War, rate hikes and rising economic risk

[Music]

Hello, I'm Alan Kohler, Editor-at-Large of Intelligent Investor and also Finance Presenter and Columnist for the ABC.

And I'm James Thomson, Senior Chanticleer Columnist at The Australian Financial Review.

And we are The Money Café.  G'day, James.

G'day, Alan, how are you?

America's still at war in Iran, so what is this...?

Week five.

Okay.

Yeah, it's not good, is it, Alan?  Where's the off-ramp?

Well, the off-ramp is, I guess, humiliation for the US and for Trump.  He has to probably admit that it was a mistake or something.  How does he get out of it without doing that?  Unless he just keeps going.  Iran's not going to basically do a regime change or give up or capitulate, clearly, so unless Israel and US continue on and basically somehow manage to remove the entire Iran leadership, which I think is probably impossible, then they're basically stuffed, I would say.

Does that mean we're all stuffed?

I don't know what you think, we're a long way away, there's people who are more expert at this than we are, but it's hard to see how the Strait of Hormuz opens again unless there's an end to the conflict, obviously.  For as long as there's a conflict - and the trouble is, it's kind of broadened now, it's not just Israel persuading the US to bomb Iran and to get it to not have a nuclear weapon, it's involving the other Arab states now in the Gulf. 

I think there's a little chance - actually, not necessarily a little chance, but I wonder if the best way out of this is starting to become that the conflict actually continues in a maybe sort of less intense, more sporadic way and we get the Strait of Hormuz partially open or open enough where Iran basically - which seems to be what they're heading towards - starts carving out deals with various countries.  They've done a deal with Thailand, they've done a deal with China obviously, partial deals with India, there seems to be a deal with 20 tankers to Pakistan.  Is there a way that the strait's partially opened, there's enough oil getting through there that we can sort of minimise or reduce the supply shock and the energy shock.  I don't know, I wonder if something like that is potentially the halfway house. 

Yeah, Iran seems to want to turn it into a toll strait so it just charges people to go through it, which would be - at least stuff would go through it but it would put up costs everywhere and Iran would make a whole lot of money which it could spend on developing nuclear weapons and obviously the cost would be passed onto consumers and that would be more inflationary, so that's not ideal either. 

The thing that really interests me, Alan, I don't know if you've seen this view from Barrenjoey's Economist, Jo Masters, who sort of has made the point that part of the problem is we've got Australia's growth momentum over the next six months, is actually quite strong.  We've got the economy running up towards a speed limit at the moment, housing approvals are finally getting off the canvas and running at a good rate.  We've got inflation already coming through, particularly through wages.  We saw last week, strikes from the ABC journalist, we saw strikes from the teachers in Melbourne.  The national minimum wage case is probably going to end up with an increase to the minimum wage of something around 4.5 per cent to 5 per cent. 

So all that is before the oil shock and so I think that's really interesting that you've got this inflationary impulse already, then you've got the oil shock on top of that.  We're probably going to need two rate hikes here, or the alternative is we get such a big growth shock that we don't need the rate hikes, in which case we're worried about something else entirely.  It's not a very pleasant set of circumstances, is it?

No and I suppose that's the problem with Stagflation, if that's what we end up with, which is inflation and a growth shock of declining growth.  Because the Reserve Bank would have to then decide which of those two problems it goes for or tries to deal with and my betting would be it deals with inflation because that's what it always does.  

Alan, I had a listen to your new podcast the other day, 'That's Business'...

That's Business, yes.

You had Jim Chalmers as your first guest.  You used that word, 'stagflation', with him, he didn't exactly like it, did he?  He didn't love that definition because there's not that unemployment side of things.

No and he's right...

Is there distinction without a difference, or...?

Unemployment is relatively low and we certainly aren't seeing stagflation at the moment.  I did ask him, "If that's what happened, what would he go for?"  Obviously, politicians go for a stimulus, stimulating growth, rather than do the opposite.  He didn't answer that question or basically any question, did he, really?

No.  He's in a really tough spot.  It's five weeks ago since we were heading towards the May budget being this big sort of centrepiece in Labor's next productivity chapter.  Is that all dead now?

Well, it hasn't really been front of mind since the productivity summit last year.  That was August 2025 and it seems like a long time ago, but look, he's still talking it up, or whether he's going to do anything about productivity, I don't know what he can do anyway.  As I pointed out in the interview, the trouble is he's being compared with Keating and Hawke, but they brought about a revolution, the changes that they did to the economy were colossal.  A double productivity growth from 1.2 per cent in the 80s to 2.4 per cent in the 90s, now we're back to 1.2 or 1 per cent productivity growth, so basically we've gone backwards ever since then.  One of the things he said was, "Basically, you can't do those things again that Keating and Hawke did."  I suppose the problem is that most of the things that they can do are more or less incremental.  Obviously, there's a lot of talk about AI, maybe AI will lift productivity, but that might be at the expense of employment. 

Yes.  I was reading on the weekend, Alan, the war has interrupted the global trade in helium, which is important ingredient in chip making.  Then the other thing is, a big financier of this AI revolution has been the Gulf states, who have many other things on their mind other than artificial intelligence right at the moment.  There is a possibility that you get growth derailed, you get AI derailed, it's a pretty tricky little time, isn't it? 

Mind you, the Gulf states are also making more money at the moment, so they're going to have to do something with the money.

That's true, but if their production gets shut in for an extended period that doesn't help them too much.  It's just like, I know this is blindingly obvious, but just the change in the world in five weeks like going from a relatively benign pretty strong growth backdrop, the prediction for global earnings growth for this calendar year is still 18 per cent.  That's probably going to prove to be too high, but at the moment, it's still 18 per cent.  Now, all of that just seems to have been blown apart for a reason that none of us can really fathom, it's interesting.

That's right and nor can we predict or understand how it's going to end up and I suppose that's the real problem, isn't it?  Nobody's got any idea what the US is trying to achieve, they're just bombing the crap out of Iran and Tehran now with what in mind?

It's very difficult to try and wrap your head around it, that's for sure.

You had an interesting piece about private credit the other day, it was based on a conference I think you went to called the Asia-Pacific Financial and Innovation Symposium.

Yes, a grand name...

It was a grand name.  Give us a brief on what came out of that?

Everybody's really worried about private credit at the moment.  Private credit's obviously - it's actually not the biggest sector in the economy by any stretch of the imagination, it's worth about $1 trillion dollars US, that's nothing compared to the banking sector or anything like that.  But what we've seen over the last few years is really strong growth in private credit as global capital rules put in place since the GFC have basically incentivised banks to concentrate on mortgages and lend less money to businesses around the world.  That's opened the door for private credit to come in, they've done that, they've had a very successful sort of decade doing that.  But now as interest rates get higher, a few little pot holes in the global economy, including the AI revolution.  We're seeing private credit come under pressure, but we've sort of got this two-speed problem.  One, is what would really hurt private credit is if there were lots of defaults in the loans they've made, that's not happening, we should be really clear about that, that's not happening.  But what we are seeing, is private credit funds, which are largely illiquid vehicles so you can't move in and out of them on a daily basis, you've got to move in and out of them on a quarterly basis.  We're seeing increased redemptions from these private credit vehicles, to the point where the private credit players are saying, hang on, we're not going to let you out, a lot of these funds, 10 per cent of investors have tried to get out, they've gated or restricted the redemptions to about 5 per cent a quarter.  So we're seeing this big shift in sentiment in the private credit sector and last week at this symposium, which is basically run by the super funds, all the big private credit players, Apollo, Black Rock, Black Stone, Blue Owl, they were down there saying, "Hang on, hang on, everything's fine, don't worry about this!  Our loans are sound, this whole thing's overdone." 

But Macquarie's Chief Executive, Shemara Wikramanayake, she sort of summed it up really well saying, "Yes, okay, there's no problem with the credit part of private credit, there's no defaults, plenty of money to go around, that's fine.  But this sentiment shift around the redemptions, if that keeps going and people keep trying to get out and panic, then you will end up with a credit problem.  It was sort of an interesting juxtaposition of those two points.  "Everything's fine..." the big American private credit players say.  Then Shemara nailed it, there is a sentiment problem here and sentiment problems have an ugly habit of snowballing. 

I think what's interesting, Alan, it's a reminder of how formative the GFC is.  We're all looking in financial markets for that next - what's the next CDO?  What's the next sort of misunderstood part of financial markets that could make everything blow up and I think everyone's decided that's private credit, rightly or wrongly, so there's a big sentiment problem here.

From what you're saying, it's basically like all kinds of banking.  Private credit is another form of banking and the thing about banking, is they borrow short and lend long, so they're always vulnerable to a run and so what Shemara Wikramanayake is talking about is a run, isn't she?

That's true, but unlike a bank, because the way these vehicles are setup, the private credit funds, they can stop the run by limiting redemptions.  They can say, "No, no, no, this is in the fine print, we don't have to let you all out now.  It's not like a bank, you can't just all get your money out, you've got to play by the rules that we all agreed to upfront."  That does stop that sort of problem of having a run, Alan.  It is a bit different, but that nuance is largely being missed. 

Oh well, that's good, I'm relieved. 

We're all looking for the next bubble, the next big trouble spot and if this war wasn't going on, the noise about private credit would be even louder which is saying something.

Just finally, before we go to questions, I'm interviewing Anthony Miller, the CEO of Westpac, today for my next That's Business podcast for the ABC, what should I ask him?  What would you ask him?

I think you should ask him, are the economics of banking quite as good as the stock market suggests they are?  You've had Matt Comyn talking about this, the other bank chief executives less so, but there's an argument that bank share prices are still very strong, but there's no growth in banking, it's far less profitable than it was even five years ago.  The banks have shrunken down to these mortgage businesses, but most of their mortgages, 75 per cent of mortgages in Australia are sold by brokers.  Is banking still a good business?  I think it's a really interesting question that we don't think about.

It is, thank you very much.  Do you think he'll answer that?

Not sure.  I think he'll say something like, "Banks are still a cornerstone of the Australian economy...  Without us, the Australia doesn't go round," which is completely true.  Most banks pay something around 0.0001 per cent on their deposits, because most of their deposits are in transaction accounts.  What if there's a technological breakthrough like stablecoins that disrupts that?  What if open AI started promising to pay you 6 per cent on your deposits at call, by having a system where they had no branches, not even an office in Australia and they used a mix of stablecoins and other technologies to do that?  It just goes back to this idea that we've got this view of banking that it's rock-solid, impregnable...  I'm not saying the banks are in any danger or any risk, but it's not the business it was and yet, we sort of treat it like it is.  It's interesting.

It is, very good.  Now, before we move to this week's questions, here's a quick word from our sponsor.

[Recording] If you're enjoying The Money Café, take the next step with Intelligent Investors Essential subscription.  Right now, listeners can access a special Money Café discount, giving you Alan Kohler's Weekend Brief, the latest market updates and economic insights, plus exclusive CEO and Talking Finance subscriber podcasts.  In-depth interviews with leading financial minds on business, the economy and investing, all for just $297 dollars.  It's everything you need to stay informed.  Sign up today to get a limited edition Money Café cap, plus three months of Intelligent Investor Premium, it's our way of saying thanks for being part of the community.  Head to offer.intelligentinvestor.com.au/moneycafe to claim the offer before it ends, T'c and C's apply. [End recording]

First question's from Belinda, "I distinctly recall during the COVID stock market drop, you responding to a question at the time saying, "Don't be a bum!"" Did I say that?  It must have been you, James.

It sounds like something I'd say, yep.

"As in, don't sell at the bottom of the market drop and lock in your losses.  I manage our family SMSF, mostly invest in blue chip stocks, Macquarie, Cochlear, CSL, Woodside, Goodman, Xero, Ramsay and so on with a portion of NDQ..."  What's NDQ?

Not sure.

"...for international exposure."  It must be an ETF, I guess.  "Since October '25, my portfolio has lost 20 per cent, $110,000, due to a number of tech, healthcare and growth stocks dropping up to 30-40 per cent and while I haven't sold anything yet, I'm starting to feel the pain.  I decided to continue to dollar cost average through it and stay the course.  I'm 44 and my husband is 52, so we have time to ride the pain, but it does feel like it's going to get a lot worse before it turns.  All insights and advice welcome."  Before we answer, we should say, we aren't able to give personal advice, apart from not being licenced for that, we don't know anything about you, Belinda, so we aren't in a position to give you personal advice, so this is general advice only.  What do you reckon, James?

You're right, NDQ is the Nasdaq 100, which went into a correction on Friday night, sorry about that, Belinda.  We would never call you a bum, Belinda, or any of our fantastic listeners, but I think Belinda's got the sentiment right here, she's got a portfolio of stocks that she believes in for the long term and she's 44, she's investing for the long-term, well and truly investing for the long-term.  She's got 40, maybe 50 years of investing in front of her and you'll look back at this eventually and think this is a blip, as we always do.  I know there's all those old adages, Alan, it's about investing in the market, not trying to time the market.  I think this is just another example of one of those, something to warm the heart, Belinda.  Michael Hartnett, the Bank of America strategist, he had a list of all the corrections since World War II.  Within 12 months of a correction, which is generally when the market's down, peak to trough, between 10 and 20 per cent, the markets bounce back on average by 15 per cent over the next 12 months.  History's not necessarily a guarantee of what happens in the future, of course, but this is a time when you do have to ride out the bumps and it sounds like Belinda's got the perfect strategy to do that, doesn't it, Alan?

Dollar cost averaging through it, I think is the right strategy for sure, which means that you continue to buy and achieve a lower entry cost, obviously, if you're buying at a lower price, you're starting to average out your costs at lower prices.  I think that is a good strategy.  But, look, it might get worse before it gets better, I don't know - nobody knows, really.  But James is right, the market always kind of bounces back, that's for sure.

I must admit, I think this fall still feels underdone, we're sitting here on Monday morning and the ASX is only down 7.6 per cent, the S&P 500 is down 9 per cent.  For the size of the potential global growth shock that we're looking at, that feels very underdone to me, so I think it is likely to get worse before it gets better, just because it's so unpredictable.  I think towards the end of last week, we saw signs that investors are starting to capitulate, but most fund managers are still highly exposed to stocks, they haven't really taken a whole lot of risk off the table, so it definitely can get worse before it gets better.  Again, Belinda's 44, decades of investing in front of her, so this isn't the time to panic.

Rhys asks, "Alan, with super balances gyrating, I want to understand how we are taxed when balances drop so much.  First contact at my super fund didn't know so I asked ChatGPT, it explained how losses are carried forward within the fund and future tax liabilities reduced accordingly.  However, it went on to say, "If you withdraw or switch funds right after a strong market gain period when tax is being provisioned then markets fall later, you may not personally benefit from the latest tax offsets."  Could you explain this?  Like a lot of retirees, we are waiting until markets rebound before we top up our bucket allocated to cash from which we are drawing our living expenses.  Are there any hidden tax pitfalls we should be aware of?"

I think ChatGPT was right, that if you don't cash in your gains, losses later are obviously taken off the gains that you eventually get, so therefore the capital gain at the end, if you've suffered some losses, are less, obviously.  But if you cash in your gains, that doesn't help.  Is that what he's talking about?

I think so, yeah.  I'm not entirely sure there and because Rhys sounds like he's in the retirement phase and drawing down on part of his super, it might be a little bit more complicated.  I personally would be - if you're really worried about it, Rhys, I'd be talking to an accountant or someone else at your super fund to try and figure that out.

Yeah, ChatGPT probably isn't quite up to being an accountant yet.

Or going through the tax issues.

Maybe it will eventually, but not yet.

Yeah.

Rob says, "Enjoyed your discussion re MMT, modern monetary theory and inflation.  Yes, MMT is simply a better description of the economic system, a bit like Newtonian mechanics is a better description than Aristotelian mechanics.  The problem isn't printing money, it's lack of goods and services, too much money chasing too few goods and services gives us inflation and we've seen this repeatedly over the past six years..." and longer than that too, Rob.  "Solution is to reduce money supply.  Increased interest rates sucks money out of the system, so my question - what about the Feds increase taxes on people and businesses who have excess money?  PM Abbott did it..." 

Did he? Rob says, "So did Keating, as part of a comprehensive inflation fighting strategy which balanced increased indirect taxes with tax cuts."  Maybe he's mistaking them for John Howard who introduced the GST and offset that with tax cuts, I don't know.  Anyway, yeah, you could do that, increase taxes on people who have excess money, a wealth tax...

This is an interesting thing and in some ways, I say good on Rob, Alan, because often MMT proponents, you often hear them in times of crisis, saying, "Just keep printing, that's the way this works..."  I rarely hear them in times of inflation saying, "Jack up taxes, that's they way this is supposed to work."  And so Rob's actually calling for that, so good on him, because isn't that a cornerstone of MMT, that when there are periods of inflation, the answer is higher taxes?

That's right, because the MMT says that the point of the tax is to take money out of the system, not to fund the Government.  The Government funds itself by printing money and does that now and the reason the taxes are imposed, is to remove money from the system to prevent inflation. 

But do you ever hear some of those bigger proponents of MMT around the world - we've had a period, as Rob says, for six years of high inflation, I've never heard any of them out there campaigning for higher taxes, it just doesn't happen.

No, that's right, no.

That's what I find hard to get my head around with the MMT, the proponents like to take comfort in the good bit and they don't like to call for the bad bit.  Anyway, maybe I'm wrong.

Sounds like a politician.

[Laughs] Yes.  Mark says, "Alan, particularly I know that you have understood the fundamentals of Bitcoin for many years, many!  I found an article by you explaining Bitcoin from 2014, for gosh sakes, so please see commentary below..."  We'll skip that - "For what it's worth, I don't think that your quip about 'Ethereal, an incomprehensible' reflected any lack of understanding, simply reflecting and commenting upon understanding of everyday mum and dad investors out there in middle 'Straya."  Did you call Bitcoin, Ethereal, an incomprehensible, Alan?

I did, yeah.  As Mark observes, I don't find it incomprehensible, but I think a lot of people do and a lot of people also buy it even though they don't understand it, because they think it's going to go up, but it hasn't been going up for a while now. 

I confess, I'm struggling to come to grips with what role it plays.  It hasn't been a fantastic store of value, hasn't been a great inflation hedge.  I certainly get the logic of a currency outside the world of fiat currencies that supply is limited and controlled, but it's utility seems to be fairly limited and some of the other roles it's theoretically supposed to play, it doesn't really appear to play them very well.  I don't know.

I think it hasn't had good branding advice, if you know what I mean.  It was invented, not as a currency - I think one of the problems with the branding is that it's called a currency, a cryptocurrency, which makes people think that it's alternative money, but actually it's a platform for transferring money between people outside the banking system, which is what it was invented for.  It was never invented as a store of value, which is what people now seem to use it for.  It won't ever be money because Governments can't possibly allow it to be money.  They can't allow there to be money outside the system, so it can't be money and it can't be a store of value because it isn't a store of value, it isn't really an investment. 

People are using it as an investment, but it's not.  What it is, is a way to transfer money between people and it does that, that's true, but the problem is, it's anonymous and so therefore the big users of it are going to be people who want to be anonymous, which are not people necessarily that you want to be anonymous.  The whole blockchain itself, is a fantastic invention and it really is a great way of employing the internet and so on to create a sort of ledger that is permanent and storable and is outside the system, I think that's all fine, but it doesn't really - I think the whole bitcoin thing doesn't really work.  Bitcoin's not going away, it won't ever go to zero, it won't ever disappear, but it isn't going to be what a lot of people think it is, in my opinion. 

Ruben says, "Great podcast, Stephen remains a 'highly likeable loon'...'  Yes, we highly like him.  "You touched last week on the classic Australian dilemma, significant wealth tied up in property, but limited liquidity.  Asset rich, cash poor.  At the same time, there's growing noise around property tokenisation via blockchain, with some heavy hitters suggesting it may not be far off mainstream adoption, including Larry Fink of BlackRock."  Is that right, did he say that?

Well, he's big on tokenisation of the world, he's talking about that.

"If that thesis plays out, do you see the potential for a structural shift in the economy?  In a world where property owners can instantly sell fractional interests to unlock liquidity, it feels like the traditional constraints around capital could change quite materially, perhaps even blurring the lines between real assets and financial markets.  Also, on a separate but equally important note, when can we expect an introduction to Stephen's AI replacement?  Would that change his outlook on the AI impact on unemployment?"  Well, I'm sure it would, but I'm sure Stephen could not possibly be replaced by any AI. 

As for tokenisation of property, about 15 years ago I invested in a business that was interested in fractionalising residential real estate called DomaCom and it wasn't a success at all, nobody wanted to do it.  My wife said at the time, "This will never work, nobody wants to fractionalise real estate," and she was right.

There you go.

She's never forgiven me for putting some of our hard-earned cash into that company.

I think the idea is okay and tokenisation of various stuff can happen, is happening.  I just don't know if property's the one thing.  Again, you often say this Alan, we need to stop thinking of property as an investment, fractionalising it or tokenising it would increase the thinking of it as an investment, wouldn't it?

It would, exactly.  Making smaller lumps would make it easier to buy and the price would probably go up some more, we don't want that, so cut it out.

Matt says, "A quick question from what I expect is a pretty common situation..."  He and his partner in their early 30s, two kids under three and they did all the right things, good education, worked hard, built solid careers and they're now earnings decent incomes.  "But the numbers feel a bit crazy, it's going to cost us close to $100,000 dollars a year in daycare just to get my wife back to work, where a big chunk then goes straight to tax.  At the same time, a pretty average house in a decent part of Perth means taking out a $1.5 million dollar mortgage, even with savings and some help from family.  On top of that, I work for a US tech company, selling an AI product and can already see how much disruption is coming to jobs over the next few years.  So, zooming out, it feels like we're running pretty hard just to stand still - half joking, should we leave the country?  But seriously, how should someone like us be thinking about this?  Is this just the reality now, or are we missing something?"  There's a big question for you, Alan.

I know, that's a big question, we'll be here all day.  Matt, for a start, it's not just Australia that's got this problem, it's everywhere, so I don't know where you would go to get away from it.

Alan, I sort of think of our old mate, Viktor Shvets, the Macquarie strategist, at moments like this.  I think Matt's summing up the feeling for a lot of people, that we're between two moments here.  We're between this old world of globalisation, financialization, where we have pushed up asset prices to really difficult levels for lots of people which is leading to inequalities and popularisation and populists and all of a sudden, you find yourself in the middle of an energy shock caused by a war that no one really understands.  And yet, to the second part of Matt's questions, we're moving towards something where we do have this AI change that's going to sweep through the economy. 

That's not a helpful - I'm diagnosing the problem rather than providing a solution, but I don't think there is a solution.  We've sort of got to ride through this and hope that the new regime that comes out of it is one where the world can find a path through and maybe that is something where the way we work and live is vastly different, the way society taxes and how it values work might be very different.  I don't know, I think this is the big problem that I keep coming back to, it's not easy to solve though and Alan's right, moving to another country is not going to help you. 

I don't know what the easy solution is, but even now, you can see with this energy shock, there'd be some who argue - I'm not one of them - but there'd be some who argue that a period of creative disruption, a clean-out of excesses in the economy would help reset things.  A big recession, something like that.  But imagine the damage it's going to do on the way through.  This is the problem we've got, the solution for all this is very painful.

I think your answer to that is very big picture and cerebral, which was great, but I read Matt's question last night as I was preparing for this and I sat down and bashed out a script for a piece for the ABC on Sunday, which started this way and I'd like to get your feedback to it, James.  I have an idea for next month's federal budget - I should say, I hear a lot about now the combination of the problem of families like Matt's, of housing affordability and childcare.  These two things are really causing a problem for people and so my idea is, nationalise childcare and kindergarten and make them free. 

That is, extend public education to kids under five.  Before you go, "Oh, Christ, nationalisation..."  The problem is that the whole childcare, kindergarten industry is 70 per cent for-profit, it's entirely profit, 30 per cent is not-for-profit, 70 per cent is for-profit.  And so, the problem is the moment the childcare subsidy is up to 90 per cent for low income families and then it tapers really quickly if you're earning money, so if both of the partners are on $150,000 or something, you're really struggling because you've got a big mortgage and you're paying a lot of childcare that's not getting much subsidy at all.  When you think about it, actually the whole childcare industry is probably worth about $50 billion dollars, the total revenue of the sector is $23 billion, so if you remove profits, it would cost the Government $20 billion to run it, but the subsidies cost $16 billion at the moment.  If they issued childcare bonds, worth $50 billion dollars, to super funds, which would be a snack for the super funds - there's $4 trillion dollars in super, so it's not that much - basically, you'd issue the bonds to buy the childcare centres and the interest on the bonds would be a total of about $3 billion dollars, so there'd be a lot left over from the subsidy to help pay for the cost of running the - that would pay for about half the cost of the Government running all the childcare centres. 

The rest of it, call it an investment in gender equality and productivity.  I just think that actually would help deal with housing affordability as well.  I think it's really anomalous, that as soon as a child turns five, their education is free or basically free as long as you go to a state school, but up to five, you're forking out hundreds of dollars a week, it's terrible.

Does that plan rely on housing supply continuing to increase though?  Because otherwise, you're just going to have more income chasing house prices, aren't you?

Well, I suppose that's right, isn't it?  The cost of childcare is helping to keep house prices down because everyone's struggling, that is true.  Trust you, James, to come up with a flaw in my plan.

Sorry!  [Laughs]  But maybe that's not a flaw, the argument is that you can't take your foot off the housing supply pedal and over time, we can get down towards a bit more of an equilibrium. 

I do think this business of having a private childcare and kindergarten sector has got hairs on it...

Your numbers are really compelling there, we are paying for all the profits anyway...

They're having to subsidise all this money, but the trouble is, they're subsidising shareholders of these private childcare centres.  Stop subsidising these private centres and just buy them!  Take them over and make it a public sector!

There you go, Matt, you've got the practical and the cerebral answer, that's a terrific question. 

Do we have to end there?  We've probably got time for one more question, do you think?

Let's go to Johnathan's at the bottom there, Alan.

Okay, is it your turn or my turn?

I think it's your turn.

Johnathan says, "I'm a hospital doctor in Australia, experimenting with open weight LLMs..."  What's an open weight LLM?  Does he mean open source?

I'm not sure.  No, I think there is open weight...  I'll come back to you on that.

"Here's a prediction I'd love your take on.  As energy constraints push cloud AI costs up, I think we'll see faster adoption of on-device models and Apple's unified memory architecture is the hardware inflection point that makes this practical.  Frontier labs won't slow down but they'll pass rising and energy and infrastructure costs through to API pricing, making local inference comparatively cheaper over time..."  This bloke's a geek!

Yeah, it's good.

He's in the weeds, this bloke!  "In healthcare, the case is already strong privacy latency cost, but I think energy economics might be what tips the balance across industries.  In short, the energy crisis will accelerate, not slow down AI, but just change how it's done."  Interesting point and question, I guess.

Yes, so an open weight model is slightly different to an open source model, Alan.  They're very similar, but the open weight models keep their own code and training data inside them and don't publicly release it, so a twist on open source models is probably the simple way or the dummies guide and Johnathan will forgive me if I've slightly got that wrong.  This is a really interesting point and something that I'm coming across a little bit more inside corporate Australia, there are discussions now about, yes, there's some really exciting things we can do, but the cost of running the models to do it is actually almost prohibitive. 

Now, that's a moment in time thing, but I think it goes to what Johnathan is saying, that there is going to be different ways that we use AI as energy costs and some technology costs adjust as this sort of revolution rolls through.  I think Johnathan's probably right, that there will be changes in what different styles of AI can do and how it's done and I think he's also probably right in that the AI ecosystem will adapt and will see different rates of progressions in different industries, which is fine.  That's the way these things usually work.  Not everybody's going to need the top AI model, we hope the top AI models get to the hospital doctors of course, but it's not going to be required everywhere.  This is the Corolla story, not everybody needs a Ferrari, for most people a Toyota Corolla is absolutely fine and it will be the same with AI.

I looked at the cost of AI, I asked one of the AIs, "Give me a list of all of the AI token costs per million tokens..." It ranged from 4 cents per million tokens, to $600 dollars per million tokens, so there's already a huge range in cost.  I downloaded it into Excel and averaged it and the average was $25 bucks, which I talked about on the news last night.  The other thing is, I think Johnathan's right in the sense that it's already happening because one of the more popular AI things going on at the moment is OpenClaw, which stays on your system, stays on your computer and it basically can operate your computer for you, if you want. 

I'm not sure whether that's what he's talking about, that kind of thing, but it's certainly true that energy costs are going to be an increasing issue with it.   The other thing is, I'm not sure whether it is a prohibitive cost if you are using AI to replace human beings.  If you're using AI to simply enhance everyone's work, increase productivity, then the cost of it really does matter, but the cost of it, whatever it is, even if it's $600 bucks per million tokens, it's less than the human costs.  I think that almost no matter what you're paying for AI, it's less than a minimum wage or average wage.

In some white collar sectors, you're not paying a minimum or an average, are you?

That's right.  Even so, I do think it depends a bit on what you're using the AI for.

Indeed, yep.

Well, that's it, thanks James, great to talk to you, as always.  Thanks, everyone, for listening to today's episode of The Money Café, it will be with Stephen Mayne next week, so send in your question and we'll answer it together by emailing themoneycafe@intelligentinvestor.com.au, until then, I'm Alan Kohler, Editor-at-Large of Intelligent Investor and Finance Presenter and Columnist at the ABC. 

And I'm James Thomson, Senior Chanticleer Columnist at The Australian Financial Review.  Have a great Easter break.

Thanks for listening to today's episode of The Money Café.  If you'd like more, join Intelligent Investor's Essentials subscription.  Visit offer.intelligentinvestor.com.au/moneycafe and you'll get my Weekend Brief, Talking Finance and CEO interviews.  Sign up now for just $297 dollars and we'll give you a free limited edition Money Café cap, plus Intelligent Investor Premium free for three months.  Get in quick at offer.intelligentinvestor.com.au/moneycafe.  T'c and C's apply.

[Music]

END OF TRANSCRIPT