Capital Flows and Asset Markets

Back in 2011, I started writing and distributing free research. The main aim was to make myself known to the investing community, but it was always enjoyable and interesting to see what comments would come back. And my last post, generated an above average number of comments. Generally speaking, when I write, I write with a point in mind. This post is more to comment on the comments.

First of all, in my post I found it interesting that Zerohedge was generally bullish. Zerohedge replied directly to me (yes Tyler Durden - and as far as I know, the real Tyler Durden, not one I imagine I am talking to from time to time) with:

Zerohedge has been “bullish” since March 2020 when the Fed started buying junk bond ETFs at which point it became clear a drop will never again be allowed. In retrospect, it has been the right call.

I missed this bullish turn in Zerohedge (March 2020 was a busy time) - but they have been correct, its been the right call. What frustrated me back then, was that there was seemingly no “market punishment” for the Fed basically committing the US government balance sheet to the market. I would have assumed inflation and dollar weakness would come from such a decision. In reality, it has happened, but with a delay. The move in gold recently has been enormous, and for me directly related the actions at the Fed. As Scott Bessent has pointed out, when you buy corporate bonds, you are starting to make political decisions.

Normally a surging gold price is not good for US equities. But what has really supported the market has been a weak oil market, which has kept inflation at bay. 2022, which was a poor year for equities was the last time oil was above US 100 a barrel, and the current oil price of USD 65 was first seen over 20 years ago.

Which leads me on to two more comments from subscribers. One was looking at the XLK/XLE (tech ETF v energy ETF) and picking up a potential turn. Rather than use these ETFs, I have used the S&P 500 sector returns to generate a much longer term chart.

Comparing oil to gold over a very long time horizon, there is no question oil looks cheap.

If there was one chart that explained the US exceptionalism, and the performance of tech versus energy, it would be spot price of Permian natural gas. It spiked in the weak years for US equities of 2000, 2008 and 2022, but as of today is pricing NEGATIVE again. A second subscriber was sure that the Permian was heading into trouble. Low prices would do that - but was there any sign of change? Like falling production, or pricing discipline?

Usually low prices drives supply destruction. There has been some signs of that, with US producers running down inventory of wells. They changed how they report this number in 2024, but the trend remains the same.

US shale producers have been able to increase productivity per well substantially. Mainly by buying bigger bits of land, and then extending the length of the horizontal drilling. The dark green line in below chart.

A Canadian subscriber (and they tend to know about oil) was saying that the Permian was a mess, and no less than the Diamondback CEO was bearish on US oil production. Diamondback is one of the best shale producers, so I was intrigued by this. And it is true that outgoing CEO, Travis Stice has commented that as current prices, shale does not make sense, and oil production will decline. But for me, this is not enough to get super bullish oil. It just says the oil is low, and short term producers will pull back from investing. However, in my reading on shale, I did find something interesting.

The US shale industry is really centred on the Permian region these days. From June 1 2025, the Texas Railroad Commission (RRC) will be changing regulations on wastewater. The Chat GPT summary of the changes are below:

Basically, for every barrel of oil, you create four barrels of wastewater, and the RRC is cracking down on this. Why? There have been instances of the wastewater leaking into freshwater reservoirs and causing old oil wells to rupture. This should put some cost pressure on the Permian shale producers. I have mixed feelings about this. Part of me wants to be bullish about oil, as I am used to gold and oil moving together. But part of me is starting to wonder if oil is a dead commodity. For me a dead commodity is when technical developments render a commodity worthless. We have seen that happen recently with diamonds as lab grown diamonds become indistinguishable from the real ones.

Oil and natural gas were vital commodities as they drove inflation. But we can see that electricity price inflation in the US is not following natural gas and oil prices. Maybe we just need to ignore oil prices.

There is a small problem with this. If I look at gold versus US electricity CPI, it also looks very overvalued. This makes sense if you think US shale pricing has kept inflation in the US in check.

So in conclusion - I was wrong that Zerohedge had suddenly become bullish - my bad. I think the market has done well more from low energy prices that Fed buying. And there could possible be a turn in energy against tech, or maybe not. There is a possibility that oil is a “dead commodity”. But if electricity prices had detached from energy prices, then an inflation shock with out an energy shock is now a possibility.

This is a public episode. If you'd like to discuss this with other subscribers or get access to bonus episodes, visit www.russell-clark.com/subscribe

This is a free preview of a paid episode. To hear more, visit www.russell-clark.com

I caught up with Rupert Mitchell of Blind Squirrel fame and we were chatting about markets. I pointed out that my friend John Hempton (@John_Hempton on X) was having a tough time. John has a different, but very effective, style of shorting selling. In simple terms, he tracks “bad” people as they move from company to company. Bad in this case means they act in ways that are “bad” for share holder value- he then shorts the absolute worse using a proprietary system. Its a great model - but I knew it would be a problem when liquidity floods the system. 2021 for example. And apparently this year too. But shorting when “bad” companies have gone up is good… so to speak.

Rupert mentioned to me that Zerohedge, which was the “most bearish website on the planet” was now also bullish. This was news to me. But a quick look tended to confirm this observation. I also noticed that the ARK Innovation fund was at 4 year highs - a bull market fund if there every was one.

Goldman was also telling me that CTAs - the true definition of dumb money - were back to max long equities.

And credit spreads were back at all time lows.

And then we had this interesting piece of research from GQG (I don’t know them), highlighting a bearish view on US tech. The key to its bearish view was that top line growth in slowing, while Capex is now increasing for tech companies.

And as I have pointed out a few times, gold outperforming S&P 500 tends to mark turning points.

So far, so bearish. We just needed one more trigger I think.

Long time readers will know that I have very divergent views on fertility going forward. To make a long story short, I saw the current drop in fertility rates being driven by a 40 year cycle that favoured capital over workers - which in essence made it too expensive to have children, which I now see changing. And secondly I saw improvement in technology allowing women to have children later driving an inflection in fertility rates. As the Economist makes clear, the trends are still decisively moving to lower fertility rates.

In 2000, you could argue that Indian fertility could offset falling fertility in China and Japan. No longer.

The Economist then goes on to demolish UN population projections by showing how it assumes countries with collapsing fertility rates are suddenly forecast to rise again. Demographers were pilloried for their Malthusian forecasts in the 1970s, are trying to make amends by not baking in collapsing fertility rates. I realise I am in some ways trying to do the same thing.

The reason why demographers, and myself, have a hard time accepting the collapse in birth rates is that is would imply nations like Japan will collapse to population of 60 million from 125 million peak. This seems absurd, so you naturally want to correct for it. The other thing about collapsing fertility rates, is that they generally only occur in nature when there is overpopulation, and potentially lack of food, neither of which seems to explain the collapse in fertility. But perhaps something else is going on. While the human population has surged from 2.5 billion in 1950 to over 8 billion today, we have also seen life expectancy increase substantially from 50 to over 70 today. If you assume work life begins at 20, the average human has also doubled the time they can be gainfully employed.

I started to wonder if human populations just naturally regulate the amount of labour available. Lets look at Japan. Its population peaked out nearly 20 years ago. But Japanese life expectancy was also rising until recently. But what we want to work out is how many years of expected life was ahead of the average Japanese person. We need life expectancy as well then.

Perhaps we should think about the sum of future years of the Japanese people? That would be the sum of all years expected life of Japanese people ((life expectancy - median age)*population). One thing I like about this way of thinking is that if we are all going to live forever, then fertility rates mean nothing anyway or should be close to zero. Doing this, we found Peak Japanese Life was early 1990s, and when Japan was its most expensive.

There are some issues with this. Firstly, we need to really only look at years over the age of 20, as this is when humans start to work, and then adjust for some decline as people get to 70 or 75. Less young people, and more older people should see “working” year remaining probably still be rising in Japan until a few years ago. This matches up with Japanese wages, which have started to rise again in the last few years.

If rising life expectancy was rapidly increasing available labour, it would also lead to rising education costs, as you have a longer work life to amortise education costs. In Japan, education costs rose rapidly until the late 1990s in Japan and have stagnated since.

In pure economics terms, I am arguing that a population boom AND an increase in life expectancy seems to create and excess of labour. Certainly you can see how the longer working lives has put pressure on younger people entering the workforce. And with a longer pay off period, the costs of education rose as people could afford to invest more. The implication is that higher the life expectancy, then the lower the fertility rate. There does seem a passing correlation.

In some ways, a longer life expectancy should imply lower fertility rates. At an extreme, if human immortality was achieved, then fertility rates would likely fall to close to zero. It makes the question of fertility really a question of population health. This sort of makes sense to me - but I still feel uncomfortable about it. I need to think about it more.

This is a public episode. If you'd like to discuss this with other subscribers or get access to bonus episodes, visit www.russell-clark.com/subscribe

This is a free preview of a paid episode. To hear more, visit www.russell-clark.com

I am a generalist investor, but once I get interested in an area I try to get to the bottom of it. My interest in clearinghouse started with this BIS article on the collapse of a small Nordic power clearinghouse. It was further stoked by an a report issued by the biggest users of clearinghouses asking for significant reform. As I was looking for a short selling trade on this, I tended to focus on listed clearinghouses. By far the most important clearinghouse is LCH, which does repo for European bonds, and interest rate derivatives globally. My biggest problem with clearinghouses was that they priced risk, but when things really went wrong, they suffered relatively minor consequences - a critique that is echoed in the report form the large banks.

In my last post, I had stated that CME cleared US treasury Repos, just as LCH clear gilts and European bond repos. This incorrect. US Treasury Repos are cleared by DTCC. DTCC is not listed, and remains owned by its members, which will be mainly large banks. As it was unlisted, I never analysed its IOSCO-CPMI disclosures. Having now had a look, the clearinghouse issue looks much clearer to me.

This is a free preview of a paid episode. To hear more, visit www.russell-clark.com

In case you missed it, Oracle, reported numbers last week. While earnings was in line, it surprised the market with a huge pipeline of new AI related cloud bookings. An already large stock gapping 36% higher is extremely rare - and this was on top of a huge run up from the April lows.

As the FT points out, Oracle has been able to out compete other cloud rivals to build bigger bookings pipeline in its cloud business.

As the FT reports, this huge growth in pipeline is driven mainly by a huge deal with OpenAI. The scale of growth in mindboggling. They expect revenue from the cloud business to USD 144 bn in four years time. Oracle sales were stuck at USD 40bn for a decade. It essentially saying its new business will create nearly 4 times the business in 4 years time.

While the FT article suggests that Oracle was able to undercut some of the bigger established cloud operators, what seems more likely is that demand is so strong, the large cloud businesses are reaching capacity constraints. Google, Microsoft, Meta all have their own AI systems that they are trying to compete with, while Oracle seems exclusively focused on cloud, making them an ideal partner for OpenAI. Oracle seems to be implying that OpenAI is going to be giving them around USD 100bn of business a year in 4 years time. For me, the only comparison I can think is when Chinese demand started to take off in 2000, the market simply could not get its head around the scale of demand.

My problem is that this sounds bullish, and is bullish. But how do I match up such a bullish announcement with the continued outperformance of gold relative to the S&P 500? It feels like a US centred tech boom should see S&P 500 outperform gold, not the other way round.

Big gold turning points against the S&P 500 have historically been times to show caution. Oracle numbers suggest the opposite.

Perhaps gold is just responding to a weakening dollar. The turning point in markets in 1970 and 2000 also marked points when the dollar weakened then too.

There is a possibly bullish read on this.

In a recent post, I wondered if the Trump administration was going to force the Federal Reserve out of the “QE Business” - or in other words ban them from buying assets, and get them to focus solely on interest rate markets. This could be a huge issue for markets. Whenever asset markets have threatened to unwind, central banks have stepped in as buyers. The UK gilt market is good example, where the Truss budget was a shock, that caused initial margins on gilt trading to rise which caused the market to have no bidders and yields to surge. It required the BOE to restart QE to bail out basis traders.

When looking at basis trades (typically buying physical treasuries and then selling interest rate futures or derivatives) there are two different market to look at. LCH owned by the LSE Group and has dominant market share in interest rates derivatives and European bonds. CME dominates Treasury trading, and JSCC dominates JGB trading. We are mainly concerned with treasuries here, so we will look at date from LCH and CME. Looking at variation and initial margin data, there is a cyclicality to this. Spikes in variation and initial margins SHOULD be correlated. As of the most recent data point, q2 2025, margins look low at LCH.

CME data is more interesting. Q2 data shows much more of a spike in both variation and initial margins, more inline with the air pocket we saw in Treasuries during the “Liberation Day” tariff sell off.

More surprising was the lack of margin breaches at LCH

Something we also saw at CME.

Having studied clearinghouses for a while, this would match up with market moves. The MOVE Index (a measure of bond volatility) tends to match initial margins, and spikes from low levels to high levels tend to march up with margin breaches. So 2020, and 2022 saw margin breaches, but since 2023, the MOVE index has tended to be moving lower.

Generally speaking, clearinghouses tend to cause trend following to work well. So when things are going well, for example the MOVE Index is going lower, then initial margins will fall, and asset markets will behave. It is only a problem when volatility inflects higher. So generally speaking, IF the Fed gets out of the bond buying business, AND the Trump administration delivers inflationary fiscal policy or we get an oil shock, then we will have a problem. But the data above suggests that will be more likely a 2026 or 2027 problem. The only data point that I found particularly bearish was a huge drop in KCCP - which determines the capital charge for trading with LCH.

This fall in capital charge is very good for basis traders and other users of the clearinghouse, as it allows less capital to be put aside. The fall was driven by a regulatory change, which in essence allows the clearinghouse to net trades. But netting is the key issue with clearinghouses. Netting is the fundamental driver of basis trade, and leverage. LCH reports initial margin numbers on a total and net basis. As I understand it, the net number is purely hypothetical. Is the amount of initial margin LCH calculates would be necessary if all trades could be netted.

Theoretically this should be fine. But analysis suggests that we are ending up with a bifurcated market, with hedge funds all one side, and banks all the other side. As an old report form the BIS noted, all US banks have become a net lender to the repo market.

Putting it all together, conditions are nearly perfect for a clearinghouse blow up - but not yet. A sustained period of the MOVE index trading in the 50 to 60 range should allow sufficient leverage to build for a blowout move. I still think long dated government bond yields go higher. But an air pocket move like we saw in gilts in 2022, or treasuries earlier this year looks unlikely until 2026.

This is a public episode. If you'd like to discuss this with other subscribers or get access to bonus episodes, visit www.russell-clark.com/subscribe

This is a free preview of a paid episode. To hear more, visit www.russell-clark.com

A few year ago, I could see the political winds were moving to getting real wages up again. This could be achieved either by raising wages, or lowering asset prices (the political aim was to get house price to income ratios down). This line of thinking lead me to a long gold (GLD)/ short Treasury (TLT) trade. It has been a good trade.

However the TLT side of the trade has not broken to a new low since 2023.

Is the GLD/TLT trade done? Or is the TLT side just a waste of time?

This is a free preview of a paid episode. To hear more, visit www.russell-clark.com

If you have never read or heard me talk about the clearinghouse trade before, then there is a tab on the home page, where most of my research is housed.

The theory of the clearinghouse trade was that the regulatory driven move from an investment bank model, where there were numerous bilateral agreements to a clearinghouse model, while simplifying the financial system, had two big flaws. First, the clearinghouse cannot go bust, and prices risk on a backward basis. If yesterday was good, then today is also likely to be good. Investment banks were far more forward looking in pricing risk. And secondly, profit maximising nature of clearinghouses means they will price risk at very low levels, before a crisis forces a rapid repricing, with causes prices to fall further.

The “practice” of clearinghouse theory has broadly been correct. A good example is the gilt blow up of 2022, when the Truss budget shock caused margins to be raised, and the gilt market went no-bid. A similar thing happened in 2020 during the Covid crisis. The problem from a hedge fund perspective, is that every time the clearinghouse model has blown up, it has been bailed out by central bank buying. Fed buying in 2020, and in the gilt market in 2022 by BOE buying bonds, even as they were raising interest rates. So the problem with the clearinghouse model as a short trade is that the politics has been supportive of “pre-emptive” bail outs.

Reading Scott Bessent’s recent interview on the Federal Reserve made me wonder if the politics may be moving “anti-bailout”. So I had a look at the share prices of the main clearinghouses, and they were surprisingly weak. Time for the clearinghouse trade?

I had viewed the political machinations around the Federal Reserve as a transparent attempt of the Trump administration to control interest rates and push them lower. The market has agreed, with 2 year Treasury yields touching 3 year lows.

The pressure on the Fed has also led the price of gold to surge, as investors question the safe haven nature of the US dollar.

My preferred inflation measure, the 30 year yield to the 10 year yield, has also surged to new highs. That is market seeing yields lower in the short term, but needing to be higher in the long term.

So far so simple. The Trump administration puts a friendly team in at the Federal Reserve, the dollar tanks, the stock market surges and everyone is happy. And that would be fine by me. But I just read an interview of Scott Bessent in The International Economy, and he had the following criticisms of the Fed policy in recent years.

And then this:

It is particularly noticeable that since Covid, (and really from 2016 onwards) the spread on corporate debt to treasuries has become low and stayed low (COVID aside).

The market has learnt that the Federal Reserve will always be the buyer of last resort. The Fed happily expands its balance sheet at the first sign of trouble. But what Bessent is saying is that this is a policy that Fed should not be doing. To be fair the Fed, it has run down its balance sheet in recent years, but I don’t think markets have priced in any removal of a backstop.

As a percentage of GDP the balance sheet is down from a high of 36% to 22% today. But prior to 2008 it was only 7% of GDP. With US GDP at USD 30 trillion, 7% would imply a USD 2.1 trillion balance sheet, or a USD 4.5 trillion reduction.

Is this just talk, or is it Trump administration policy? Would it really allow credit markets to operate freely again? If true, it would be extremely bad news for leveraged bond trades - or basis trades. After dipping in 2022, basis traders have come back in vengeance.

The basis trade has expanded massively since the GFC for two reasons. QE and clearinghouse reform. Clearinghouse reform has meant that hedge funds can get all the leverage they want from clearinghouse, as the clearinghouse can not go bust, and the lenders have no transparency on who is borrowing. This should be a recipe for disaster, and we have seen moments when this looked like it would be a disaster, but central banks have stepped in to take the other side and bail out hedge funds. If Bessent is putting political pressure on the Fed to reduce its balance sheet, then the risk of malfunction in the treasury market is rising.

I think the base case for most investors would be the the Trump administration and the Fed would bail out investors, but politics is a funny game. And not all asset owners are the same. Hedge funds and private equity guys being zeroed would not cause much of a political problem I think. Let’s see.

This is a public episode. If you'd like to discuss this with other subscribers or get access to bonus episodes, visit www.russell-clark.com/subscribe

The Gold/S&P 500 ratio got back to the 200MDA, and then spiked again.

Which makes a larger turn in the market more likely, implying either a bigger fall in the S&P 500 or a larger spike in gold.

And GLD/TLT finally broke out of its range.

Credit spreads have widened a tad - from very low levels.

8 month VIX continues to trade richly - something I have taken to imply that investors are keen to hedge risk.

Spot VIX still looks too low compared to UX8.

With tame commodity prices, my best guess is that the movements in equities, bonds and UX8 are markets positioning themselves for some sort of government debt crisis. The widening in French bond yields to German bond yield remains unresolved.

And the back up in long dated JGB yields suggest a market that wants higher yields.

For now, long dated US treasuries remain well behaved. But a decisive move about 5% beckons in my view.

Markets seem to be saying that inflation is building, and the current enthusiasm for interest rates cuts seem misplaced. I would agree.

This is a public episode. If you'd like to discuss this with other subscribers or get access to bonus episodes, visit www.russell-clark.com/subscribe