Small Modular Reactors will unlock data centers and other projects that are currently not feasible

Like Nuclear-Powered Submarines – Small Modular Reactors (SMRs) are not new – we have had small nuclear power generators for decades — most famously in nuclear-powered submarines and icebreakers.

Global Race – Companies and governments around the world are now moving to use SMRs for commercial electricity generation.

Mass Produced in Factories – SMRs are intended to be produced on factory production lines (like cars). This should make them far cheaper and faster to get running than traditional, on-site-built nuclear power plants.

Smaller – SMRs are smaller – an SMR might provide between 5 and 500 MW of electricity, while a traditional nuclear power plant might provide between 1 and 8 GW of electricity.

Leading Countries – Only China and Russia currently have operational SMRs. Other leaders in developing SMRs include the US, the UK and Germany.

Regulation as the Key Barrier – Licensing and permissions are likely to be a key barrier to rolling out SMRs. Regulation is important for safety – as any accident could turn public perception strongly against SMRs and stop their roll out (like the Three Mile Island accident in 1979 stopped the growth of traditional nuclear power plants).

Google and Amazon SMR Deals – Google and Amazon have in the last few months signed agreements to buy SMR-generated energy with SMR developers. These agreements should help the SMR developers to commercialize their products. The first energy to be produced by SMRs under these deals is expected to be in the early 2030s.

Timing – SMRs will continue to slowly grow over the next decade. We expect the first SMRs in the US, the UK and Europe in the early 2030s. With commercially relevant numbers of SMRs in place by the late 2030s.

Unlocks More Datacenters and Other Projects – Because of their size and cost, SMRS can be deployed in locations that were not feasible for other power generators (nuclear and other). Because of this, SMRs could allow large numbers of datacenters to be built around the world in locations that are currently not possible because of power constraints. SMRs could also make possible a very large number and wide range of projects – from manufacturing facilities to new types of residential area.

Quantum computing – when should companies start paying attention

The answer is – companies should start to take action for long term strategy and investments now. There are parallels with AI – companies that “front-ran” AI are big winners now. These include NVIDIA, Alphabet, Microsoft, and Tesla.

Timing

Our estimate of widespread quantum computing use is around 2035. There will be some use before then.

Given how big the impact of quantum computing will be – large companies should start acting now to protect their long-term assets, and to build assets that will become valuable when quantum computing is applied. It’s a case of “skate to where the puck is going”.

Quantum computers are like planes. Classical computers are like cars.

Quantum computers are a new type of computers. An analogy is that classical computers are like cars, and quantum computers are like planes.

Planes allow you to get from place to place much faster than a car. But there are still many things for which a car is still the much better choice than a plane.

You do not need to understand how quantum computing works

Business people don’t need to understand exactly how quantum computers work – in the same way that very few business leaders understand all the processes in a classical computer. It is the same as you not needing to know exactly how a car or television works to use it.

What will quantum computing change?

The key strength of quantum computers is that they can model very large numbers of scenarios at the same time. A classical computer, on the other hand, has to go through them one-by-one.

This is a good rule-of-thumb for working out what quantum computers will change.

A few examples of important things that quantum computers will dramatically change that follow from this:

Pharmaceuticals – finding new drugs. Molecule-level modelling (of the drug and the body) that is too complex for current computers will mean a step-change in medicine. This will have major implications – including on longevity, on the value of drug patents, and on global demographics/geopolitics.
Materials – finding new materials. Quantum computing will allow us to create a large range of materials that unlock things that we can’t currently do. This will include materials that make better large-capacity batteries, and better fertilizers that reduce food costs.
Financial Markets – modelling complex systems. These will create opportunities for firms that are ahead of the curve in quantum computing – potentially creating the Renaissance Technologies of the 21st century. At a broader level, this should mean more efficient capital markets – unlocking global growth.
Manufacturing and logistics at scale – modelling complex systems. This allows companies to operate efficiently at a larger scale (where currently things get too complex and start to suffer from inefficiencies). This could lead to ultra-large companies in sectors like manufacturing, shipping, airlines, retail.
Artificial intelligence and robotics – quantum computing could make training AI models far more efficient. This could lead to much more intelligent AI, and AI for much more specific use cases trained on proprietary data sets. An important use of this more intelligent AI is likely to be in robotics.
Pure science – quantum computing will allow us to make new and in many cases unexpected discoveries in basic science. This will result in big but difficult to imagine changes. Thinking about the changes in the 20th century that came about because of breakthroughs in basic science – from satellites to antibiotics to nuclear power.

What companies should do now

Large companies should start to develop a quantum computing strategy. This should look at what will happen to the company’s existing long-term assets once quantum computing is adopted, and what new assets the company can build/acquire that will become valuable in a world with quantum computing.

Positioning for a 1980s /1990s-style global economic boom from upcoming financial markets deregulation

Financial markets regulation has increased dramatically since the 2008 crisis. This increase in regulation has been particularly strong in the US and Europe.

Higher regulation has been made up of regulation on banks (higher capital and liquidity requirements, pay restrictions to reduce employee incentives, etc.), investors (limits on investors from asset managers to pension funds to insurance companies), traders (including limits on short selling), companies (rules on borrowing like risk retention rules for securitization issuers), and individuals (like new rules on individual borrowing limits).

Some of these rules have reduced some risks – but many of them have also prevented productive economy activity. Some of these rules have also created unexpected second-order effects like the development of a large private credit industry which have other risks.

The Trump administration has made it clear that it plans to deregulate banking. The US government has spoken about reducing bank regulatory capital, reducing bank liquidity requirements, and making bank stress tests less difficult to pass.

All this could unlock hundreds of billions of dollars of regulatory capital – which could turn into trillions of dollars of new debt in the US. This new money could flow into investments by companies, M&A, government, financial assets, and consumer spending.

Given current geopolitics, and the socially stressed situations of many countries around the world (developed and developing) – we expect that US deregulation will be quickly followed by deregulation in other western markets. We also expect that geopolitics might also extend this quick deregulation to other markets including China.

For companies, the upshot is to position yourself for the major economic boom that could come about if this scenario comes about. In this case, companies who have the production and distribution capacity in place (including capital, technology and people) will be years ahead of competitors. There are cases that could mean this does not happen – but there seems to be enough of a probability of this now that it is now worth taking actions to be positioned for this state of the world.

AI will result in much more software. What this will mean for companies

AI is going to result in much more software. From working through the implications, we think that fears that software proliferation because of AI will be bad for most companies are wrong – our conclusion is that it will be very positive for the average company in terms of profits and valuation. What is needed though is for almost all existing companies to reassess the new “rules of the game” as this effect on software will change almost every industry – and adapt their strategy to continue creating value in this new, more productive world.

In this short note, we first think about software and its general impact on business (looking at the last 30 years), we then think about how AI will dramatically increase the amount of software in the world over the next five years, we then think about the effects that all this extra software will have on the world and on companies, and finally we think about what companies need to do now in response to this fundamental change.

Thinking about software and its general impact on business (looking at the last 30 years)

Software has changed the world over the last 30 years.

It has changed how people work, how humans communicate with each other, how democracies work, and the policies governments implement.

This has had big implications for companies. Many of the world’s biggest companies 30 years ago no longer exist, and many of the world’s biggest companies now did not exist 30 years ago.

Software has increased standards of living. It has driven GDP increases in developed and developing markets – and has created a lot of value that is not reflected in GDP (things like the benefit of having a smartphone or fewer deaths from healthcare improvements are not fully reflected in GDP measures).

The big constraint has been software developers – even simple software has been expensive to create – because there are a limited number of people who can write code. Software developers have been among the best-paid professionals in society because of this high value that their software creates.

This has meant that we have had far less software than we would ideally have if we were not limited by the number of software developers that we have.

How AI will dramatically increase the amount of software in the world over the next five years

AI has dramatically reduced this societal constraint – it has reduced the number of software developers we need to write each piece of code.

AI can already produce code. Companies like Google say that over 25% of their code is now written by AI.

At the moment though, AI-produced code is still highly imperfect – it is still full of errors. But this is improving fast. These improvements are partly coming because LLMs are improving, but they are also coming because of other software that is being created to work with LLMs to write code together. These include software tools which get LLMs to write lots of possible versions of code for a prompt – and then run all of those versions in a “walled-off computer for testing” to see which ones work, which ones work fastest, and which ones can’t be hacked.

As LLMs improve and as these bigger systems that work with LLMs to create code get developed, we will be able to produce much more code.

This is not a new phenomenon for us – it is similar to the entire history of programming. We went from very basic programming languages like punch card patterns, to assembly languages, to early high-level languages like FORTRAN – all the way to modern high-level languages supported with large numbers of frameworks and libraries. The outcome of these improvements in programming languages and the outcome of AI-written code generation is likely to be the same – more software created per programmer. Today, one programmer can maybe produce as much software (in terms of what it can do) as perhaps 100 programmers 30 years ago because of programming language advances and advances in the computing environment in which their programs run. AI-generated code means that today, one programmer can produce as much software as perhaps four programmers a year ago. In five years from now, one programmer might be able to produce as much software as 20 programmers today.

This is not likely to mean that in five years, we are creating a similar amount of software to today – just employing far fewer developers.

We expect that the result will be that we have much more software produced. We expect that the total number of software developers will also significantly increase – as we have seen historically as output per programmer increases. The nature of what a “software developer” is will change – as it has over the last 30 years. The average developer will be working at a higher level of abstraction than developers today. There is a chance that we are wrong, and that AI systems will actually be able to fully design and create all software alone in this time. But this seems unlikely – at least in the next decade. In this scenario, the rate of software growth would be even faster.

First-order, second-order, and higher-order effects

We can break down the effects we expect into first-, second-, and higher-order effects.

First-order effects

We expect the first-order effects to be that much more software produced. This will be both custom software produced in-house and third-party software. In-house software will replace some software that companies currently buy. Large software companies will expand the features of their software – creating much more value for buyers. Some software services will be protected by non-software assets attached to them – like the data of Bloomberg, or the network effects of Microsoft’s Office file formats. New software companies will be created that solve new, often more specific problems. Existing large software firms that adapt will often have competitive advantages that help them continue to succeed in this new environment. New firms will replace some existing large software firms that do not adapt well to this new paradigm.

Second-order effects

More software will mean more output. Standards of living will increase. The value of companies’ assets (customer bases, brands, assets, etc.) will rise in general. Some industries will get highly disrupted – but in general for companies that are proactive in adapting, we expect the future to be increasingly profitable.

Higher-order effects

More output will increase standards of living – in developed and developing countries. This will affect social, geopolitical, political, and regulatory systems. There are a lot of wildcard outcomes that could occur because of how powerful software can be – including scientific breakthroughs in energy and longevity.

What companies need to do now in response to this fundamental change

The key takeaway for all companies is to adapt. The future should be better on average for companies – but the rules of the game have changed so companies need to adapt to those new rules. Companies need to look at their products – and work out if the effects of much more software in the world will mean they need to change their products. Companies also need to look at their operations – to work out how to make best use of the new capabilities available to companies. This includes finding better ways of getting new customers and delivering value to customers. They also need to find new ways to increase efficiencies that are unlocked by this new lower-software development price structure (and new capabilities of software) – so they can stay competitive on price.

Understanding what China’s new Five-Year Plan will mean for the world

China is finalizing its next Five-Year Plan – for 2026 to 2030. This decides what China will do and become in the medium term. As the world’s second largest economy, and one of the world’s most important technological and military actors, this is important for governments and companies around the world.

Putting Five Year Plans in context – At the top level, China has broadly-defined long term goals. These include: i) Vision 2035 – which covers things like becoming an innovation-driven economy and national defense goals, and ii) the 2049 Centenary Goal – which marks the People’s Republic of China’s 100th anniversary and is to “build a modern socialist country” that is “prosperous, strong, democratic (China’s format for democracy), culturally advanced, and harmonious”. The Five-Year Plans help achieve the country’s long-term goals.

Timing of the upcoming Five-Year Plan – studies and discussions began in 2024, the Government’s Central Committee is currently preparing the plans, the plan is expected to be formally published in March 2026.

What we expect the upcoming Five-Year Plan to contain (largely based on what has been spoken about by officials and state press recently):

Leadership in pure science and in technology – to become a leader in scientific research, breakthroughs in fundamental science, and the application of science in society and business by being a leader in technology development and adoption. Focus areas are likely to include advanced computing chips, AI, robotics, advanced manufacturing, aerospace, space, quantum computing, biotechnology, new energy, energy storage, neuroscience, 6G telecommunications, carbon capture, and advanced materials.
Moving up the value curve – to shift manufacturing to increasingly higher-value advanced manufacturing, to continue the rapid growth of the services sector as a proportion of total GDP.
Domestic demand – increasing domestic consumption to increase economic activity that is not reliant on exports and property.
International influence – increasing international foreign investment and diplomacy.
Green transition – continue the move to renewable energy production and electrification (e.g. EVs).
Demographics public policy – more investment in health and eldercare. Policies designed to increase birth rates and the working age population.

From a geopolitical perspective, this is likely to mean changes in global influence, alignment, and trade policies – which will all affect companies’ trade and operations – not just within China but elsewhere in the world. Demand for international goods and services from China might increase as the country increases consumer consumption. Some goods and services will face more competition as Chinese companies start to compete in those markets. There could be more supply of high-value goods and services from China to the rest of the world – which could increase standards of living. Competition for some of some goods and services is likely to increase. These types of new supply will unlock new market opportunities. We should see China become a bigger contributor of pure science and new technology to the world.

Companies should position themselves to benefit from new opportunities that the implementation of the next Five-Year Plan creates, and should act to change their holdings and strategies that will be disrupted.

Recent defaults in the private credit market and the risk of a 2008-type crisis

We have had two large defaults in the private credit market in the last few weeks – First Brands and Tricolor. Both might be at least partly down to fraud. The private credit market has grown exponentially over the last 15 years – to now around $3 trillion of assets globally. The 2007/2008 crisis started with defaults from fraud – so it is worth looking into whether we might see something similar happen now.

In some scenarios, these recent defaults might mean nothing – and private credit will continue its trajectory of growing rapidly (and supporting economic growth). But in another scenario, this could be the start of a 2007/2008-like credit markets crisis – which then turns into an overall economic crisis.

Some possible scenarios:

Isolated defaults, no problem. These are just two isolated defaults. Some defaults are expected – as with any kind of lending.
A few more defaults, no problem. There will be a few more companies that have large private credit borrowings that default. These will be managed, some losses will be booked, but private credit will continue fine.
Many more defaults, private credit problems, but economy fine. The private credit markets experience a flight – with end investors (pension funds, insurance companies, high net worth individuals, banks, hedge funds, etc.) liquidating their private credit positions. There could be a government response – increasing regulation on private credit lenders (as we saw with structured credit in 2008). Banks would also likely suffer losses – with banks also being lenders to many private credit borrowers – as is the case for First Brands and Tricolor. The economy is fine – there is a reduction in credit available from private credit lenders but banks and public bond markets are able to cover most of this borrowing need.
Major economic disruption. Private credit contracts quickly, and other lenders are not able to take their place – so money in the real economy materially contracts. It gets much more difficult for companies to borrow and stock markets fall. This then results in lower total spending in the economy – resulting in lower GDP and higher unemployment. There would likely be a quick monetary policy response which could reduce the size of economic disruption.

We do not know how things will play out. We expect things will be fine (scenarios 1 or 2), but think that there is a non-zero risk of scenario 3 or 4 – and so companies should take steps now to make sure they are not vulnerable in these cases.

Securitization as a solution for our $2 trillion+ data center build out

We expect total global data-center capex over the next three years to exceed $2 trillion, with 2025 spend projected at more than $500 billion. Build-out is accelerating as demand from AI applications and the large amount of new software that AI is helping to write, grows.

Securitization – issuing bonds backed by data-center assets and cash flows – began appearing in the sector in 2018. It has expanded quickly and is set to become a major source of large-scale, long-term financing for data centers.

In the 2000s, data centers were funded through the balance sheets of data center owners (large technology companies, specialist data center operators, government and telecommunication companies). In the 2010s, loans secured on data centers became more common – adding bank and private credit fund assets to the mix. And since 2018, securitization has brought pension fund, insurance company, fixed-income fund, and sovereign wealth fund assets into data center funding at scale.

As the number of data centers around the world grows rapidly to meet new demand from AI, we will need to intelligently apply the large, long-term sources of capital that come to this sector at scale through securitization.

Securitization will fund new data centers of hyperscalers (AWS, Google, Microsoft, Oracle, IBM, etc.), of specialist operators (Vantage, Digital Realty, Equinix, etc.), and new entrants.

Using AI agents as investment committee members

Investors including Singapore’s GIC, Schroders and Mubadala have announced that they are starting to use AI agents in their investment committees.

AI agents can be a useful tool in identifying risks and spotting risk mitigants – that give an investment committee an advantage over the competition.

They can ingest far more deal data than any human committee member and can spot subtle relationships across many dimensions. You might get an extra edge if you have proprietary data about your previous investments in an asset class that only you can feel into an AI agent.

Problems include that AI agents frequently make mistaken conclusions – especially on novel situations. There are also regulatory considerations depending on the type of investor.

Overall, at this stage, AI agents can be valuable voices on an investment committee – particularly as voices to provide contrarian perspectives that can be discussed and surfacing information that would otherwise have been missed.

How fintech lenders can use rating agency criteria to reduce funding costs

Aligning your asset origination risk criteria with rating agency criteria can be a key competitive advantage for fintech firms.

Why

The ultimate financing for most fintech-originated loans and other other assets is securitization transactions – in the form of public asset-backed bond deals or private credit deals. On the road to this long term financing, many fintech firms using “warehouse facilities” where a bank gives the fintech a funding facility they can use until they have a large enough portfolio (say $200m) to securitize.

Investors use ratings provided by rating agencies to help decide whether to buy securitized bonds and at what price. Rating agency securitization criteria provide the methodology that rating agencies use to assess the risk of asset pools that will be securitized. There include asset level and portfolio level assessments of the loans or other assets in the pool. These assessments then determine the ratings that rating agencies provide.

Most fintech firms start by originating assets based on their own judgment of asset quality. The problem with this is that it will not be aligned with rating agency criteria. This will result in a lower rating on the assets than could have been achieved. This directly translates to a higher cost of funding those assets for the fintech company.

How

The rating criteria from Moody’s, S&P, Fitch and other rating agencies is publicly available.

By studying the criteria and putting together a simple excel or python model, you are able to estimate how changing what your projected pool of assets looks like will change the ratings you get on your securitization of those assets. From that rating, you can estimate the cost of financing that pool.

You can use this model to work out what asset pool will be best for you – based on ratings as well as demand from your customers, market rates on those assets, and areas in which you think the rating agencies’ view on risk is wrong.

Outcome

You end up with a pool of assets that you can fund at a lower rate than your competitors.

As fintech markets get increasingly competitive, competing on the customer side of the business is getting more difficult.

But fintech firms often put much more effort into product-market fit and customer marketing, than they do into product-financing fit and investor marketing. This can give fintech firms that can figure this out a major competitive advantage. It can create network effects as lower funding costs allow fintech firms to originate the best assets – which in turn provide lower funding costs.

AI has made custom software development much cheaper. What should you do?

You need to reassess your entire technology plan. AI has made software development far cheaper – which means you should be thinking about replacing the existing software you use (internally developed and external), and you should be looking at if you should be changing entire parts of your business to use custom software.

There is upside – in that cheaper technology can improve your profitability. But there is also a threat – in that if your competitors take this opportunity to build technology superiority, you will be left at a competitive disadvantage. There is an investment lag between starting to build new systems, to implementing them, to seeing the effects – so a wait-and-see approach can be dangerous – as you could find yourself one to three years behind your competitors. We saw this dynamic happen with companies that embraced the internet in the early 2000s, and those that didn’t – and we are likely to see this again now.

To start thinking about what to do, we suggest starting with the limit case. Consider if you could develop any maintain any software for free. What would you build? Would you replace external software that you currently pay a subscription for with custom internal software? Would you build new software to automate more of your operations? Would you develop new software-driven solutions for your clients? This provides a starting point. Where you land on what makes sense will then be somewhere between where you are now and that limit case.

Once you have an idea of what you might want to build, you will need to build the software. If you build this externally, you will need to find software development firms who have rapidly adopted to AI – and will pass the savings on to you.

AXX can help you work out how to change you business with custom software being much cheaper than before. We bring the domain and technology experience to do this. AXX can also then develop the new software for you.