors.
Ever since the shock of the GFC, recessions have kind of been abolished, or at least the monetary and fiscal authorities are trying to make it so. They don't always succeed, of course - the pandemic got under their guard but the response was zero interest rates and massive deficit spending, because ... recessions have been abolished.
Viktor Shvets puts it a bit differently; he calls it "cycle suppression", but I think the idea is the same. He says: "Mean-reversion depends on economic and capital market cycles that drive investment from one sector to another in relatively predictable phases of exuberance and clearances. This was the world of the 1960s-00s.
"However, over the last two decades, suppression of economic and capital market cycles combined with emergence of highly disruptive technologies, have eliminated Buffett's moats while creating unprecedented excess capital and confusing economic backdrop of rapidly alternating cycles."
The other thing that got me thinking along these lines was this chart from Gerard Minack comparing PE ratios of the top AI firms over the past 7 years with those of the tech media telecoms (TMT) bubble of the late 1990s:
While the average PE ratio of the big TMT stocks increased from 15 to 44 times between 1995 and 2000, before crashing back to 20, the average PE of the AI stocks has remained roughly the same.
This has happened while their prices have done this:
So as investors we now have a binary choice: either AI does unleash huge productivity gains, especially for those companies selling it but also for those using it, or it doesn't.
Scenario one means the valuations are justified because profits grow sufficiently; scenario two implies a reversion to the valuation mean and value outperforms growth once again.
Which will it be? It's impossible to know.
Meta CEO Mark Zuckerberg addressed the question in a recent interview and basically concluded that he doesn't know, it could go either way.
"If we end up misspending a couple hundred billion dollars, that would be very unfortunate, but I would say the risk is even more on the other side. If you build too slowly, and superintelligence is possible in three years, but if you built it out assuming it would be there in five years, then you're out of position in what I think is going to be the most important technology".
Which explains why they're all building like crazy - they don't want to be left behind.
But that's what always happens in bubbles. It's fundamentally about FOMO, or fear of missing out.
I think the most likely scenario is a combination of both one and two - that is, the profits do eventually justify the spending, but it takes longer than Zuckerberg and the rest of them think, including the investors, and there is a 25% correction along the way.
When? That could take longer as well: as Mr Keynes said the markets can stay irrational longer than you can stay solvent. Or, as my 50 years in the markets tell me, wait at least a year from when people start calling it a bubble.
But there's a separate problem with scenario one, which is expressed in this "AI quandary" decision tree published by Louis Gave of GaveKal this week:
AI is not just, or even mainly, a financial thing; it has the potential to change the way human society functions completely.
What does happen to white collar workers everywhere? Will they lose their jobs? What if Viktor Shvets is right and most current jobs disappear and there will need to be a universal basic income to compensate them for their irrelevance? How does that even work?
The next box in Louis' decision tree is "much bigger budget deficits, rise of socialism/populism and higher corporate taxes and wealth taxes", none of which sounds very appealing, and you could write a book about each of those statements.
So this issue is a lot more than simply whether value investing or growth investing will win in the end.
Sure, tech might continue to lead as the last pink box says, which is nice for growth investors who took the bet, but that's not necessarily so great if you have to live in an underground bunker in a tax haven.
Also, I'm not sure the US does outperform, as Louis says.
Saturday 4th October 2025: Just to continue on with regards to the recent comments by Viktor Shvets (Head of Global Strategy at Macquarie) about "this time IS different" and there no longer being any true cycles so no mean to return to, therefore mean reversion is now a myth, in a week in which the US Government shut down due to an impass over the federal budget, this is how global markets reacted:
Those were the moves across the past week amongst the major global markets, we actually outperformed the US, which is not usual, but mining stocks, especially copper and gold stocks, were rising, and the US is tech heavy, not mining heavy like the ASX.
But the fact that the world can shrug off a US Government shutdown really plays into Viktor's assertion that you can have a crash on Thursday, it's fixed on Friday and Monday everything's back to normal. Or rather, you don't have the crash at all, because of the expectation that the market has guard rails in place now that won't allow it to fail, similarly to almost everything else - there's just an expectation that all such events - like inflation, government shutdowns, growth concerns, wars, terrorist attacks, etc. - are going to be temporary distractions and be fixed soon enough - so why stress about it in the first place?
Which makes me think that when we do have a correction, it's probably going to be a big one, because it has to be triggered by an event, or events, large enough to change the current mindsets that everything can and will be fixed, and quickly.
Even ETF flows are still rising - the passive money is still piling into the market. The only ETF in the list below that went backwards last week was the BEAR ETF.
Investors and the market in general are certainly still bullish.
We only need to look at Australia's second largest listed company, BHP, as an example of everybody shrugging off bad news. In the past couple of weeks, Albanese has been talking up a planned critical minerals "reserve" here in Australia which the Government would establish and/or control - the details are not clear - and there has been talk that Australia could use our critical minerals abundance as a bargaining chip with Trump to get reduced tariffs.
See Here: https://www.abc.net.au/news/2025-09-28/albanese-offer-trump-critical-minerals/105825654 [28 Sep 2025]
And here: https://www.abc.net.au/news/2025-09-24/trump-and-albanese-to-meet/105809460 [24 Sep 2025] Meeting scheduled at the White House on Monday October 20th.
One week ago, China expressed their displeasure at this by banning Chinese steelmakers from purchasing any iron ore from BHP, see here: https://www.afr.com/companies/mining/australia-urged-to-intervene-in-bhp-china-iron-ore-deadlock-20251002-p5mzhv [02 Oct 2025]
Excerpt: China’s state-run buyer of iron ore, China Mineral Resources Group, has told its steelmakers not to purchase iron ore from BHP, according to a Bloomberg report, escalating a stoush over pricing contracts which has cast doubt on trade worth more than $100 billion annually.
--- end of excerpt ---
Reports in that article and elsewhere suggest that BHP's exports of iron ore to China have continued without disruption despite the reported ban. Although it's not clear when the ban is supposed to start and if shipments that are already on their way over are exempt or included. It's hard to imagine China trying to slap us on the back of the hand and missing our hand completely and then not doing anything to rectify that. They've proven in the past that they're quite happy to use export bans and huge tariffs to "punish" us for doing anything that displeases them.
As for BHP, they haven't responded to questions about it and haven't made any statements. And what did their share price do over the past week? It finished basically flat; down just 14 cents @ $42.08 yesterday, from $42.22 the previous Friday, a drop of just one third of one percent (-0.33%), despite the high probability that this move by China has yet to play out. China is after all BHP's largest customer, by a BIG margin.
RIO, who are part owners of the huge Simandou iron ore project in Southeastern Guinea (in West Africa), which also has substantial Chinese ownership, and is due to come online this calendar year, so within the next three months, rose a modest +1.5% for the week. It is quite possible that China actually believe that with Simandou coming online and iron ore being available from RIO, FMG, Vale in Brazil, and other suppliers, that they don't need BHP's iron ore any longer.
People might say, well BHP is just one company. Why does it matter if they lose one customer?
BHP reported a total economic contribution to Australia of over A$50 billion in FY24, with iron ore being a significant component. In FY23, BHP's overall contribution to the Australian economy was A$60 billion, and it was the third-largest corporate taxpayer in the country. In terms of the commodity itself, in 2023, iron ore was Australia's largest export, valued at $139 billion, making it a major contributor to the economy. If China was to shut out one of our two largest iron ore players, it would have an impact on Australia for sure.
But I'm sure it'll all be fixed by Monday, right?
Disclosure: I do not hold BHP, RIO, FMG, MIN or any other iron ore companies at this point in time.