Snippets are works in progress and so ask questions rather than attempt to provide answers (yet).

On commodity volatility - 2 February 2026

¹Source:ICE BofA via St Louis Fred

Natural Gas

Natural gas had its largest 1-day drop in 30 years, on easing of tensions between the US and Iran, and warmer weather forecast in February. And as we predicted over the weekend, a provider of leveraged natural gas funds experienced a restrike event. When was the last time the same ETP provider had a restrike event on consecutive trading days for different underlying commodities? We’ll be keeping a diary of those. An exploding commodity price volatility regime is only good for market makers and commodity traders, and certainly not for end users or buyers trying to plan for the year ahead. Options prices spike during volatile periods and so those trying to hedge themselves will spend a fortune on premiums.

Eggs

We recently talked about egg prices in the US likely hitting a bottom, and on that basis looked into Vital Farms and Cal-Maine Foods. Well either there’s a glitch on the website or eggs in the US just dropped 95% to $0.06 per dozen. Most recent USDA data is still pointing at $1.05 and perhaps it has to do with the US government shutdown. Otherwise, something fishy is going on with eggs.

Credit Spreads

Finally, we saw that the ICE BofA High Yield Option Adjusted Credit Spread spiked again last week. In the past significant spikes have been accompanied with market volatility. The 10-day change is the highest since November when the NASDAQ 100 saw a slight correction. But this spike could have been Friday’s metals panic and might not have broader market implications. The rest of this week might give us a clearer signal but on the whole we’re positioned for some downside here.

---

The Return of FX¹ Volatility - 29 January 2026

¹Stock data from Yahoo! Finance

Around 15:00 GMT the DXY US dollar index was around 95.88 when it started to appreciate sharply. By 15:30 it was trading at the high of the day at 96.40, up barely 0.5%. But other assets reacted violently over this same time window, potentially because of this move. Silver fell nearly 11% from 120 to 107 over this same time frame, and gold nearly 7% from 5500 to 5100. There were a few pieces of economic data that came out around then that might have explained it. US factory orders were higher than expected, growing 2.7% instead of a forecast 1.6%, potentially signalling a more robust economy.

We’re inclined to say that the cause of the US dollar move was less important than the effect - we believe that move rattled the precious metals trade. And given that gold and silver collectively comprise $44 trillion in "market cap" at the moment, one should not underestimate the extraordinary asset price deflation that a surge in the US dollar could cause. Even if a directional surge is not the base-case here, we are increasingly confident of a pickup in currency volatility more broadly.

Several US and EUR currency pairs have started to show some increased volatility over the last few days, with the 10-day rolling average of intraday high-minus-low for some of these pairs hitting the highest since the April tariff turmoil. Rhetoric around foreign currencies has picked up significantly in the last week. This includes President Trump and Treasury Secretary Bessent’s comments about the dollar recently, comments around the South Korean won being too weak and the Chinese yuan being too weak. The semi-annual US currency report was released today as well and while it did not name any specific currency manipulators, the report did state that the US treasury would strengthen its monitoring criteria of other countries.

Giving a strong directional view on the US dollar remains difficult at the moment - keeping it low or lowering it further seem to have significant benefits for the Trump administration at the moment. And even today its brief spike higher resolved back down to the 4-year low just below 96. But with the new Federal Reserve Chair in play, increased US currency rhetoric, Japan’s fiscal and monetary policy, and worries about currency debasement coming up, we expect to see this recent surge in currency volatility to continue.

And currency volatility has plenty of downstream effects, few of them good.

Updates

We noted yesterday that one of the most liquid leveraged semiconductor ETFs (SOXL) was approaching a level that it had failed to exceed at previous market tops (around 70). Today at the low it was down -10% to 63. Maybe there was a bit of superstition in the markets. Maybe it's still there.

---

Some Chart Superstition in Tech Stocks¹ - 28 Jan 2026

¹Stock data from Yahoo! Finance

These are two of the most liquid and most highly traded ETFs - the ProShares Ultra QQQ (TQQQ - a 3x leveraged long Nasdaq 100 fund) and the Direction Daily Semiconductor Bull 3X (SOXL). Sophisticated speculators and retail traders alike favour leveraged ETFs because of their liquidity and because they provide opportunity for magnified gain (and losses) and so require less capital to trade for a certain exposure.

The gain for TQQQ from 20 to 60 represents a 200% gain over a period of around 6 months. The SOXL gain was closer to 600% over the same period. These are obscene numbers for a 6-month return and we could imagine lucky owners being desperate to cash in on these gains. If you were a US investor looking to realise your gains - would it be worth risking the market turning, in order to be able to wait long enough so that your realised gains would be taxed as capital gain rather than income or would it be better to simply sell off and take the tax hit?

Usually we don’t pay much heed to technicals or levels. But with the prevailing conditions, there might just be enough caution in the market for superstition and so it might be worth exercising particular caution when considering semiconductor stocks at these levels.

---

A Commodities¹ Supercycle…in this economy? - 27 Jan 2026

¹Tin price data from Trading Economics

A handful of cherry-picked articles highlight the difficulty of predicting a commodities supercycle. We do not mean to disparage the brave souls who dared publicly claim a commodities supercycle was at hand, at or near the top of the previous spike in commodities prices. But perhaps there is a certain kind of hubris in declaring that the input costs to much of the world economy is about to soar to new long-term or even permanent highs¹²³⁴.

Niels Bohr is thought to have said “prediction is hard, especially about the future”. Entirely tangentially there’s a great quip about Niels Bohr here from a book by George Gamow. We’ve all been in the position where a confident prediction went spectacularly wrong. We’ve also been in a position where being early on a narrative or trend is the same as being wrong. Or sold too early. But in order to make money in markets you need to be able to at times sit back and see others make a fortune on something you passed on, or dismissed. This sounds like the pontification of someone who was/is out of the silver rally. No comment.

So commodity supercycles. This phrase has been floating around on our radar a bit more in the last few weeks. Precious metals’ historic rally has been something to behold. And something to get fired for if you missed out on, or bet against, in a trading position. With the US dollar dropping purposefully below a 4-year low today, a day after we (seemingly prematurely) claimed that rumours of the death of the dollar were premature, the threat of further dollar devaluation and support for the next leg higher in everything denominated in US dollars (i.e. nearly everything), looms large…

Federal Reserve Chair Powell slides into view for a rather important meeting of the FOMC, during a storm that has closed federal government offices in Washington DC. Markets expect them to hold rates steady. They’re between a rock and a hard place. At what point does further US dollar decline start to become an imported inflation risk? If they cut interest rates, do they risk spurring inflation? If they hold steady, and inflation has continued to decline in the background, are they effectively pushing up real rates which could hurt bonds and other asset prices. We don’t think the Fed would tolerate much higher yields, or a much weaker dollar.

The unexpected move tomorrow would be a cut, or an increase in the treasury buying the Fed has already been engaging in since mid-December 2025. This would put further pressure on the US dollar and probably provide enough impetus to see the Nasdaq back through to all-time highs and beyond. Because it means real interest rates are coming down. It also, incidentally, would play into Steve Donze of Pictet Asset Management Japan’s rhyming US dollar chart which is keeping us up at night, and making us wonder whether there are still a few rungs lower to go for the USD.

That might just be enough to keep this “commodities supercycle” that we’ve been hearing about going, or off the ground. The only other thing to do so, in our view, would be that state actors have been acquiring essential raw materials because they are preparing for war or other catastrophic supply chain disruption. And if that is the case, then this commodities supercycle is indeed just getting started.

Short of that, however, we think we are at or close to the turning point in the USD and therefore in everything priced in USD. Expensive commodities are very bad for inflation and for politicians unless there’s a war or a pandemic to blame it on.

---

The Death of the Dollar¹² (has been greatly exaggerated) - 26 Jan 2026

¹Stock data from Yahoo! Finance

²Federal Receipts Data from St Louis Fred

Every now and then a chart does the rounds on social media highlighting the decline of US dollars as a percentage of foreign exchange reserves. There is something compellingly monotonic about the decline in that chart, granted. The US dollar index is down nearly 10% from the start of 2025. The yen recently strengthened dramatically against the USD. Precious metals have exploded upwards in price in the last year.

President Trump’s recent rhetoric on tariffs and Greenland and threats of another US government shutdown have all weighed on the dollar. But in addition to the contrarian notion that when social media starts to say “the dollar is dead” (or some such) it usually means the opposite, the world (and specifically all US taxpayers in it) is about to be reminded that there is something very important one can only do with the US dollar - and that is pay US taxes.

Tax day, April 15, is fast approaching. And given that 2025 was another stellar year for US equities (even though it underperformed other markets), following two other stellar years, we expect the amount of tax receipts flowing into the Treasury General Account (TGA) to be extraordinary.

The barchart below shows how dramatically Total Federal Receipts spike in April in particular, but also in some other months (typically quarter-end like June, September, December).

Perhaps the most interesting observation is how often the months leading up to a significant April-spike in TFR seem to exhibit a declining equity market (as shown here by the Nasdaq 100). This pattern is observed in 2020, 2021, 2022, modestly in 2023, 2024 and dramatically in 2025.

One explanation for this might be that as taxes are paid and money flows into the TGA, it exits the banking system and thus reduces banking reserves and effectively drains liquidity out of the monetary system. This is particularly pronounced in the tax season following a year of significant equity gains. It is interesting to note that in 2020 the tax deadline was postponed to July 15 due to the Covid outbreak, announced on March 21. This happened again in 2021 and was announced on March 17 that year. As a result the April-spikes are not so pronounced.

This same pattern can be summarised somewhat differently as below, for a longer horizon. The chart compares annual Nasdaq 100 performance with the % change in annual tax receipts. The relationship is particularly pronounced during years where the equity market has declined significantly (e.g. 2000-2002, 2008, 2022).

We are not the only ones evaluating this. The Federal Reserve was sufficiently concerned about a liquidity drain running up to this date that it restarted buying US treasuries. We are also curious whether the strong performance of equity markets in 2025, combined with the dramatic rise in certain assets already in 2026 (e.g. precious metals), will exacerbate this phenomenon.

Given that some of the realised gains from 2025 might have been redeployed already (which means they would need to be sold again, if some of the proceeds had already been earmarked for tax payments), we also wonder whether this could lead to some downward pressure on equity markets going into April.

What lies ahead?

Here we’ll float some ideas, in increasing order of drama, in which either the liquidity drain could be averted/postponed and/or the potential financial market impact reduced:

• Federal Reserve cuts interest rates and/or significantly increases its Treasury buying for ”technical reasons” before April
• A government shutdown in April postpones the tax deadline
• The US Supreme Court rules the tariffs illegal and that they need to be repaid to businesses
• Other assets are allowed to be used to pay tax (e.g. US treasuries or precious metals)
• An as-yet unknown crisis provides basis for significant fiscal or monetary stimulus

We believe that President Trump does not want the current equity market rally to end and that this period is known for financial market instability. So it seems likely that one of these or another tool may be used to keep the liquidity flowing.

But until we see something like that happening - what do we say to the death of the dollar?

Not today.

And if the US dollar starts to rally dramatically from here, it goes without saying that everything priced in US dollars is going to be in for some turbulence.

We’re looking at you, silver, and at you, USDJPY.

---

Software as a service¹ (as an economic indicator) - 25 Jan 2026

¹Stock data from Yahoo! Finance

Software-as-a-service companies have experienced a bloodbath in share prices over the last year. One of the main reasons given for this is that many people seem increasingly of the opinion that software has become a commodity that can be much more easily produced with LLMs. They are quick to declare “SAAS is dead”. There are critics of this view, and we resonate with some of them. Just because some of these tools are easier to build (or rather, prototype) does not mean they are easy or cheaper to maintain and internally developing anything means that the cost is brought forward, whereas with SAAS it’s spread over time.

An alternative explanation for the decline of SAAS is as follows: many software companies are often initially valued at very high revenue multiples for a long time because of the expectation of significant continued market share growth and earnings in the distant future, because of “winner takes all” dynamics in this business. However as more competitors join the market, market share growth plateaus, and the expectations of significant future growth for each individual company diminishes.

So instead of being valued like a company that will have a 60% market share in the future, it may be closer to e.g. 20% or 30%. If this is why SAAS companies have tumbled - that the total addressable market was either smaller than expected, or each of these companies’ relative control of this market was overestimated, then it seems unlikely that SAAS companies will recover quickly (or ever) to their recent highs. Caution warranted.

A third potential reason for why SAAS companies are down is that subscription-as-a-service is struggling. This could be because of subscription fatigue and/or cost-cutting drives. We think of SAAS expenditure as a barometer for overall recurring corporate expenditure. And if this is being tightened, this may not bode well for the business cycle and perhaps for equities more broadly. This could also explain weakness on the consumer side, with video-streaming companies like Netflix and music-streaming companies like Spotify performing poorly since the end of Q2 2025 and particularly since November. It might be useful to consider online payments processors like Paypal, Block and Adyen in this equation as well and so Paypal's earnings call on the 3rd of February will be closely watched.

In conclusion we think SAAS might be down (beyond broader market movements) for one of three reasons:

1. Competition from LLMs
2. Declining future market share expectations/market saturation
3. Wider subscription fatigue/economic weakness

In our view only if reason 1 explains the weakness then we believe SAAS companies could be worth more investigation. If either of the other two explain it, more caution is warranted.

---

After plunge US Egg Prices are up 120% in a week - 23 Jan 2026

Data from: Trading Economics

Farmers have been struggling in recent years. In the US, farmers are facing $44 billion in net cash income losses across corn, soybean and wheat for their 2025-26 crop due to rising input costs, lower crop prices and trade war impacts. There are similar stories in Europe. And in the UK, a report found that farming was no longer profitable for the average farm. This could pose risks to the future supply of food.

Food is one of the many things we’ve become accustomed to importing and therefore for which we have enjoyed lower prices. In a less collaborative world, it may become more important to be able to produce food domestically. The US relies on imports for roughly 20% of its food, and the UK about 40%. Covid-related supply chain disruptions showed us what can happen to the price of various foods during extreme events.

When prices in a market fall significantly, the less efficient producers either go out of business, get acquired, or get subsidized. In many such cases market concentration will rise which increases suppliers’ pricing power, which can lead to price increases later. And given that agricultural commodities’ supply are significantly affected by the weather, which is effectively random (or very hard to predict) at long time horizons, their prices can move in cycles punctuated with shocks.

A good example of such a product is eggs in the US. The price of eggs, as shown above, spiked dramatically during Covid and again even more so during the HPAI outbreak in 2024 and 2025, reaching $8 per dozen. Since then, the price collapsed to $0.33 per dozen, down nearly 96% from the peak. When adjusting for inflation, as far as we can tell this is the cheapest eggs have been since 1990 possibly ever.

This is likely due to a reduction in the avian flu cases, some amount of demand destruction at such high prices, and a reduction in overall calorie consumption as well as that of many baked goods like cakes (due to Ozempic). This excellent article by the St Louis Fed discusses the market for eggs.

Some of the listed egg producers in the US (Cal Maine Foods and Vital Farms) have seen substantial declines in their share prices since August 2025, by which time the collapse in egg prices was well underway. Perhaps this suggests that their earnings are not so sensitive to declines in prices or perhaps there hasn’t been enough time for that effect to pass through.

Egg prices have now fallen to a level such that many consumers will a) be able to afford them and b) may be less sensitive to even large percentage increases in their price. This presents a potentially asymmetric opportunity for egg producers and their equity stocks. Egg prices are clearly very volatile. But 120% in a week is a staggering rise. Whether we saw the bottom in prices is uncertain but at this rate it is one of the cheapest sources of protein - $0.33 for ~ 80 grams of protein or $0.004 per gram is incredible. And this will likely not last.

We have taken a position in Cal-Maine Foods and are evaluating Vital Farms as well.

There is one important caveat. In a lawsuit filed originally in 2011, Cal-Maine, along with other producers, were found guilty of conspiring to limit egg supply to raise prices between 2004-2008. Beyond the moral outrage at fixing the price of a staple protein, we are also concerned about a repeat of this outcome with potentially far more significant damages.

Several class action lawsuits accuse egg producers and a particular media company (formerly Urner Barry Publications, now part of Expana) of participating in price fixing behaviour. Under the Trump administration there has been an increase in antitrust investigation into agricultural producers, including egg producers.

We might just be counting our eggs before the producers have chickened out.

Question for the future

• Urner Barry publications - could this be a form of algorithmic price fixing?
• Are there other agricultural commodities that have experienced similar collapses in prices which may be at or near turning points?
• Has any industry consolidation taken place for these commodities during these downturns?

---

The new widomaker trade? - 22 Jan 2026

Opinion Survey Data from: Bank of Japan

The Bank of Japan (BoJ) is announcing its interest rate decision soon. Markets seem to expect them to hold rates steady at 0.75%. With a chart looking like the above, for the BoJ to do that either means we misunderstand something, or they are ignoring their current survey respondents. These respondents said they expect inflation to average 10% per year for the next 5-years. Expectations are the highest since they began collecting this survey data, and the last time it was this high was just before the Global Financial Crisis.

And they do not seem to be budging lower. With Japanese Government Bond (JGB) yields at skyhigh levels, the BoJ can’t really afford more inflation but they also might not be able to afford an interest rate hike and the ensuing impact. They are between a rock and a hard place and if they hold rates as the market seems to hope and consumer inflation expectations do not start to come down rapidly they might be in trouble. Will they really tolerate such high consumer inflation expectations? Is it just a matter of time before those expectations translate into wage negotiations?

In the past, the widowmaker trade was to be short JGBs. In the last couple of years, that has been flipped - and shorting JGBs has been one of the best trades available. Perhaps now we’re about to move into a regime with such volatility in these instruments that participating in the market at all could be the new widowmaker trade, especially on the very long maturities like 30-year or 40-year bonds. Bloomberg noted that a mere $280 million of trading volume in the long-dated bonds caused much of Japan’s bond market ructions this week.

On an unrelated note. Intel’s disappointing performance after earnings is somewhat worrying. Does this mean anything for the portfolio of companies that the US government has invested in? We were at least vindicated in our begrudging decision to sell the remainder of our Intel portfolio. Or is this a buying opportunity and possibly the last time you could get Intel at that price?

TACOs at Davos, and Intel - 21 Jan 2026

Bond yield and credit spread data from: St Louis FRED

President Trump delivered a one-two market-boosting combo from Davos today - first indicating that he would not use force to take Greenland (as we suspected), and second taking tariffs off the table because of the prospect of a “Greenland deal” that arose from negotiations with NATO’s Mark Rutte. We’re somewhat surprised by the extent of the apparent climbdown and markets were clearly bursting for this good news - with the NASDAQ 100 ripping from the day’s low of 249 to a high close to 255, an intraday range of nearly 2.5%. We won’t go so far as to point out the things that can be purchased with such a gain in US equity market cap…

We are hesitant to trust the rally today. It smells a bit like a short squeeze. It seems unlikely that uncertainty around Greenland and associated tariffs is over and if not, there’s more market volatility to come. If it is over, we worry about the credibility of President Trump’s threats, more specifically the loss thereof, if he is seen to have climbed down so quickly. If he can no longer inspire fear and awe with threats around tariffs, we are concerned about the tools he may resort to in order to achieve his goals. Arguably, this time around he already resorted to threats beyond tariffs in the shape of potential military force.

Equity markets are an important metric for Trump. And we’ve seen in the past how he was willing to tolerate a certain amount of pain in equity markets, which added to the credibility of his threats. So the worry is that their current lofty levels mean there’s a lot of ammunition left for him - he can make wild statements and threats and afford to sit back amid the ensuing market chaos (and, effectively, implicitly, take credit for whatever market swings apparently result from his threats).

One caveat we have here is that it might not be equity markets that prove to be the limiting factor. The 30-year US Treasury Yield is sitting at around 4.8%. This is close to the pain threshold of 5% that we’ve seen the yield retreat from a few times this year, coinciding in April with dramatic tariff climbdowns or deals. And with the additional threat (although ludicrous and self-immolating, in our view) of European funds dumping US government bonds looming above him, perhaps Trump pulled back so quickly because we were already close to the pain threshold in long term yields. If this was the thinking, then this equity rally probably has some more legs.

Another risk factor worth looking at again are credit spreads which we covered shortly before the tariff meltdown in April 2025. Junk spreads are very close to the Jan 2025 lows (2.65 yesterday vs 2.59). And therefore, close to all-time lows. Does the market really believe, at the moment, that junk-rated companies are at the lowest risk relative to investment grade companies, ever? This seems a bit optimistic, especially with the redemptions that have been happening in private credit. As a result we’re liable to interpret this number as something of a floor - it is unlikely to go much lower, and when it does go up, it tends to be paired with market turmoil. Indeed today it did spike 8 basis points, an increase of around 3% or in the 93rd percentile of daily changes. Not great, not terrible. But if your dosimeter only reads 3.6 roentgen, then maybe you should check your dosimeter.

Intel

On a somewhat unrelated note, Intel is reporting earnings tomorrow evening. We exited the remainder of our Intel position today and are already sad about what seems to be the inevitable rip higher. What if it’s another Oracle-type earnings call and this multi-hundred billion dollar company simply goes up 40% after hours? This earnings call could be a bellwether for a few things - US semiconductor manufacturing, as well as, perhaps more importantly, what happens to companies in which the US government takes a stake. If Intel earnings disappoint, could it mean something bad for the other companies the government has taken stakes in?

The US has made investments in several listed and even private companies including recently a private alumina (and gallium) producer. And we’ve been trying to determine what variant of late-stage capitalism or statism this really is: does the government want stakes in these companies because

1. they are strategically important companies and the government wishes to exercise control over them
2. because they expect to make gains on their equity stakes
3. or is it purely a kind of equity lifeline or bailout for important companies that might otherwise struggle (as evidenced by Intel in previous years)

Reason 1 and 2 are likely positive for future share price prospects. However 3 could mean the government were effectively planning to subsidize an otherwise inefficient on uncompetitive company, which we think bodes less well for share price growth. A lot is riding on this particular company’s earnings tomorrow.

There was an eerie moment today at Davos - President Trump said that the Dow Jones would reach 50 000 soon and double after that. James K. Glassman gave a rather critical address in early 1999 on the topic of government ownership of stocks. He also co-authored a book with Kevin Hassett, now director of the National Economic Council and potential future Federal Reserve governor, called Dow 36 000 calling for the Dow Jones, then at 9000, to double and then double again. It's a good thing we're not superstitious here.

If we had to make a call, we’d say a bit of uncertainty is back. The Nasdaq 100 didn’t even get close to the October lows. The VIX could still have further to fall and if so quite soon options are going to be very cheap again. Overall we’d warrant caution in both directions. There are potential headlines which could appear over the next few weeks, including the supreme court tariff result, which could spike equity market volatility again. Today was a reminder that face-ripping rallies can come out of nowhere.

When do Japanese Government Bonds¹ Become Attractive Again? - 20 Jan 2026

¹Bond yield and USDJPY data from: St Louis FRED

Japanese Government Bonds (JGBs) would likely have been the main financial news story today, had the spotlight not been stolen by the rest of the world’s response to President Trump’s rhetoric about Greenland and his impending arrival at Davos tomorrow. The yield on 40-year JGBs hit 4% for the first time. You can now lock in a 4% return for 40 years. Admittedly, it’s in Japanese Yen which has been bumping up against the 160 USDJPY ceiling, and it seems like the whole world is holding its breath to see whether it breaks through that ceiling with purpose, and whether the Bank of Japan steps in to calm markets. US Treasury Secretary Bessent pointed out that the two-day yield spike in JGBs was a “6-sigma” event, the largest move in yields since tariff “Liberation Day”.

USDJPY trading at 160 feels like a psychological barrier. It most recently reached near that level in early 2025 before the tariff market meltdown and in Aug 2024 before the yen carry-trade unwound. As a refresher on the latter - with interest rates so low for so long in Japan, investors had borrowed Yen to buy equities abroad, increasing the supply of Yen, the demand for currencies (mainly the dollar) and for US equities. Some weak US economic data and the largest rate hike since 2007 later, and the Nikkei had its worst day since 1987 and the Yen strengthened 11% against the USD within a month. Part of the reason for the panic was the expectation that the yen would appreciate suddenly and so borrowed yen came flooding back to Japan under duress.

On the one hand, we’d like to think that the 160 ceiling is likely to hold (even if it gets pierced temporarily). And if so, the prospect of a 4% yield in a (potentially) appreciating currency seems irresistible. However this might be too good to be true. There are a few reasons why caution may be warranted.

Results of the 104th Opinion Survey of the General Public’s Views, published this Monday 19 January 2026, show that the median household 1-year inflation expectation is 10% and a mean expectation of 12.8%. And survey respondents said they feel like prices are up 20% over the past year. A paper by the BoJ shows how inflation expectations, having been seemingly securely anchored around 2% for most of the last 15 years, have become completely unanchored and even at the time of publication (April 2025, with data seemingly up to 2024) was already exceeding levels reached before the Global Financial Crisis in 2007.

The BoJ is still reducing its bond buying program by ~$3 billion per quarter. It has also started a mostly symbolic sale of its equity ETFs and REITs - it plans to sell ~$4 billion in market value per year and in September 2025 its portfolio was worth over $500 billion, so this will take more than a century to complete. It is currently sitting on nearly $300 billion worth of unrealized gains on its ETF holdings. This is however in rather stark contrast to the $200 billion in unrealized losses on its JGB portfolio as rates have continued to climb. If the BoJ hopes to realize the stock market gains and transfer these profits to the government coffers, it may need to do so at a slightly faster pace than $4 billion per year.

Finally, Japan’s Prime Minister Takaichi has called a snap election for the 8th of February in the hopes that she can translate her personal approval ratings into a larger more tenable majority for her Liberal Democratic Party and the Japan Innovation Party with whom the LDP have a coalition. An unprecedented amount of defense spending, market-friendly policies, a spat with China about Taiwan, and an election the result of which is by no means a foregone conclusion - could this be the main macroeconomic act moonlighting as second fiddle?

Again, the temptation to buy JGBs at these rates and with USDJPY at this level is high. And if you're generally of the view that longer term yields and inflation are trending down, this might be the last time you get a deal like this. Someone has clearly been selling them - but maybe this time the widowmaker trade is actually being long JGBs.

The Old Man and the Sea(s) - 19 Jan 2026

Attribution: Gorak tek-en

China-Taiwan relations have featured prominently in the media in recent months. 2027 is seen as a critical year by some foreign policy experts because it is the centennial anniversary of the founding of the People’s Liberation Army, the 21st convening of the party congress and the year by which China’s army needs to have completed material preparation to be able to conduct major combat operations. Some recent analysis characterises thoughts around this in a dashboard or a scorecard.

This leads us to ask - what are some examples of disruptive policy that China could implement which would provide it with sufficient plausible deniability thereby not provoking a justified (or legal) military response or drawing sufficient international attention or response?

The New York Times reported recently on thousands of civilian fishing boats being mobilised into 200-mile long walls in the South China Sea. They quote Gregory Poling of Centre for Strategic and International Studies saying that this could be used in support of a “quarantine”, a measure that would fall short of an act of war, but could cause disruption to Taiwanese trade. A quarantine is also more likely in the short term than a blockade or invasion. It is worth noting that Taiwan’s GDP is extraordinarily exposed to trade and therefore to any disruption of its shipping routes. It is also nearly entirely reliant on imports for its energy.

This recalls the accidental (probably) disruption caused by the 2021 blockage of the Suez canal by the Evergreen and ensuing market reaction. The South and East China Seas, the Yellow Sea and the Sea of Japan have trade corridors that are narrow at certain points and could feasibly be disrupted by coordinated, ostensibly civilian, fishing vessels.

In essence, these routes seem like potentially vulnerable logistics targets which could be disrupted with relative ease and cause significant uncertainty and financial market volatility for Taiwan and the US and Europe.

The Return of Uncertainty? - 18 Jan 2026

President Trump’s tweets recently around a potential acquisition of Greenland over the weekend has stoked the ire of EU members and NATO have issued a joint statement on this. There have been talks of readying a “trade weapon” by the EU, as well as up to EUR 93 billion of potential tariffs. Weekend futures markets are showing the Nasdaq down -0.9% and Dow Jones down -0.7%. Interestingly, major German and French indices are indicating down -1.3% and -1.1%.

This could very well just be posturing and classic negotiation tactics - start with a flurry of worst-case scenario threats in the run-up to negotiations in Davos, which gets a harsh initial response from the EU, and then resolve to strike a deal somewhere quite reasonable in the middle. It might even be another case of the controversially termed TACO. From a financial markets perspective, we don’t believe they are pricing in a significant ramping up in tensions between the US and EU (yet?) and certainly not a military intervention or standoff in Greenland. If, as we are led to believe, a key complaint of the US in this recent rhetoric is the EU and NATO’s lacklustre military spending and insufficient military presence in Greenland in particular, then ironically the latter may already have been partially addressed with NATO troops sent to Greenland.

There is a caveat to this. Equity markets are at or near all-time highs, opening up some questions about stretched valuations. There have been some significant private credit fund redemptions. Bond yields have remained relatively elevated while inflation rates and expectations have softened somewhat. Truflation’s measure of inflation has been coming down quite quickly. This means real yields have been spiking. For two Fridays in a row, CBOE Put-Call Ratios have dipped outside of typical territory. The Federal Reserve has been adding liquidity for “technical reasons” for over a month now, and seemed concerned about the April 15 tax deadline and ensuing liquidity. The Supreme Court may rule as early as this week on the legality of President Trump’s tariffs, the impact of which on the economy, financial markets, and on his policy response, is unclear. There is also some contention in crypto land, with Coinbase CEO Brian Armstrong apparently managing to kill a senate committee vote with his opposition to the CLARITY bill.

Financial markets are very good at pricing risk - specific, measurable quantities and relatively well defined outcomes with tractable probabilities. They are less good at pricing uncertainty. During the March 2020 Covid market meltdown and the April 2025 Tariff crash, US economic policy uncertainty spiked dramatically. And there is some potential here for a political misstep to start a chain reaction of policy responses that reignite this uncertainty.

In the face of this uncertainty, one can only look to the incentives at play here. If President Trump is truly getting on war-footing, and suspects escalation from, or wishes to escalate with, Russia or China, further attempts at acquisition and ensuing rhetoric are highly likely. Furthermore, tariffs with a deadline could be an effective measure to frontload imports or stockpile important goods, and also consumer spending (boosting Q1 GDP), especially the latter if the threat of tariffs reignited inflation concerns.

A somewhat fringe notion is also interesting to consider. With all this uncertainty going on, and the US government likely having additional information on other potential sources of disruption that may emerge over the coming months, President Trump may well suspect that financial markets will turn down in this quarter, independent of his policy actions. If markets are already “due” to turn down, for whatever reason, there is some logic in “leaning into” it. Perhaps it is better to be seen to be responsible for a market crash, and possibly also for its end when it finishes, than to be blamed for being able to avert it. It is admittedly a bit far-fetched and 4-D chess. But it would certainly not be the strangest thing to have happened in the last year.

Having observed how this market has been virtually unwilling to go down in any meaningful way since April 2025, it is very tempting to call this resurgence in tension posturing and an opportunity to buy the dip. And it's at moments like these where all the sagelike platitudes of "it's about time in the market, not timing the market" and "more money is lost waiting for a crash than in the crash itself" tend to resurface. But on the other hand, people probably weren't expecting April 2025 when they were standing at all-time highs in February 2025.

So just ask yourself - do you feel lucky, punk?

CBOE Total Put Call Ratio¹ vs Equity Indexes² - Extended - 17 January 2026

¹Put-Call Ratio data from Chicago Board of Options Exchange

²Stock data from Yahoo! Finance

On Friday the 9th of January the CBOE Total Put-Call Ratio (TPCR) hit 0.75. This is in the 4th percentile for 2024-2026 and values equal or below 0.75 have occurred 25 times in the last ~500 trading days. The average 5-day change in the Nasdaq 100 starting from a day the TPCR hit 0.75 or below, was -0.81%. The average 5-day change across the entire 2-year history was +0.4%. From Friday 9th Jan to 16th Jan close of trading day (5-trading days later), the Nasdaq 100 was down -0.34%. So technically, this would count as another confirmation of the signal’s predictive potential, which we covered on 11 January 2026.

We modified the chart from that day by:

• altering the time–horizons (adding 2 and 3-day horizons and dropping 6 and 12-month horizons)
• adding two new indexes: the S&P 500 and the Dow Jones Industrial Average
• extending the TPCR thresholds (now from 0.75 to 0.95)
• adding the index “benchmark” comparison, represented by a yellow cross, which represents the proportion of times the index was negative over a given time horizon.

One should exercise caution interpreting these results, however. We would have had more confidence in the predictiveness of this signal if there was more consistency or monotonicity across the time-horizons.

Overall the interpretation is as follows: for a given threshold and time-horizon (e.g. 0.75 and 10-days), if the coloured circle lies above the yellow cross, it means that the relevant index was negative more often over that time horizon after hitting a TPCR threshold or below, than over that horizon in general (i.e. any random 10-day period).

The CBOE TPCR hit 0.73 this past Friday 16th January, lower even than the lowest threshold we tested. This value or below only occurred 13 times in the last 500 trading days. If you expect this signal to hold again then the Nasdaq100 could be lower again by the end of this week. Weekend markets are currently showing some negative expectations following US President Trump’s tweets about putting tariffs on European countries that disagree with the US approach to Greenland’s “acquisition” but there is a lot of time between now and Tuesday's market open, nevermind Friday.

The Price of Protein - 16 January 2026

The US government recently upended the old food pyramid and is calling for Americans to emphasize protein in every meal. Whether this is in service of public health or the cattle lobby, we may never know. But we expect it probably won’t arrest surging beef prices. We previously postulated that Ozempic might have a deflationary impact on spending not just on waistlines, and a paper in the Journal of Marketing Research found that households reduce their grocery spending by 5-8% on average within 6 months of starting.

We should have caveated our statement - it may reduce overall spending, and it may reduce spending on (and prices of) specific goods a lot. However it seems likely that the cost of some foods may go up significantly. And with the US government now pushing protein harder than ever, it seems likely that sources of protein could see continued price increases. We have a particular hunch about cottage cheese (because parmesan is already expensive).

In the UK, demand for whey protein has gone up. And annual beef price inflation hit 27% despite demand volumes dropping since before the recent price spike, a dystopian reminder that maximum profit (even for food producers) might occur with selling lower volumes at higher prices.

Question for the future

• How to build a Protein Price Index for different countries?
• How has a “protein portfolio” performed in recent months/years?

Consumer Surplus and a Shrinking Pie - 15 January 2026

This is not really a snippet but rather a collection of thoughts that will be expanded on in more detail later

At the root of much of the geopolitical, social, and financial market events of the last few years lies, we believe, a fundamental shift in political attitude - a shrinking (or less slowly growing) economic pie. And given that the expectation not that long ago was of a growing pie, even a less slowly growing pie may feel like a shrinking pie. A local proxy for this is simply inflation, particularly of things like house prices.

For many years we were happy with globalisation - relatively free movement of people, goods and ideas. Specialisation and division of labour, offshoring, and regulatory arbitrage saw declining prices and rising living standards across much of the world. There is now significant opposition to this and increasingly so on an ideological level.

What was the cause of this shift? Part of it can be explained by shifting demographics. Much of the world has both a fertility rate below replacement and an aging population able to produce less output so there is reason to believe that output per person (the pie) will decline unless productivity increases. Covid and the realisation of supply chain vulnerability likely also played a role.

Poor policy is also to blame for the shrinking pie and while there are many contenders for worst policy choice, surely amongst the top ones is the approach that much of the Western world has taken to homebuilding. We hesitate to call it a nimbyist approach because of its connotations, but through incompetence or malice, and with a sprinkling of climate panic (not without reason, admittedly), the Western world has made it increasingly difficult to build homes where people want them. This effectively limits the supply of homes which maintains (and increases) house prices. Housing shortages have extraordinary downstream effects including lower fertility, as many people put off family formation and having children. In cities like London it plays role in suppressing economic growth, as skilled workers are unable to move to where the jobs are.

Another underappreciated cause of this we would like to highlight relates to the notion of consumer surplus, specifically with regards to the role that consumer data collection plays in transferring consumer surplus to producers. With an increasingly detailed consumer profile of every consumer being constructed through data collected on them, it is getting easier to find out exactly how much someone is willing to pay for anything.

In combination with algorithmic pricing, a practice whereby prices for products are set automatically with the help of data collected about both customers and the market environment, this makes it hard to imagine why the prices for many things may ever come down outside of a consumer-spending driven recession. Arguably one of the most significant recent developments of this might be the introduction of LLMs to the personal healthcare space (more on this in the near future).

Looking to the future - there are certain industries and products where consumer surplus is still very large and therefore present potential targets for surplus-transfer by producers:

• Electronic/digital communication - particularly long-distance (international voice and text communication)
• Internet connectivity
• Digital search (arguably being disrupted by LLMs)
• Personal computing and data storage
• Food
• Electricity
• Entertainment (although perhaps in e.g. streaming content we might be approaching some sort of saturation point)

One of the manifestations of "shrinking pie" behaviour is seeking to increase the yield from an existing resource. Unfortunately for consumers this means getting people to pay more for the same thing, instead of producing more of the thing and selling it to more people for less.

Should the US invade (or buy) Greenland? - 14 January 2026

Let’s begin with a strong statement. The US should under no circumstances militarily invade Greenland while it is under the sovereign control of Denmark. Does this calculus change if China or Russia, for example, invaded or otherwise established a strong military presence there? Maybe. Under the current circumstances, we also think a US invasion is very unlikely as it would represent an extraordinary act of aggression against an important ally (i.e. NATO by way of Denmark), despite the recent rhetoric of the Donroe Doctrine. Greenland is not Venezuela. What’s more - why invade when you can buy?

Should the US buy Greenland? That is a far more interesting and relevant question. Would it be good for the US? Probably - Greenland is rich in important resources and has a critical strategic location. Would it be good for Greenland? For Denmark? For the rest of the world?

Perhaps one should ask this question with an eye toward the future. What was 2015’s 10-year view of the future? This was before President Trump’s first term, the first trade war, Brexit, China’s major growth scare, Covid, the Russian invasion of Ukraine. And so the likely answer would probably have been something involving more of the same - a continuation of the plus-sum, pro-trade, globalisation-centric, collaborative, international rules based order. In 2015.

Now, however, this does not seem to be the case. The order is fraying, to put it mildly. And the prospect of a major conflict in the next 10 years between nuclear powers seems much more likely than it did in 2015. This should dramatically change the calculus for all parties involved in this issue. And so now we must ask - how much?

In a very crude way, what might Greenland be worth? One could use the GDP as an estimate of annual “revenue”, assign some sort of discount factor or multiple and value it on that basis. However Greenland’s GDP is 40-60% government spending, nearly half of which is significantly reliant on a Danish grant, which comprises about 20% of GDP. We can already hear someone saying “this is not a profitable country”.

Alternative estimations might use the supply of natural resources. It is rich in oil and natural gas with a US geological survey estimating nearly 18 billion barrels of oil and over 4 trillion cubic meters of natural gas. Even with an unrisked estimate of 4 billion barrels of oil, an initial 1 billion barrel target, at current prices alone, could be worth $60 billion. There are several estimates out there, taking into account real estate and rare earth minerals. Nevermind paying each Greenlander $100 000 - might as well make it $1 000 000 or $10 000 000.

If the US offered each Greenlander $1 000 000 and US citizenship, along with a stronger military presence, would that really not pass a popular vote?

We are skeptical.

Did the Fed’s treasury-buying¹ dampen stablecoin² redemptions' impact on short-term US treasury yields? - 13 January 2026

¹Fed Balanace Sheet data from St Louis FRED

²Stablecoin market cap data from CoinMarktetCap

The Federal Reserve started adding to its balance sheet again in mid-December by buying up $40 billion worth of short dated treasury securities. This was purportedly purely for liquidity management and follows signs of liquidity stress in October. We wonder whether the impact of the Fed’s securities purchases may have dampened the impact of stablecoin redemptions on short-term US treasury yields, after the dramatic selloff in cryptocurrencies in the last quarter of 2025.

Tether, the issuer of USDT, and Circle, the issuer of USDC, together comprise the overwhelming majority of the stablecoin issuance market. And stablecoin issuers are, collectively, one of the largest holders of short-dated US treasuries, as large as some countries (e.g. Germany, South Korea). And while this may be a small share of overall US debt, a BIS paper finds that significant flows in and out of stablecoins have an impact on 3-month treasury yields, with outflows increasing yields 2-3x as much as inflows decrease them.

Treasury bills and other cash-like securities are bought and held by stablecoin issuers as collateral to support the peg of their respective stablecoins against the US dollar at a ratio of 1:1. In theory the peg is safe as long as the issuer holds enough money-like collateral to support the peg. At time of writing the USDT market capitalisation was around $186 billion and the USDC market cap around $75 billion.

In the past when the price of cryptocurrencies has been rising, the demand for stablecoins has been high or increasing. This has led to an increase of issuance and as a result an increase in the purchase of collateral by these issuers. In theory, the converse could be true - if cryptocurrency prices decline, demand for crypto and stablecoins may decrease and people may wish to redeem their stablecoins. This would lead to a reduction in the market capitalisation of the relevant stablecoin, and a sale of that issuer’s collateral. In this case, quite possibly, short-term US treasuries which are liquid and easy to sell, which could lead to their yields increasing.

After the recent market peak in October 2025, major cryptocurrencies declined significantly. Bitcoin went down from a peak of $123 000 to around $82 000, a drop of nearly 33%. All told, the broader cryptocurrency market cap declined ~33% from around $ 4.3 trillion to $2.9 trillion. It is not a surprise that USDC saw its market capitalisation decline during that selloff. Tether, interestingly, remained resilient.

Hunting Academic Fraud - Prediction Markets Part 2 - 12 January 2026

A British scientist has been granted a $2.5 million+ reward from a settlement for a claim filed against the Dana-Farber Cancer Institute under the US False Claims Act.

This brings to mind an initiative by Bill Ackman in 2024 when his wife was embroiled in a plagiarism case involving Business Insider and the faculty of MIT. He could likely dedicate significant resources exclusively to the task of finding plagiarism in academic journals but not everyone has these sorts of resources.

In the US the False Claims Act provides for whistleblowers to receive a bounty of up to 30% of a settlement amount, if they disclose information showing that a fraud led to financial loss for the government. But a lawsuit can also be a costly and time-consuming endeavour. Perhaps this presents an opportunity for prediction markets.

PolyMarket offers prediction markets on lawsuits, including the outcome of the Supreme Court decision on President Trump’s tariffs, as well as whether President Trump will sue Federal Reserve Chair Jerome Powell. There is clearly already a precedent.

We imagine that a prediction market could be set up around questions such as:

• Will a published journal research paper be retracted?
• Will a fraud case be opened against Organisation XYZ?
• Will a fraud case against Organisation XYZ reach court or be settled?
• Will the settlement value be disclosed?
• Will the settlement value be within or beyond a certain range?
• Will it be appealed?
• Will an appeal be heard?

Our tests for a good prediction market:

• Is there potential public benefit in knowing the outcome, range of outcomes, uncertainty around outcomes, or market participants’ beliefs around these outcomes?
• Does the existence of this market pass some sort of cost-benefit test where the cost would include, amongst other things, the potential for abuse or corruption around influencing the outcomes of the market?
• Is there a robust, impartial verification system in place for the market outcome?

We posit generally that this practice can be a good thing. There is enormous potential for abuse, of course. And so this should be closely monitored by relevant authorities.

Prediction markets aside, this lawsuit result and its associated press coverage may very well spur on other sleuths to go looking for flaws in academic articles.This also ties in to a broader theme of prominent figures taking a strong stance against fraud in public spending (e.g. Elon Musk’s notorious DOGE efforts in 2024, and more recently comments by Chamath Halipitiya regarding waste and potential fraud in California).

Question for the future

• Who are the largest vendors in terms of receipts of government funds?
• Which of these vendors’ contracts rely on bidding processes which involve highly technical/scientific knowledge and therefore may have higher risk of the government being defrauded (if those vetting the bids are not as sophisticated as those applying for the funds)?
• What are the equivalents of the above and of the False Claims Act, and associated whistleblower provisions, for other countries?

Updates

The Wall Street Journal yesterday (11 January 2026) covered the use of prediction markets as a form of price insurance in real estate, which we covered on 6 Jan 2026.

Testing Predictive Signals - 11 January 2026

See Main Page for today's publication

Stock Market¹ Extremes? - 10 January 2026

¹Stock data from Yahoo! Finance

Just because a stock is at an all-time high does not mean it is overvalued or due for a correction. We live in interesting times - markets are being driven by AI companies, and there is sufficient geopolitical strife and uncertainty that getting access to the resources that power it, which currently depend on an interconnected global web of supply chains, is no longer a foregone conclusion.

And when one thinks of financial markets as being, at their core, about capital allocation then the US financial markets are allocating a huge amount of capital towards, effectively, securing the resources needed to continue the AI revolution in a more uncertain future. Nevertheless, it is quite impressive to see just how many indexes are at all time highs.

Across a number of large equity markets, in developing and industrialised nations, equity market indexes are near record highs. But another observation can also be made - in the past, some markets set new all time highs for extended periods of time before experiencing any sort of substantial drawdown. That said, it might be worth beginning to investigate the reliability of certain market “top” signals.

Some caveats

• Not denominated in same currency
• Poland and Hong Kong’s market peaks were actually before the 2012 start but current levels are very close
• Not adjusted for inflation
• Other asset classes (e.g. precious metals and some currency pairs like USDJPY) are also at extreme levels

Do institutional investors significantly impact the market for US single family residential houses? - 09 January 2026

US President Trump recently tweeted that he was going to take steps to ensure that large institutional investors (IIs) could not continue to buy single family residential homes (SFRs) in the US. Almost immediately after, the share price of two major IIs (Invitation Homes and American Homes 4 Rent) as well as a large private equity firm (Blackstone) dropped sharply.

Some pundits were quick to point out that policy failures like zoning issues are what has led to housing shortages and high prices, not institutional ownership. They point out that IIs are a very small proportion of owners of the overall US SFR housing stock (ranging anywhere from 3-5%) and so can’t likely influence prices.

This misses a few important caveats as investors make up a large proportion of:

• Certain geographical markets - e.g Atlanta, Georgia (25%) or Jacksonville, Florida (21%)
• SFR transactions (i.e. buyer/seller rather than owner) - e.g. a third of all SFRs in Q2 2025
• The available rental market (the vast majority of SFRs are owner-occupied and so not necessarily available to rent)

It is worth noting that mortgage rates in the US hit their highest levels in 20 years in October 2023, with 30-year mortgage rates hitting nearly 8% and only coming down to 6% this year. And with nearly 54% or 30 million households with mortgage rates below 4%, many do not have an incentive to sell or attempt to trade up.

Question for the future

• How large does a market participant have to be, in terms of market share (either of the overall stock, or of transaction volume) to influence prices?
• Do institutional buyers act as volatility dampeners or enhancers?
• What impact do these buyers have on price discovery, especially when conducting large portfolio transactions?

Can hyperscalers help drive traditional media...higher? - 08 January 2026

Much was said in 2025 about the “circular deals” phenomenon amongst the companies involved in building the hardware/infrastructure for LLMs (including the chips and, increasingly, energy provision). This has, possibly, helped to inflate/elevate the share prices of some of these companies. We wonder whether this may soon happen, and perhaps has already been happening, with traditional media companies.

Copyright holders of various sorts are suing, and have sued, LLM companies with mixed success. NYT is suing Perplexity (and Microsoft and OpenAI). Anthropic agreed to pay $1.5billion in a settlement with book authors. And there are some licensing deals already out there.

In May 2025 the New York Times signed a licensing deal with Amazon for its content. The deal is in the range of $20-$25 million. Its share price, incidentally (could be entirely unrelated), is up nearly 30% since then. So let’s take a closer look at some of the NYT’s numbers.

The New York Times group had a trailing 12-month revenue of $2.8 billion and net income of around $338million. So this single licensing deal is ~1% of revenue or 8% of net income. If you classify the licensing deal directly as earnings, the extra 8% in earnings is worth an additional 8% in market capitalisation.

Admittedly, this is very hand-wavy and back-of-the-envelope math, at best. But this is also only one licensing deal, from one hyperscaler. There are potentially more of these licensing deals to be made, with other hyperscalers. And as AI slop becomes more ubiquitous, ironically, the value of verified, high quality “traditional” editorial media content may increase due to relative scarcity.

There is another factor that makes this market interesting - a large number of publishers and traditional media companies are privately held which means there might be more scope for somewhat creative accounting.

Question for the Future

• What is the most appropriate way to classify/account for such licensing deals in terms of likely profit or revenue impact?
• what is the scope for other such licensing deals and what might they imply for valuations?

The End-of-the-World Trade - Part 1 - 07 January 2026

¹Stock data from Yahoo! Finance

With the US “invasion” of Venezuela, the US and UK seizure of a Russian-flagged ship in the North Atlantic, and protests in Iran spreading to more provinces, there is a whiff of “end of the world” in the air. Not necessarily apocalyptic, but probably an end of the old world order of relatively peaceful, plus-sum globalisation and trade. Following hot on the heels of a year of tariff tit-for-tat, this is not new. But the fever seems to be building. (Interesting note - Venezuela supplies about 4% of China’s oil imports, and Iran around 14%).

This “end of the world” may happen to varying degrees of calamity. But at the core of this admittedly still somewhat amorphous idea lies a disintegration of the rules-based, collaborative international order. And stemming from this, a dramatic increase in the need for self reliance, at the very least on a national level, and an associated increase in decentralisation. So what does preparing for decreased collaboration, or even for outright conflict, look like for a superpower like the US? Well, probably a stockpiling of materials and reshoring of capacities that enable modern society and financial markets to operate:

• Power generation capacity and resilience (solar, batteries, electrical transformers, now possibly other countries’ oil reserves…)
• Base and Precious metals (for industrial use and money alternatives)
• Everything needed for “compute” and AI (memory, GPUs)
• Defense
• Satellites (for military and potentially civilian communications use in the event of disruption)
• Cryptocurrency

If one looks at the relative performance of the above over the last year, and particularly since the bottom of the market in April 2025, even in comparison to high-performing equity indexes like the NASDAQ 100, there has been enormous demand for these.

A series of social media posts today have singled out defense companies in particular, with President Trump calling them out for dividends, stock buybacks and executive compensation, even going so far as to single out Raytheon as a specific manufacturer.

Whether or not the “world is ending”, one thing is for sure - like selling downside protection to retail “perma-bears” in an AI-fuelled market rally that doesn't seem to want to end, preparing for the end of the world is very big business indeed. The biggest.

Some thoughts on Prediction Markets - Part 1 - 06 January 2026

Prediction markets could, in theory, facilitate something that financial derivatives have never been able to - highly specific risk hedging (insurance) for the general public. Futures contracts, the original derivative, were invented to enable farmers to insure themselves against nature - a very useful feature of financial innovation.

But these contracts and other derivatives tend to be highly standardized, require a fair amount of financial literacy and sophistication, and tend to be traded on margin (with debt) which for most of history made them risky and somewhat inaccessible. And until recently, most bookmakers were limited to sports or maybe some political races, in terms of markets they would make.

But now with platforms like Polymarket and Kalshi, you can place bets on virtually anything, ranging from whether Trump might acquire Greenland before 2027 to which movie will win the Best Picture Oscar. Extraordinary entertainment value aside, the granularity of these markets means that it is easy to imagine a world in which one could e.g. protect oneself from rising house prices, by simply placing “Yes” bets on “Will house prices increase in Area XYZ?”.

At the moment, the only way to hedge yourself against rising house prices is to buy real estate, shares in public or private real estate companies, or other inflation-resistant or real assets (e.g. silver or gold).These have significant disadvantages, including being non-specific (what if you care specifically about houses in a certain area?) which could potentially be resolved with prediction markets.

We are not so naive as to imagine that only good outcomes will arise. Prediction markets attract, among others, gamblers and “retail traders” who are notoriously good at losing money. And with market maker firms like Jane Street (allegedly) and Jump Trading (officially) joining in, the retail trader is now sharing the pool with the best paid sharks (said with affection) in the business.

Be careful out there.

Question for the future

• What makes for a good/useful/interesting prediction market?

Examples of some interesting potential prediction markets:

• What gets discussed in the UK parliament
  • Relevant/public interest
  • Easily independently verifiable (via Hansard)

Moore, Swanson, and Batteries (and maybe a bit of Jevon’s Paradox) - 05 January 2026

¹Stock data from Yahoo! Finance

Note: We have previously held a substantial holding in the VanEck Rare Earth and Strategic Metals ETF (REMX) and continue to hold a position in it.

Prices can seem to drop year after year, until they don’t. Moore’s law is dead, so declared Jensen Huang in 2022 and predicted Bob Calwell back in 2013. The doubling of transistors on an integrated circuit every two years does not seem to be happening anymore.

Swanson’s law refers to a similar finding - an industry specific application of Wright’s law whereby the price of solar photovoltaic modules seem to halve for every doubling of cumulative shipped volume.

There appears to be a similar phenomenon for electric batteries - a significant reduction in cost over time, as installed capacity (and EV vehicle production) has increased dramatically.

Whether causal or not, a commensurate increase in industry and supply concentration has been associated with these cost reduction phenomena, both geographically and by manufacturer capacity. The solar PV, semiconductors, and some battery minerals supply chain exhibit this phenomenon. In more concentrated markets, it is easier for suppliers to raise prices and control the supply and such a market is more vulnerable to external shocks.

For microchips, one example of such a “shock” was the advent of LLM-driven AI and the sharp increase in demand. The resultant meteoric rise in price of the hardware underlying this technology, and the share price of manufacturers, speaks for itself - supply was not able (or willing) to meet the increased demand with production quantities that would have held the market price steady.

While it may be too early to tell, we suspect that we find ourselves in a similar position now with batteries and the solar PV supply chain:

• The entire supply chain has become highly concentrated - geographically and by manufacturer
• Supply chain concentration makes it more vulnerable to demand and external shocks
• The shocks (may) have arrived:
  • Geopolitical tension between the demand (US) and supply (China, Latin America) nations
  • Grid limitations meaning that solar PV is the “marginal electron” and that demand for backup energy will likely increase

Fire in Berlin (and a Battery in Every Home) - 04 January 2026

A fire in Berlin has left nearly 50 000 homes and 2200 businesses without power. Officials suspect arson and have compared the events to previous suspected arson attacks. This time however it’s -6C in Berlin and the local heating system has been affected as well. It may take several days before electricity and heating are restored to many of these houses, and emergency housing has been set up for people affected.

In March 2025 when the Hayes substation fire that shut down Heathrow happened, airline stocks fell in the next trading session. It made us wonder whether “market manipulation by sabotage” would increase during 2025. Both the Heathrow and Berlin fires have underscored the importance of having substantial backup power facilities - both at household and industrial scale.

With utility-scale backup battery installations becoming cheaper, and grid failures and interruptions seemingly becoming more frequent, we expect the pressure for action on this front to increase. And given that a combination of solar and battery storage are the most readily deployable backups, demand for these might increase.

Also, if there is an expectation of increased geopolitical conflict (e.g. with China or Latin America, key producers of rare earth and battery minerals) that could raise future prices of the critical minerals needed to build these systems, it may well create demand in the present for these minerals.

Question for the future

• Is there a Moore's (Swanson's) law for battery prices?
• What happens to prices in industries/technologies that have exhibited Moore's law type phenomena?

Can Real Interest Rates¹ Predict the Stock Market²? Maybe - 03 January 2026

¹Real yield data from Econforecasting.com

²NASDAQ 100 Index data fromYahoo! Finance

Real interest rates matter and could impact financial markets. They are inflation-adjusted interest rates. One reason is that real yields on government bonds e.g. US Treasuries are seen as the benchmark risk-free rate. If you can get a guaranteed 5% inflation-adjusted return on your investment in bonds, then equities could seem less attractive and so investors may allocate more capital towards fixed-income and less to equities.

Real interest rates are also used as a “discount factor” - they are used to calculate the present value of future cash flows of an investment. A higher discount rate (larger denominator) implies a lower present value.

Real interest rates can go up, mechanically, for two reasons - either nominal interest rates are going up or inflation is coming down. In this manner, perhaps counterintuitively, a higher rate of inflation means a lower real interest rate.

While the real rate is not the only important variable, it is interesting to compare it with the NASDAQ 100 over the last 5 years in which there are 3 examples where tightening real rates preceded or coincided with a significant stock market correction:

• Market top in Dec 2021
• Market top in July 2024
• Market top in Feb 2025

It is interesting to note that in the most recent market high was reached in October 2025 and that real interest rates have been climbing steadily since then. There was a small (but still significant) ~8% correction in the NASDAQ100 around then.

Note: There are other interpretations for what real rates and their changes signify e.g. real rates going down might signal expected economic weakness (a good contender for why there was a 2025 correction in Q1, in addition to the market turmoil caused by tariffs) and vice versa, which would actually make the relationship positive between real rates and equities. And given that inflation data is only released with a lag, and may be somewhat backward looking, getting the direction of causality, if any, right is a challenging task. But remember - if someone could predict inflation or interest rates accurately, you would not know about it. They’d be sitting in a hedge fund on a pile of gold that stretched to the sky.

Question for the future

• What is the predictive power of real rates (dispersion, turning points, changes) for securities e.g. equities?

Use UK Wind Curtailment Payments¹ and Energy to Mine Crypto in Scotland - 02 January 2026

¹Curtailment cost data from Robin Hawkes

In 2025 the UK spent nearly £1.5 billion in curtailment and turn up costs for wind power generation - that is, money paid ultimately by the consumer for wind farms to stop producing power (curtailment) and for gas or other electricity providers to step in (turn up), due to electricity grid constraints.

Grid constraints are increasingly the limiting factor in power distribution (in the UK and elsewhere) however given the significant annual curtailment costs, there are some proposals to reduce waste:

• Mining cryptocurrency like bitcoin with excess energy
• Battery energy storage systems are becoming cheaper
• Datacenters

Surely there is a combination of the above which can convert the wasted curtailment expenditure into a more useful resource. Deploying a bitcoin mining rig, for which the primary expenditure after deployment is energy/electricity, in conjunction with a battery energy storage system, seems most promising.

This idea has clearly been around for some time, with Bitcoin talk forum members alluding to it as early as May 2018, and with much more specific interest since January 2025. It was also proposed on the Scotland Government website in 2022.

Major DRAM Manufacturers¹ - 01 January 2026

¹Stock data from Yahoo! Finance

Consumer surplus is a good proxy for happiness. It is the monetary benefit consumers gain when they buy something for less than they are willing to pay. It is the feeling of exceeded expectations. Producers (companies) want to identify consumer surplus and transfer it to themselves - every dollar of consumer surplus is a dollar that the producer could be charging more (to that specific consumer - there are caveats/assumptions).

How do producers do this? Simply put, by raising prices. But raising prices is not always easy. If you raise prices, consumers may find alternatives. If there are no alternatives, this could be because the producer is monopolistic/anticompetitive. Ideally, a producer wants to have market control or pricing power. However this could attract regulatory attention which is dangerous for the producer.

So in addition to pricing power, a producer wants a narrative. A narrative is a believable explanation, or story, for why prices “are going up” rather than simply “we are increasing prices because we can”.

And the enormous demands that the AI ecosystem is placing on resources is a very good narrative for increasing prices.

Crypto Treasury Companies¹ - Part 1 - 13 July 2025

¹Stock data from Yahoo! Finance

Some Companies Have Seen Huge Stock Price Moves After Announcing Cryptocurrency Treasury Intentions

Additional Warning:Investing in cryptocurrencies and cryptocurrency companies, listed or otherwise, may carry even greater risk than other investments. As this chart shows, crypto companies may experience extreme volatility, both positive and negative price action. This post is in no way a recommendation to trade in cryptocurrencies or related assets.

The Smarter Web Company has been in the news for rocketing to a £1 billion valuation after its IPO with the aim to become a bitcoin treasury company, looking to capitalise on the valuation premium that Strategy (formerly Microstrategy) is currently enjoying over its bitcoin holdings. Given how well that has worked, other companies are likely to follow suit. And they have been.

BitcoinTreasuries (not affiliated) maintains a database of companies that have become bitcoin treasury companies. Monitoring news websites (e.g. Twitter or MktNews) for companies that announce plans to become cyrptocurrency treasury companies could be a lucrative (if very risky) trading strategy.

Question for the future

• What separates sustainable crypto treasury initiatives from pump-and-dump/scams?
• What are the characteristics of these companies?
• Can these characteristics be used to predict which companies will in future become crypto treasury companies?
• What are the implications for cryptocurrency prices of increased adoption?
• What financial innovations (e.g. PIPE) are facilitating this phenomenon?
• What are the regulatory burdens in different jurisdictions to become a partial/primarily crypto treasury company?

The Ozempic Economy - 25 May 2025

Will Ozempic (and other GLP1s) have a deflationary impact on the global economy?

In Apr 2024 12% (~30m) of American adults had tried a GLP1 and 6% (15m) were using one. Studies show calorie reduction of 700-900 per person, reduction of 16-39%, for users of GLP1s. Consumption reduction mostly hit processed foods, refined grains, and sodas or sweetened drinks.

Starting with the conservative estimate of 15m current users and a 700 calorie reduction, that is roughly equivalent to removing 5 million people (assuming a typical 2000 calorie diet) from the market for certain foods. Could this already be having an impact on demand?

The prices for certain foods have come down significantly. The prices of eggs, potatoes, orange juice, wheat, and sugar are down significantly over the last few months. Admittedly some of these foods are still up hugely since their pre-Covid level, and other factors (weather, avian flu) have undoubtedly had large impacts.

And if you look at the shares of some of the largest processed food companies in the US, some of them are down significantly from their highs (Kraft Heinz down 25% since Oct 24, Tyson Foods down 42% since Apr 22, General Mills down 40% since May 23). Of course, again, other factors could have been at play.

Question for the future

• If GLP1 uptake continues to increase - how large will the demand reduction get?
• Which sectors will most benefit from substitution (e.g. fitness, healthy foods)?
• If GLP1s are used to regulate other addictive behaviours (alcohol consumption, gambling) which other non-food/health sectors will be affected?

Risk Spectrum - The Flip - 14 April 2025