Connect with us

Finance

Artificial intelligence and asset pricing: The power of transformers

Published

on

The rapid advancement of artificial intelligence (AI) has reshaped numerous fields, and finance is no exception. Asset pricing, a domain traditionally dominated by linear models and factor-based approaches, is now experiencing a transformative shift due to AI’s capacity to uncover persistent predictive patterns in financial data. The introduction of large-scale AI models – particularly transformer-based architectures – has significantly enhanced our capability to model complex relationships among assets and firms (Eisfeldt et al. 2023), leading to improved forecasting and risk assessment.

In this column, we explore how AI-driven asset pricing models, particularly those incorporating transformers, leverage cross-asset information sharing to reduce pricing errors and enhance predictive accuracy. These innovations offer a novel perspective on how financial markets process information and determine asset prices.

Traditional versus AI-driven asset pricing

Most traditional asset pricing models, such as the Fama and French (1993) framework, rely on predefined factors that explain asset returns. While effective, these models assume a fixed and linear relationship between asset characteristics and expected returns. AI, on the other hand, introduces a non-linear, data-driven approach, identifying patterns that are often invisible to traditional methods.

Machine learning models, including tree-based methods and neural networks (Gu et al. 2020), have improved asset return prediction by capturing complex relationships between firm characteristics and returns. However, these models typically focus on ‘own-asset prediction’, meaning that they use only an asset’s individual characteristics to estimate future returns. This approach ignores the broader context in which assets interact.

The role of transformers in asset pricing

In a recent paper (Kelly et al. 2024), we introduce the Artificial Intelligence Pricing Model (AIPM), which embeds transformer networks into the stochastic discount factor (SDF) framework. While initially developed for natural language processing, we show that transformers are remarkably effective in financial applications due to their ability to capture cross-sectional dependencies across assets.

Advertisement

Unlike traditional machine learning models, transformers incorporate the ‘attention mechanism’, allowing them to dynamically adjust the weight placed on different inputs based on their relevance. In the context of asset pricing, this means that the model not only considers an asset’s own characteristics but also how these characteristics interact with those of other assets. This approach significantly enhances predictive power by leveraging market-wide information.

Empirical findings: Performance of AI-based models

We evaluated the performance of the transformer-based AIPM using a dataset of US stock returns and conditioning variables. Compared to traditional asset pricing models and other machine learning approaches, the AIPM demonstrated:

  • Lower pricing errors: The model achieved significantly smaller out-of-sample pricing errors compared to traditional factor-based models and neural networks without attention mechanisms.
  • Higher Sharpe ratios: By integrating cross-asset dependencies, the transformer-based model outperformed existing approaches in terms of risk-adjusted returns out-of-sample.
  • Scalability and complexity gains: We found that increasing model complexity – by incorporating deeper transformer layers – consistently improved predictive performance. This supports the notion that AI models benefit from higher parameterisation when applied to asset pricing.

Empirical evidence from the literature on large language models (e.g. Kaplan et al. 2020) indicates that adding more transformer blocks enhances the capacity of models to effectively represent language. Models with deeper architectures are able to capture more abstract features and longer-range dependencies than shallower models. Each additional layer refines the attention distributions, allowing the model to consider both short-term and long-term relationships. Interestingly, the same benefits of transformer complexity emerge in the context of the asset pricing model.

These findings suggest that AI-driven asset pricing models are more efficient in processing vast amounts of financial data, leading to more accurate and robust predictions.

Implications for investors and policymakers

The integration of AI into asset pricing has profound implications for market participants and policymakers. Investors can benefit from improved portfolio allocation strategies driven by AI’s capability to identify subtle pricing inefficiencies, a concept aligned with research on AI and personality traits shaping economic returns (Makridis 2025). Meanwhile, regulators and policymakers must consider the impact of AI-driven trading on market stability and efficiency.

Moreover, the adoption of AI in asset pricing challenges traditional views on market efficiency, echoing broader concerns about AI’s macroeconomic impact and productivity gains (Filippucci et al. 2024). If AI models consistently outperform classical frameworks, it may suggest that markets are less efficient than previously thought, opening the door for further research into the nature of pricing anomalies.

Advertisement

Conclusion

The fusion of artificial intelligence and finance is revolutionising asset pricing. Our research demonstrates that transformer-based models significantly enhance return predictions by leveraging cross-asset information sharing. As AI continues to evolve, its role in financial decision-making will only grow, offering new opportunities for investors and reshaping our understanding of market dynamics.

References

Gu, S, B Kelly and D Xiu (2020), “Empirical Asset Pricing via Machine Learning”, Review of Financial Studies 33 (5): 2223-2273.

Eisfeldt, A, G Schubert and M B Zhang (2023), “Generative AI and Firm Valuation”, VoxEU.org, 4 June.

Fama, E and K French (1993), “Common risk factors in the returns on stocks and bonds”, Journal of Financial Economics 33(1): 3-56.

Filippucci, F, P Gal, and M Schief (2024), “Miracle or Myth? Assessing the Macroeconomic Productivity Gains from Artificial Intelligence”, VoxEU.org, 8 December.

Advertisement

Kaplan, J, S McCandlish, T Henighan et al. (2020), “Scaling laws for neural language models”, arXiv preprint arXiv:2001.08361.

Kelly, B Kuznetsov, S Malamud and T Xu (2024), “Artificial Intelligence Asset Pricing Models”, NBER Working Paper No. w33351

Makridis, C (2025), “The Role of Personality Traits in Shaping Economic Returns Amid Technological Change”, VoxEU.org, 31 January.

Vaswani, A, N Shazeer, N Parmar, J Uszkoreit, L Jones, A N Gomez, L Kaiser and I Polosukhin (2017), “Attention is all you need”, Advances in Neural Information Processing Systems 30.

Advertisement
Continue Reading
Click to comment

Leave a Reply

Your email address will not be published. Required fields are marked *

This site uses Akismet to reduce spam. Learn how your comment data is processed.

Finance

Rogers Sugar AGM: Shareholders Approve Directors, KPMG Auditor and “Say on Pay” Resolution

Published

on

Rogers Sugar AGM: Shareholders Approve Directors, KPMG Auditor and “Say on Pay” Resolution
Rogers Sugar (TSE:RSI) shareholders approved all resolutions brought forward at the company’s annual meeting, including director elections, the appointment of auditors, and a non-binding advisory “say on pay” vote, according to preliminary results reported by the meeting’s scrutineer. The meeting w
Continue Reading

Finance

Block vs. PayPal: Which Fintech Stock Is Better Positioned for 2026? | The Motley Fool

Published

on

Block vs. PayPal: Which Fintech Stock Is Better Positioned for 2026? | The Motley Fool

Two companies battling to win the global payments market.

Great businesses win by solving problems, and the $2.5 trillion global payments market is a goldmine for companies that can make money move effortlessly.

Two of the firms competing in that space are Block Inc. (XYZ +4.85%) and PayPal Holdings Inc. (PYPL +1.30%).

Image source: Getty Images.

As each pushes into new technologies and revenue streams, the next year could define their long-term trajectories.

Advertisement

With this potential turning point, I’ll examine which fintech stock may fit best in your portfolio.

PayPal’s moves into AI, global payments, and stablecoins

PayPal shares have dipped 37.28% over the last year, but the company has three initiatives that could help reverse that trend: PayPal World, artificial intelligence (AI) agents, and cryptocurrencies and stablecoins. PayPal World and AI agents enhance the current services, while crypto and stablecoins open up entirely new financial terrain for PayPal.

PayPal Stock Quote

Today’s Change

(1.30%) $0.52

Current Price

$40.42

Advertisement

Announced in June 2025, PayPal World will allow customers to pay global merchants using their payment system, or wallet of choice, in their local currency. In essence, you’ll start seeing PayPal integrate seamlessly with other payment services.

Advertisement

For AI shopping, PayPal says a customer can tell an AI agent they need a ride to the airport at 4:50 a.m. The agent can both book that appointment and pay for it.

Finally, that brings us to cryptocurrencies and stablecoins. The company enables the buying, selling, and sending of crypto within its wallets. PayPal also offers its own stablecoin pegged to the U.S. dollar called PayPal USD (PYUSD) for fast, global payments. As of this writing, holding PYUSD offers a 4% annual yield.

Its peer-to-peer payment service, Venmo, can also boost revenue over time. As a reference point, in 2021, PayPal said it generated roughly $900 million from Venmo. PayPal expects it to generate $2 billion in revenue by 2027.

Block’s next growth chapter

Similar to PayPal, Block shares have stumbled over the last year, dipping 22.48%.

Block Stock Quote

Today’s Change

(4.85%) $2.59

Advertisement

Current Price

$55.97

Once again, the key is looking at what lies ahead.

Advertisement

Its flagship Cash App service still has the reputation of friends just sending each other money, but Block is focused on turning it into a complete financial platform. Through banking, savings, direct deposit, bill paying, an AI-powered money assistant, and more, users are gaining fuller control of their financial lives through just one app. In Q3 2025, Block reported $1.62 billion in gross profit from Cash App, a 24% year-over-year increase.

Its global lending products have now surpassed $200 billion in provided credit. Defaults remain low, with 96% of buy now, pay later installments paid on time and 98% of purchases incurring no late fees.

Outside of its consumer products, Block is building out a robust suite of merchant tools to provide businesses with everything they may need, including credit card terminals, payroll services, and loyalty program marketing campaigns. Business owners can also build websites through Block, which could lead sellers to adopt more of its tools over time.

Block has also leaned deeper into cryptocurrencies. In October 2025, it launched Square Bitcoin, which will automatically convert credit card sales into Bitcoin. Block also holds roughly 8,800 BTC, worth nearly $770 million.

The PayPal vs. Block winner

PayPal and Block are both stocks that could rebound in 2026 if their initiatives gain traction.

Advertisement

Block has high-growth segments in cryptocurrencies and lending, and its expanding suite of services and tools for businesses can help it generate more revenue from its current customer base. That high upside potential also comes with a high beta of 2.66, meaning it is more than two and a half times more volatile than the general stock market. Despite those issues, the balance sheet is strong, with $8.7 billion in cash compared to $8.1 billion of debt.

PayPal has steady, transaction-based fees from its global payments platforms and even pays out a dividend of $0.56 per share. Its beta of 1.43 also means it’s less volatile than XYZ. This may appeal more to risk-averse investors. The key here will be if PayPal’s recent moves can take it beyond being just a steady and mature business. With $12.17 billion in debt and $10.76 billion in cash, PayPal operates with a slight net debt that’s reasonable considering its consistent earnings.

Ultimately, the choice comes down to whether you prefer owning PayPal as a dependable revenue machine that could grow meaningfully as it enhances its services and features, or Block’s higher-risk path that could deliver outsized returns if its bets pay off.

Continue Reading

Finance

Bond Markets Are Now Battlefields

Published

on

Bond Markets Are Now Battlefields

As the Greenland crisis came to a head in the days before Davos, Europeans sought tools that could be reforged as weapons against the Trump administration. On Jan. 18, Deutsche Bank’s global head of foreign exchange research, George Saravelos, warned clients in a note that “Europe owns Greenland, it also owns a lot of [U.S.] treasuries,” and that the EU might escalate the conflict with a “weaponization of capital” by reducing private and public holdings of U.S. debt instruments.

U.S. Treasury Secretary Scott Bessent reported later that week that Deutsche Bank no longer stood behind the analyst’s report, but Saravelos was far from the only financial analyst to discuss the idea. Within days, a few European pension funds eliminated or greatly reduced their holdings of U.S. Treasurys and—perhaps as a result—U.S. language about European strength became considerably less aggressive.

As the Greenland crisis came to a head in the days before Davos, Europeans sought tools that could be reforged as weapons against the Trump administration. On Jan. 18, Deutsche Bank’s global head of foreign exchange research, George Saravelos, warned clients in a note that “Europe owns Greenland, it also owns a lot of [U.S.] treasuries,” and that the EU might escalate the conflict with a “weaponization of capital” by reducing private and public holdings of U.S. debt instruments.

U.S. Treasury Secretary Scott Bessent reported later that week that Deutsche Bank no longer stood behind the analyst’s report, but Saravelos was far from the only financial analyst to discuss the idea. Within days, a few European pension funds eliminated or greatly reduced their holdings of U.S. Treasurys and—perhaps as a result—U.S. language about European strength became considerably less aggressive.

Advertisement

It’s unclear how much of an impact Europe’s moves had on the White House backing off. But it poses a number of questions: Can Europe take advantage of weaponized interdependence to wage financial warfare against the United States? How big are the obstacles in the way, and how much impact can such moves have?

Financial flows and financial policy are instruments of coercive power. There is some evidence of financial flows putting pressure on the United States last year; in the wake of his triumphant declaration of mass tariffs in April, movement away from Treasurys reportedly persuaded President Donald Trump to partly change course.

However, this seems to have been an organic, unplanned development and a short-lived one.

Despite the precipitous fall of the dollar, and lively discussion over the past year of the United States losing its reserve currency status, the evidence points to mundane concerns about inflation and policy uncertainty leading to a slow reallocation of investment from the United States to other countries rather than any kind of coordinated response. Expert observers have asked if it is even possible for Europe to do anything further given its active trade with the United States, its smaller markets, and its interdependence. The Financial Times’s Alphaville blog summarized the idea of weaponization as “implausible.”

Yet the potential is there. History can be instructive. The state weaponization of finance feels new but, in fact, is centuries old. In the last decades of the 19th century, European governments—particularly France and Germany—aggressively used finance to advance their interests. The subservience of finance to diplomacy was considered natural; to propose otherwise could be dismissed as “financial pacifism.” At a critical moment in conflict with Russia, German Chancellor Otto von Bismarck banned the Reichsbank from accepting Russian securities as collateral. After the Franco-Prussian War an “official but tacit ban” was used to prevent French investors from putting any money into Germany.

Advertisement

How might similar action look today?

The main battlefield for weaponization is markets for sovereign debt—Treasurys on the U.S. side and the mix of national and European Union-level debt instruments on the European side. If Carl von Clausewitz had been a banker instead of a general, he would have pointed to these instruments as the “center of gravity” of any coercive financial operations. Here, the United States has a distinct advantage: Treasurys are the core market of international finance—large, very deep, very liquid. They form the backbone of world financial flows, a major channel of supply and demand for local markets everywhere.

Virtually all national financial markets are tied to the U.S. Treasury market, and it greatly eases the U.S. ability to borrow. This makes it a potentially powerful target for European pressure but also, at best, a delicate one—it is very difficult to launch pressure that does not boomerang back against the EU. Much of EU ownership of Treasurys is also in private hands.

Despite all this, European governments still have the means to go on the offensive. Finance is notoriously sensitive to the arbitrage opportunities created by regulation, such that leading textbooks on the industry include extensive discussion of loophole mining. (This may also explain why lawyers can now earn more than bankers on Wall Street.) If clever bureaucrats at the European Central Bank and EU and elsewhere created the right loopholes, then European funds could move accordingly. Instead of banning use of Treasurys as collateral à la Bismarck, slight adjustments of their risk weight or tax impact under EU or national law should do the trick. There are great technical and political challenges, but it is absolutely doable.

On a defensive basis, Europe can improve its financial position by further developing common  EU debt, building on the large-scale Next Generation EU issuance during the COVID-19 pandemic. In December, EU leaders agreed to raise 90 billion euros ($106.3 billion) for Ukrainian defense, and further steps are very much under discussion. The political and technical challenges to full development of common debt options are obviously enormous, requiring the historically unprecedented establishment of a large, stable market for supranational debt.

Advertisement

EU common debt tends to trade at a discount relative to comparable national debt, showing investors’ concerns. However, the potential payoffs are significant. In addition to facilitating EU-wide defense planning and creating a clear substitute for the Treasurys market, a strong common debt market could create a new and more powerful backbone to European finance, investment, and economic growth.

None of the above analysis should be viewed as prescriptive; by far the best path forward is a negotiated return to the rules-based order as opposed to a collapse into the full anarchy of unrestrained interstate competition. Unfortunately, the Trump administration seems committed to an aggressive policy that puts that order in peril. From at least the Napoleonic wars to the end of World War II, national interests regularly hijacked international markets, pushing them away from their idealized Economics 101 role as mechanisms of price discovery and efficient allocation into channels of pressure and coercion.

In an effort to bottle up these destructive spirits, the Franklin Roosevelt administration—with the assistance of economist John Maynard Keynes—used the United States’ status as the most powerful surviving state to implement the Bretton Woods system of financial and political controls. The success of the Bretton Woods project can be measured in part by how many of the tactics of the previous eras have been forgotten.

As the past month shows, these tactics and their destructive side effects are reemerging as the order collapses. Once again, bond markets are now battlefields.

Advertisement
Continue Reading

Trending