Throughout the spring, Paul Krugman went on a bit of a crusade to argue that the standard of living in the US is not better than in Europe and that Europe did not fall behind in the last decade or two in terms of living standards.

As a proud European, I agree with what Krugman said on this topic. The key argument is that GDP per capita expressed in US dollars is a stupid and highly misleading measure of living standard. I have addressed this at length in this post and will skip that bit today.

Instead, I want to point your attention to a study by the European Commission that compares broader metrics of well-being in the US and the EU. Yes, I can hear American readers complain that this is just a bunch of motivated reasoning, and naturally, the European Commission would come up with a study that shows that well-being in Europe is better than in the US.

I will contemplate this criticism during my two weeks of holidays in a European country, enjoying the food, nature and local culture. And when I am back, I might respond to it while looking forward to my remaining three weeks of vacation time left in 2026.

The European Commission looked beyond GDP and productivity alone and included measures of well-being, which include a range of indicators from real disposable income, the affordability of houses and employment rates, to such things as access to healthcare, pollution and self-reported life satisfaction. This broader measure of wellbeing increased by 9.5% between 2014 and 2025, while in the US it only increased by 1.2%.

Wellbeing in the EU and the US

Source: Benczur et al. (2026)

Then they went a step further and broadened the wellbeing indicator to a total of 42 indicators, including those that cover resources for the future, economic resilience, inequality and quality of institutions.

This index of sustainable and inclusive wellbeing rose by 4% in the EU since 2010 but declined by 0.5% in the US.

Sustainable and inclusive wellbeing in the EU and the US

Source: Benczur et al. (2026)

So, let me finish by posting here a meme that has made the rounds this year that compares the UK (which clearly has a lot of issues) with the US on a range of economic and well-being metrics.

Enjoy August, everyone.

First, my case for 3% inflation targets rather than 2%.

In the US and Western Europe, we face two major problems over the coming decade or more. The first one is an ageing population and a declining workforce. This demographic headwind leads to increasing shortages of labour, which in turn creates wage inflation and pushes inflation rates structurally higher. According to my calculation, this lifts the average annual inflation rate by about 0.5 percentage points over the coming decade.

Then there is the end of globalisation. I am not talking about de-globalisation or the fragmentation or localisation of markets. Rather, I think that going forward, we will no longer reap the benefits of ever-increasing globalisation and outsourcing to cheaper producers in emerging markets. I have written here how this increases prices for European and USS consumers. According to my calculations, this adds another 0.5 percentage points in structural inflation.

Add these two drivers together, and structural inflation should be one percentage point higher in the next decade or more than it was in the last twenty or thirty years.

Before you panic about runaway inflation (I have many German readers, after all), let me assure you that 3% inflation is no problem at all for financial markets or businesses. Empirically, businesses and stock markets alike can deal with inflation anywhere between 0% and 4%. If inflation rises above 4% to 5%, businesses start to suffer because end demand declines (cost of living crisis) and margins get squeezed as businesses struggle to pass on rising input costs to end customers.

Economically and financially, there is no fundamental difference between a 2% and 3% inflation target.

But Kara makes the argument that there is, and that once inflation surpasses 1.9%, businesses change their behaviour and monetary policy loses some of its effectiveness.

He uses microdata of the price changes of some 340,000 products sold to consumers in the UK from 2003 to 2021. Then he asks a simple question. If you are running a business in the UK and you see that your competitors are increasing the prices for their products. Should you increase the prices for your goods as well?

If the prices are trending structurally higher, a business needs to increase its prices to avoid a margin squeeze. The longer you wait for your price increases, the more money you leave on the table because you are paying higher input costs while charging the same output prices.

But if price changes are only temporary and inflation pressures decline after a while, then you are better off keeping your prices low because your input costs will come back down again, and you can retain your margin while offering your products at a lower price than your competitor. By raising their prices, your competitors have effectively put themselves at a disadvantage the moment input costs drop again.

Kara shows empirically that once price increases surpass 1.9% in the UK, businesses tend to change their behaviour. If competitor prices increase just a little, businesses tend to ignore these prices and count them as a competitor reacting to firm-specific strategy decisions or firm-specific cost issues. Once price increases by competitors surpass the typical annual price increase (which for most industries in the UK happens to be around 1.9%), firms take the price changes of competitors as a signal that the cost pressures they see are a universal and persistent cost pressure. Hence, they decide to follow suit and raise prices themselves as well.

The result is that once inflation rates for individual products rise by more than 1.9%, this herd behaviour reinforces inflation trends. With every company that joins in the trend of raising prices, consumer price inflation accelerates a little bit, and you start to get an inflation spiral.

This behavioural change has two implications. First, there seems to be a ‘natural rate of inflation’ which in the UK is around 2%. If inflation remains above that level for too long, an economy becomes increasingly prone to self-reinforcing inflation loops.

Second, monetary policy becomes less effective once inflation rates surpass this natural rate of inflation. Because businesses are increasingly busy trying to monitor and emulate their competitors’ price actions, they pay less attention to monetary policy incentives, and hence changes in interest rates by the Bank of England have to become larger and more aggressive to break the inflation loop.

I find this an intriguing observation, but my counterargument is that it is based on the past behaviour of businesses during a period of low and steady inflation when the Bank of England focused on an inflation target of 2%. If the Bank were targeting 3% instead, would firms change their behaviour only once competitors start raising their prices by more than 3% rather than 2%? How did behaviour change during the recent inflationary episode, 2022 to 2023 (the analysis only uses data to 2021)? How did companies behave in the 1970s? I don’t know the answer, but it would be good if someone could check.

A study from the Singapore Management University examined the portfolio holdings of US hedge funds and compared how well these portfolios aligned with the economic policies of the incumbent US President. To do this, they measured the share price reaction of every US stock to economic policy announcements by the President at the time, which allowed them to measure the portfolio sensitivity of each hedge fund to these announcements.

The chart below shows the hedge fund alpha after correcting for the Fung and Hsieh hedge fund risk factors for portfolios sorted by political alignment. Hedge fund managers that are politically aligned with the President (measured by the political donations they make) also run portfolios that are more aligned with the President’s economic policies.

But the portfolios that are most aligned with the President’s economic policies underperform the least aligned portfolios by about 4.4% per year. In essence, these aligned portfolios are betting on something the market already knows, namely, the economic leanings and intentions of the President. So a lot of that information is already in the price. At the same time, stocks that do not benefit from the President’s economic policies tend to be shunned by investors, making the same mistake, and as a result have higher future returns.

Hedge fund portfolio alpha by political alignment

Source: Chen et al. (2026)

Of course, we don’t know if the explanation I just gave is the real cause or whether this study just measures correlation. But what the study does is look at unanticipated events that increase political polarisation, like mass shootings and political protests in cities where the hedge funds are based. And they find that after such events, hedge fund portfolios that were already aligned with the President’s policies align themselves even more and subsequently underperform even more.

How fast money is warping market logic

LONDON, July 27 (Reuters) - A tidal wave of "fast money" may be behind a series of seemingly irrational price swings in recent years, as fundamentals are seemingly being overwhelmed by the rush for the next big trade.

Check out the full article here: https://www.reuters.com/commentary/reuters-open-interest/global-markets-roicolumn-pix-graphics-2026-07-27/

Recently, I wrote a post about how the proliferation of AI will, in my view and according to some early evidence, not lead to widespread job losses. Instead, Jevons’ Paradox ensures that AI will be a net job creator. But that doesn’t mean that these new jobs will be created equally across industries and age groups. And while it is still early days (I emphasise this all the time because things are very much in flux), a new analysis by Lee Tucker from the US Census Bureau indicates that one group of people is losing in the job market: Recent university graduates.

He tracked the employment situation for US citizens across industries and age groups since the advent of ChatGPT in 2023.

My first chart today shows the change in employment among people aged 22 to 24 (i.e, people at the start of their career) since the end of 2022, sorted by the AI exposure of industries. The dark blue line indicates the 20% industries that are most disrupted by AI. These are obviously IT, but also finance, insurance and professional services (Law, HR, Accounting, etc.). Clearly, this is where employment has declined, while employment has grown in industries that are least affected by AI (Agriculture, construction, tourism, etc.).

Employment by AI exposure, ages 22-24

Source: Tucker (2026)

To check if this is a general trend across all age groups or just among young people starting out, we can look at the second chart, which shows the change in employment across age groups in those industries that are most disrupted by AI (i.e. the dark blue line in the previous chart). More experienced people did not see a decline in employment, while younger ones clearly did.

Employment by age in industries most disrupted by AI

Source: Tucker (2026)

We can even go a level deeper and look at the change in employment by individual industry for people aged 22 to 24. Again, the main culprits are the tech industry, financial services, and professional services. Agriculture also shows up, but that is likely just part of a secular decline in employment in this sector that has been going on for decades.

Change in employment for 22- to 24-year-olds by industry

Source: Tucker (2026)

One word of caution, though. Tucker also shows that the decline in employment may not only be due to the advent of AI alone. Some of these trends already started with the pandemic and were simply accelerated by AI. But at the very least, AI didn’t help young professionals trying to find a first job.

I mean, is this link between owning pets and being happier in life correlation or causation?

As an animal lover who is passionate about helping animals in the developing world, my answer is clear (and by the way, if you want to donate to help animals in St Vincent and the Grenadines, you can do so here; this is the charity that I am involved with).

But Michael Gmeiner and Adelina Gschwandtner tried to get causal evidence, not just correlation. To do this, they used a trick. They wanted to know if people are happier when they are asked to watch over their neighbours’ pets while they are travelling.

Unfortunately, there is no real-time survey of people and their happiness that also includes a question about whether they are watching their neighbour’s cat or dog. So, they did use a second-best option in the UK Household Longitudinal Survey that asked how often people are asked to watch over their neighbour’s property while controlling for the variables that measure social ties and closeness to neighbours in general. They show that this approach is correlated with pet ownership, though admittedly it is an imperfect measure (but this is a Friday post, so I won’t be too critical).

Their results are unequivocal. People who own a cat or a dog are clearly more satisfied with their lives, and their approach shows that this is causal. Once you get a pet, you do become happier. End of story – no more discussions.

Except one. They also tried to measure the monetary value of owning a pet. They asked people how much they are willing to pay for a cat or a dog and how much that would influence their happiness. This ‘willingness to pay’ put the value of owning a pet at about £70,000. And if you ask me, that is a low estimate.

In any case, if you can give them a good home, get yourself a pet. I promise it will make you happier, healthier, more successful, richer and in every way a better person. And to motivate you a little more, here is a picture of my two corgis for the weekend.

Marco Zanotti analysed the websites of US funds and asset management firms and the kind of technologies they employ to harvest customer data from each visit. These can be simple data points like the IP address (and hence location) of the visitor to A/B-testing different website designs, to more complex data sets and dynamic websites built based on the data collected from the visitor.

The idea is that if you dynamically adapt a website to the type of visitor looking at it, you can focus the information presented to the visitor on the elements that make the fund more attractive and create a nudge to contact the fund manager.

And clearly this seems to work, because in the 12 months after a fund manager introduced such data mining technologies on their website, net inflows in the funds increased by an average 1.5% of AUM, with most of these flows coming from retail investors.

Net inflows increase after fund managers introduce data analysis on their websites

Source: Zanotti (2025)

However, there is a caveat. Obviously, there is a diminishing return on investment. Installing more and more data analytics tools will bring less and less in additional flows. But more importantly, most of the inflows go to first movers in the industry. The chart below shows that the size of net inflows declines if a larger share of a fund’s competitors also uses data analytics tools on their website.

Effect of competition

Source: Zanotti (2025)

This indicates two things.

First, introducing data analytics tools and optimising the webpage to attract customers has, over time, moved from being a competitive advantage to a must-have just to keep up with the competition (and prevent losing ground to peers).

Second, we are in the middle of a new technology emerging that allows for new ways to analyse customer behaviour: AI. I think asset managers need to think about how they can use AI to analyse customer data and use that to attract investors.

The people at the Bank for International Settlements (BIS), famously called the ‘central bank of central banks’, are clearly concerned about this possibility. They investigated how rising central bank policy rates lead to rising default rates in loan markets. But crucially, they provide some useful rules of thumb for investors depending on the prevailing macro and interest rate environment before rate hikes began.

The headline result is that a one percentage point increase in central bank rates on average leads to a 0.1 percentage point increase in default rates for private market loans. That sounds like nothing, and I admit it isn’t large, but note that the current default rate for private market loans in the US is 1.3% and in Europe 1.4%.

However, the study also tests for different macro environments before the rate hikes begin, and they can increase the effect significantly.

Effect of interest rate hikes on loss rates in adverse economic conditions

Source: Fandl et al. (2026)

The figures above need some explanation. The blue line in each figure shows the change in loan default rates in reaction to a one percentage point increase in policy rates if none of the adverse conditions stated in the headline of the figure is present. The light blue area around these lines gives the 90% confidence interval. This is the reference benchmark against which the black and pink lines need to be assessed, and they show that rate hikes can easily be digested if the economy is doing great before the rate hikes begin.

The dashed black line shows the change in default rates if the stated adverse conditions are in place before the rate hikes begin, but monetary policy is loose. The pink line shows the change in default rates if the adverse conditions are in place and the monetary policy is tight beforehand.

Looking at the four figures, I would argue that we are currently entering a phase of potential rate hikes with an economy that has high debt levels and lots of loans outstanding that have floating rates, rather than fixed rates. Hence, the bottom right figure seems most appropriate for the current situation. In the US, we have the additional complication of high inflation rates, while in Europe, we face the challenge of an economic slowdown (but not too high inflation). Each of these additional adverse factors is likely to increase the default rate even more.

But even so, the figure shows that a one percentage point increase in policy rates (something that is currently priced in money markets for the Eurozone, but not the US) is likely to trigger a 0.3 percentage point increase in default rates in the first year. And given current default rates, that is the same as saying that default rates in private loan markets may rise by a third.

This study found that for every company, the news flow is predictable. That reflects that narratives drive share prices and markets overall. Narratives that dominate the headlines take the spotlight, and investors overreact to these stories. In contrast, narratives that don’t fit the narrative tend to be overlooked and incorporated into share prices only slowly. This effect is particularly strong in smaller, less liquid stocks, as you can imagine.

The new study looked at US stocks between 1996 and 2022 and used LLMs to analyse the news flow across multiple dimensions and how the share price reacted to it. Based on this analysis, the researchers then predicted the news flow for the coming month as a function of past news flow. Deviations from this predicted news flow are a news surprise.

Interestingly, these news surprises seem to get incorporated only very slowly into the share price because they have to overturn or at least modify the prevailing narrative. And this can take up to 18 months, according to this study. The chart below shows that the value-added from paying attention to news surprises that contradict the prevailing narrative is particularly high for smaller companies, but it remains substantial for large caps as well.

Performance of portfolios based on news shocks

Source: Didisheim et al. (2026)

Most of us won’t be able to run LLMs on a huge universe of stocks all the time and predict future news flows to identify unexpected news shocks. But what every investor can do is to focus on two things. First, understand the narrative that is driving a company’s share price or the market overall. In 2026, the dominant market narratives are the potential fallout from the Iran war and the AI boom.

Then look for news items that may change the narrative in the future because they are important for the market or a company, but contradict the current narrative or are at least unrelated to it. These news items will likely take longer to be incorporated into prices.

The chart below shows that these news surprises are increasingly coming from three areas: (i) corporate actions, deals, financings and restructurings, (ii) news about a company getting distressed or delisting (credit analysis can help here identify financial stress early), and (iii) changes to corporate guidance and outlook. Together, these three areas generate about 70% of news surprises that fly under the radar.

Which news gets incorporated slowly?

Source: Didisheim et al. (2026)

There clearly is a trend that younger people get married later, have children later and have fewer children. But I am pretty sure that lower sperm count in young men is not the reason for this trend. Rather, it is more likely the fact that younger people are financially less secure, thus have to live longer at home and take longer to have the economic means to start a family, particularly in countries where child care is increasingly expensive.

But are worries about climate change also to blame for low birth rates? Strangely, a group of researchers in Poland looked at British people and the determinants of having a first child.

The good news is that most factors they tested didn’t matter. Only the obvious ones did. Age and marital status are the only factors that have a large influence on the probability of having a first child. Nobody will be surprised to learn that – up to a certain point – people are more likely to have a child if they are a little older, and people are much more likely to have a child if they are married or live with a partner. Meanwhile, educational attainment or political orientation did not influence the decision to have children.

The only other factor they found that had a small effect on first births is concern about the environment, but only under certain constraints. People who profess to be more concerned about the environment and climate change are not, per se, less likely to have children. Similarly, people who get a lot of news about climate change in their news feed are no less likely to have children.

But people who are very concerned about climate change and get a lot of exposure to news about it are about 62% less likely to have a first child than people who are concerned about the environment but don’t get trapped in a doom loop of climate news.

And in the UK, it seems these people are Guardian readers. The charts below show the number of news stories on climate change in the left-leaning Guardian newspaper and the right-leaning Sun tabloid.

Number of news articles on climate change in the Guardian (left) and the Sun (right)

Source: Weychert et al. (2026)

The failure gap across a range of questions

Source: Eskreis-Winkler et al. (2026)

• Health: What percent of pregnancies end in miscarriage? Estimated: 19% - True: 21%

• Consumerism: What percent of online purchases get returned? Estimated: 22% - True: 30%

• Gender Equality: Globally, what percent of women have been victimized by violence? Estimated: 32% - True: 33%

• Cybersecurity: What percent of Americans have experienced identity theft? Estimated: 24% - True: 33%

• Medication: What percent of people who take Claritin experience relief from acute pain (relief is defined as at least a 50% reduction in pain over 4-6 hours)? Estimated: 31% - True: 36%

• Health 2: What percent of people will be diagnosed with cancer at some point in their life? Estimated: 33% - True: 40%

• Education: What percent of adults lack basic math skills? Estimated: 26% - True: 40%

• Diplomacy: In what percent of countries worldwide have U.S. diplomatic relations broken down? In other words, what percent of countries worldwide has the U.S. fought or invaded? Estimated: 23% - True: 43%

• Poverty: Globally, what percent of people earn under $5.50 per day? Estimated: 40% - True: 46%

• Education: What percent of college students at a standard college graduate on time? Estimated: 33% - True: 48%

• Education: Globally, what percent of high school age students are enrolled in school? Estimated: 37% - True: 48%

• Medication: What percent of people who take Aleve experience relief from acute pain (relief is defined as at least a 50% reduction in pain over 4-6 hours)? Estimated: 37% - True: 49%

• Sports: Consider the ___ , a team of professional hockey players in the National Hockey League, the premier hockey league in the world. What percent of games do you think this team won last season? Estimated: 44% - True: 50%

• Wellness: What percent of people with a mental health disorder do not seek treatment? Estimated: 41% - True: 50%

• Underwater Biodiversity: What percent of underwater life has gone extinct since 1970? Estimated: 31% - True: 50%

• Wellbeing: What percent of people will be diagnosed with a mental health disorder in their lifetime? Estimated: 45% - True: 50%

• Healthcare: What percent of hospital personnel in U.S. hospitals meet national standards for hand-hygiene compliance? Estimated: 28% - True: 50%

• Medication: What percent of people who take Advil experience relief from acute pain (relief is defined as at least a 50% reduction in pain over 4-6 hours)? Estimated: 29% - True: 52%

• Infrastructure: Globally, what percent of people have a working toilet in their home? Estimated: 36% - True: 60%

• Economy 1: What percent of restaurants close within the first year? Estimated: 56% - True: 60%

• Medication: What percent of people who take Aspirin experience relief from acute pain (relief is defined as at least a 50% reduction in pain over 4-6 hours)? Estimated: 39% - True: 62%

• Medication: What percent of people who take Tylenol experience relief from acute pain (relief is defined as at least a 50% reduction in pain over 4-6 hours)? Estimated: 37% - True: 62%

• Law Enforcement: In what percent of crimes do police not identify a suspect? Estimated: 54% - True: 65%

• Wildlife Biodiversity: What percent of wildlife has gone extinct since 1970? Estimated: 25% - True: 69%

• National Security: What percent of the time does the TSA fail to detect weapons or explosives when they are there? Estimated: 14% - True: 70%

• Health 1: What percent of Americans age 20 and older are overweight? Estimated: 56% - True: 74%

• Environment 1: Globally, what percent of people litter? Estimated: 49% - True: 75%

• Sanitation: Globally, what percent of waste water flows into the environment untreated? Estimated: 38% - True: 80%

• Environment 2: What percent of the world’s land area has been affected by climate change? Estimated: 63% - True: 80%

• Energy: What percent of the world’s energy comes from fossil fuels? Estimated: 68% - True: 80%

• Economy 2: What percent of new businesses started by entrepreneurs close? Estimated: 60% - True: 80%

• Finances: What percent of Americans are in debt? Estimated: 67% - True: 80%

• Education: What percent of college students at a community college graduate on time? Estimated: 36% - True: 80%

• Equality: What percent of developed countries offer the same full legal protections to women that they offer to men? Estimated: 49% - True: 82%

• Food: What percent of people in the world lack proper nutrition (for example, they fall below recommended levels of daily fruit and vegetable intake)? Estimated: 46% - True: 82%

• Economy: What percent of countries are considered developed nations? Estimated: 52% - True: 83%

• Pollution: Globally, what percent of city-dwellers are exposed to air pollution levels that drastically exceed the safety levels set by the World Health Organization? Estimated: 54% - True: 85%

• Human Rights: In what percent of countries today do human rights violations like modern slavery exist? Estimated: 21% - True: 85%

• Innovation 2: What percent of new drugs fail (i.e., never get FDA approval)? Estimated: 58% - True: 85%

• Relationships: What percent of romantic relationships end in break-ups? Estimated: 65% - True: 85%

• Innovation 1: What percent of new patents are rejected? Estimated: 64% - True: 86%

Odd numbers convey information, and typically, it is not the best kind of information. Market makers know that orders with odd numbers of stocks or odd volume (e.g. buy 11 stocks of company X or sell $4,650 of shares Y) tend to come from less sophisticated investors. And I can assure you they would never try to profit off these investors by changing their bid or ask price. That never happens because market makers are kind-hearted people who want to foster equity market participation across the public rather than maximise the profits they are making for their employer in the short term.

Similarly, lenders don’t adjust the interest rate of a consumer loan depending on whether the borrower asks for a round-number loan or an odd amount. That one is true, though a study by Xiaowen Hu and Andreas Kraft shows that they would be better off if they adjusted their loan interest depending on the amount borrowed.

The study used anonymised data from Lending Club, the peer-to-peer consumer lending platform, to examine whether the loan amount tells you something about the credit risk of the borrower. And indeed, people who borrowed odd amounts were more likely to default on the loan than people who asked for round amounts. This was true across time, across consumer credit scores and even across loan amounts. People who borrowed amounts that were multiples of 10, 100, 100, 1,000, etc., were less likely to default on their loans.

Default rate by loan amount

Source: Hu and Kraft (2026)

What is going on there? Well, think back to the guy trying to auction off 11 bottles of wine. What likely happened is that this person bought a case of 12 bottles, opened one, found it was off and then tried to get rid of the other 11.

In the case of odd lot trading by retail investors, they probably buy a few shares or an odd volume because that is all the money they have available. They are trying to max out their investment portfolio and inadvertently give away the limitations of their means.

And when applying for a loan, people who are more desperate to get a loan will try to maximise the loan amount while staying within the budget they can afford to pay on the loan every month. But of course, if they max out the monthly payment to the absolute limit they can afford, even a small unforeseen emergency can mean that they no longer can afford their monthly repayments. And hence, they may have to default on the loan.

In short, applying for an odd amount is a sign that the person is already in some distress, even if it doesn’t show up in the credit score or other indicators of credit risk.

For those who don’t know Jevons’ Paradox, let me quickly recap. During the Industrial Revolution, when steam engines became ubiquitous and increased the productivity of manufacturing, many feared that increased productivity would lead to lower employment. After all, if a man with a machine can make twice as many shoes as a man making them by hand, you can theoretically reduce employment by half without reducing output. In practice, the opposite happened. As steam engines and machines proliferated, employment in the manufacturing sector soared.

The relationship that higher productivity leads to increased employment can be observed time and again. The introduction of the cash machine (ATM) did not reduce the number of bank tellers. Rather, their number grew, but instead of simply handing out cash, they now offered advice to customers.

And when it comes to AI, Microsoft CEO Satya Nadella argues that Jevons’ Paradox holds as well and will lead to higher employment in software engineering and other areas penetrated by AI.

The basic idea is that as productivity increases, coding becomes easier to do and thus people will do more of it, which ultimately creates more jobs.

That’s as far as theory goes, but what does the data say? It is early days, but in recent years, we have seen surveys that indicate that the proliferation of AI has so far not cost any jobs. But recently, I have come across another study that looks at actual employment data in the US between 2017 and 2024. It then combines this employment data with output and productivity data from the GDP data releases and wage data from the American Community Survey.

Hence, it is a top-down estimate of the macroeconomic impact of AI proliferation, starting with industrial enterprise AI in 2021 to the end of 2024. And this study indicates that Jevons’ Paradox may really hold in the case of AI.

First, the proliferation of AI in industrial applications starting in 2021 and later the proliferation of generative AI, starting with the launch of ChatGPT in 2023, has likely increased productivity. The study finds that a one standard deviation increase in AI exposure in any business increased output by 7% between 2020 and 2024.

To measure job exposure to AI, they used the index created by OpenAI and the Centre for the Governance of AI at Oxford University, which is really just a number that is hard to explain. For what it’s worth, a one standard deviation increase is roughly the equivalent of the difference in AI exposure between a worker at the US Postal Service and a worker at an online retailer.

Second, the study finds that industries across the US that had higher AI exposure also saw employment and real wages rise. A one standard deviation increase in AI exposure increased employment by 3.9% and the wage bills of affected employers by 4.8%. These numbers are smaller than the growth in output, indicating that not all productivity gains drop through to higher employment or higher real wages, but at the very least, there is statistically significant job growth.

Again, it is early days, and new data is becoming available all the time, so we need to take these results with some caution, but they are encouraging nevertheless.

Impact of AI exposure on worker earnings (top) and employment (bottom)

Source: Johnston and Makridis (2026).

Why NATO should ask Ukraine for help

LONDON, July 14 (Reuters) - Western governments want to rearm, but they may be spending billions rebuilding the wrong arsenals. While tanks and missiles are still necessary, so are drones and algorithms. Ukraine may be well positioned to show the West what’s needed today.

Read the full piece here: https://www.reuters.com/commentary/reuters-open-interest/why-nato-should-ask-ukraine-help-joachim-klement-2026-07-14/

A series of surveys across 11 EU member states showed that between 2020 and 2024, about 10% of non-investors invested in the stock market for the first time each year. This is quite a change of pace for a continent where stock ownership is famously low. We don’t have the same investment culture as investors in the US or even the UK.

I have mentioned in this note on the investment performance of Germans that this reluctance to take investment risks contributes to poor wealth accumulation. The result is, for example, that despite having higher savings rates than other nations, Germans don’t necessarily have a larger retirement nest egg at old age.

The study by researchers from Italy shows that by November 2024, more than 50% of German households were investing in stocks or stock funds, up from about 40% in the middle of 2020. Across the Eurozone, participation rates rose from about 30% to 40% in the same period.

Stock market participation rates in the EU

Source: Christelis et al. (2026)

While many investors came to the stock market during the pandemic, they also left the stock market again once the pandemic ended. The percentage of investors leaving the stock markets is about 20% of the existing investor base, with spikes in the last winter of the pandemic and the onset of the inflation spike in 2022. But while many investors left the stock market after the pandemic ended, many stayed around, as the rising participation rates into 2024 show.

Share of stock market investors exiting the market

Source: Christelis et al. (2026)

PS: The study also gives some insights into crypto ownership in the EU. And the two figures below show that there still is no crypto boom in Europe. A constant share of about 10% of EU households invest in cryptocurrencies, with many joining during the pandemic, but new joiners slipping away once the pandemic ended.

Household participation in crypto assets

Source: Christelis et al. (2026)

To give you an example, let’s look at a study by the National Bank of Belgium on the impact of the EU’s Sustainable Finance Disclosure Regulation (SFDR) on fund flows of conventional and ESG funds.

The SFDR was announced on 27 November 2019 and came into effect on 10 March 2021. It was implemented in several stages and split funds marketed in the EU into three broad categories: (i) Art. 6 funds that are basically conventional funds without any specific ESG integration, (ii) Art. 8 funds that ‘promote environmental or social characteristics’, which must provide additional reporting how they do that, and (iii) Art. 9 funds that are impact funds.

The gaps between the announcement and the phased implementation give us an idea of how fund flows were affected by the regulation. It is clearly visible in the figure below that once the new regulation was announced and when it was initially implemented, fund flows in Art. 8 and Art. 9 funds increased significantly. Conventional Art. 6 funds also saw more inflows, but the increase was not statistically significant.

I should add that the data below is pure fund flow, cleaned of performance effects and of any reclassified funds that moved from one classification to another.

Note also that once the initial rush in ESG funds was over, additional, stricter regulatory requirements didn’t create that many new flows. This implies that with additional regulation, the funds incurred higher costs to comply but didn’t benefit in terms of new inflows and higher assets under management.

Fund flows as the SFDR regulation was implemented in the EU

Source: Emiris et al. (2026)

When examining the drivers of these flows, the authors of the study found that institutional investors were more likely to switch to ESG funds and particularly so in countries where concerns about sustainability are culturally more accepted. In a society where buy-in into environmental and social risks is higher, private investors were shifting more of their assets into ESG funds as the regulation made it clear which funds are really green and which ones are not. And institutional investors in these countries did so as well, given the societal pressures and reputational risks of sticking with conventional funds.

Meanwhile, fund managers reacted to the changing regulatory environment as well. Many funds that were mildly green before the regulation was announced became greener and more sustainable as the regulation was introduced to attract new assets and new investors. Cleverly, and unsurprisingly, these ESG funds also charge on average higher fees, which means that not only did fund managers gather fresh assets, but their profitability increased as well.

Yet, many people, particularly in the US and Europe, will also have a somewhat more positive view of Saudi Arabia than of Iran. As someone who has been in Saudi Arabia several times, I can assure you that a slightly more positive view is thoroughly undeserved. As far as I am concerned, there is no difference between Iran and Saudi Arabia except that Saudi Arabia is a long-standing ally of the US and Western Europe and invests billions in sportswashing.

But that is the point. Positive news coverage is easier to get if your country is more closely aligned with the country that is reporting on the news. In a beautiful study, Sibo Liu and his collaborators analysed 1.4 billion news articles from 55,240 media outlets in 176 countries.

The core result is shown graphically in the chart below. The closer the institutional similarity between the country that reports and the country that is reported on, the more positive the news coverage on average.

Institutional similarity and news coverage

Source: Liu et al. (2026)

Perhaps unsurprisingly, news coverage of political adversaries is particularly negative in autocratic countries and in both Autocratic and democratic countries among state-controlled media outlets. The latter also has an important corollary. In a country that experiences democratic backsliding (e.g., Poland and Hungary during the 2010s), the news coverage of other democratic countries is rapidly turning more negative, particularly among state-controlled media outlets.

And of course, as we get more and more bad news from specific countries, public opinion about these countries deteriorates as well, which in turn gives politicians the opportunity to act more hostile towards these countries.

As Austrian writer Karl Kraus famously said: “How is the world ruled and led to war? Diplomats lie to journalists and believe these lies when they see them in print.” He was pretty close to the truth, it seems.

As European regulators have tightened their screws on banks and demand higher capital backing for their loans, banks have increasingly turned to SRT to free up existing Tier 1 capital. The idea of SRT is that if a bank sells first losses in a senior loan tranche to an unrelated entity, the capital requirement for these senior loans is reduced by 85% to 90%. This capital can then be used to back additional loans to other borrowers. Technically, the loans remain on the bank’s balance sheet, and the bank will still have to hold some capital against these loans, but if these loans default, the first losses are borne by the outside investor who bought the SRT. Only once a specific loss threshold is reached does the bank have to recognise losses on these loans.

For banks, these SRT are obviously an attractive proposition. For investors who buy these SRT, they can be lucrative as well since they receive not only the loan repayments but also an additional risk premium. The result is that SRTs have grown rapidly since 2019.

According to an analysis by the ECB, the main investors in these SRTs are government entities and the European Investment Bank (EIB). Still, in recent years, pension funds, insurance companies and specialised investment funds have also joined the party.

SRT holdings by investor type

Source: ECB (2026)

Anyone who is old enough to remember the financial crisis of 2008 will remember the risk transfer via CLOs that allowed banks to package loans and mortgages and sell them on to other investors who supposedly were better able to bear that risk. But when these loans incurred losses, it turned out that everything came right back to the banks to haunt them.

If there is one saving grace about SRT vs CLOs, it is that banks that use SRT still have to hold some capital against these loans, even if they sell first losses on to other investors. But the ECB studied the market and found some concerning tendencies:

• Banks use SRT more to sell off riskier loans that require more capital backing. Hence, the incentives for banks are such that banks hold on to the safest loans themselves and sell the riskiest loans to outside investors.

• Once loan risks have been transferred by banks via an SRT, banks reduce the monitoring efforts on these loans. After all, if these loans default, the bank no longer suffers losses. Investors in SRT thus end up with the riskiest loans that are also monitored less diligently.

• Finally, the investors in these SRT themselves are linked to the banks. Banks are 57% to 66% more likely to sell an SRT to investors with which the bank has a lending relationship. The ECB estimates that via these loans, banks fund about 29% of SRT volume themselves by extending loans to investors who then use that money to buy SRT.

I mean, what could possibly go wrong?

The goal of these new trade deals is to reduce the negative impact of the US tariffs. On the face of it, this seems a very difficult task indeed, since the US is the largest economy in the world and its largest export destination. Yet, in the EU, officials have coined the term ‘P7’ to denote the seven free trade deals it wants to accelerate to offset lost trade with the US: the members of the Mercosur trade zone in Latin America, India, Australia, the UAE, Indonesia, Malaysia, and Thailand.

Together, the P7 currently account for roughly 10% of EU exports, still only about half the export share of the US. But the difference between trade with the US and trade with the P7 is that tariffs between the EU and the US had already been close to zero, and trade was highly integrated. With most of the P7 nations, tariffs are currently high, and trade relationships are not very deep. Hence, the upside to trade deals with the P7 is much larger than the loss of trade with the US.

Andreas Baur and his colleagues have simulated the potential impact of trade deals with the P& for the EU in the light of the Trump tariffs under Section 122 (i.e. after the Supreme Court ruling IEEPA tariffs illegal).

They find that the Trump tariffs, as they currently stand, may cost the EU about 1% of its annual exports. The manufacturing sector is especially hard hit, losing some 1.3% of its output. The result is an estimated drop in EU GDP due to the US tariffs of 0.1%, as I said, not very big, but concentrated in specific manufacturing industries.

If these US tariffs remain in place and the EU were to close free trade agreements with the P7 that eliminate tariffs in manufacturing, mining, and agriculture but do not change existing non-tariff barriers like minimum quality standards, etc. (so-called ‘shallow integration’) then EU exports are expected to rise 1.3% and manufacturing output in the EU is expected to rise by 0.1% despite the drag from the US. EU GDP in this scenario is expected to get a boost of 0.2% while the GDP of the P7 is expected to rise by 0.4%. This is why free trade is a win-win.

One can even go a step further and reduce non-tariff barriers to a minimum (so-called ‘deep integration’), in which case EU GDP would get a boost of 0.4% and the P& get a growth boost of 1.0%.

Thus, the US tariffs could become a blessing in disguise for Europe and developing countries alike. Because these countries have to find trade partners other than the US, they are accelerating free trade between them and, as a result, gain more from these new trade deals than they lose in trade with the US. The only loser in this world will be the US, which sees its GDP shrink due to higher tariffs.

Impact of US tariffs combined with P7 trade deals

Source: Baur et al. (2026)

Momentum crash warnings flash red. Here’s how you might survive

LONDON, July 6 (Reuters) - As the AI frenzy creates new market cap trillionaires and worldwide stock performance continues to be driven by tech winners, investors have developed a severe case of FOMO, leading many to plough more funds into equity leaders. But momentum booms tend to be followed by busts, and the current ​one is looking dangerously stretched.

Is there a way to avoid getting run over without abandoning the momentum train? A new study suggests yes.

Read the full piece here: https://www.reuters.com/commentary/reuters-open-interest/momentum-crash-warnings-flash-red-heres-how-you-might-survive-2026-07-06/