In an environment marked by rising correlations, the transformation of private banking and the rise of AI, Bekim Laski deciphers new diversification strategies, the changing role of wealth management and the trade-offs between industrial excellence and technological growth potential.
In times of market stress, correlations between asset classes tend to rise sharply.How should investors rethink diversification in the current environment?
Unfortunately, the empirical data is unequivocal, even if the observation may seem uncomfortable. In times of market stress or regime change, traditional diversification often tends to fail precisely when investors need it most. This phenomenon is no longer confined to listed assets, but now also affects certain pockets of the alternative investment universe. When liquidity contracts and margin calls multiply, investors first sell what they can sell quickly, and not necessarily the most exposed assets from a fundamental point of view.
In this context, diversification based less on traditional asset classes than on risk exposure, liquidity profiles and real performance drivers seems more relevant. Above all, it is important to distinguish between structural diversification and diversification that depends on favourable liquidity conditions. A traditional 60/40 portfolio, particularly in low-yield markets such as Switzerland, no longer offers the same convex protection as it did in the past, when central banks did not permanently compress risk-free rates.
Investors therefore need to think more in terms of value, momentum, quality, low volatility or carry factors, since these exposures tend to behave differently depending on market conditions. Illiquid assets can still play an important role, but only if their illiquidity is real. Private credit, infrastructure and real assets can capture an illiquidity premium, but this becomes less reliable when leverage, secondary market structures or the increasing financialisation of non-listed assets reintroduce liquidity risk. Nevertheless, differentiated return streams still exist, for example in insurance-linked securities such as cat bonds or certain royalty strategies, but they require rigorous due diligence. Explicit convexity via option strategies can also help, although this protection is often costly.
You often stress the importance of a more global approach to asset management.Is the role of the CIO becoming more strategic and cross-functional?
Broader wealth management is not a new concept. It has long been part of the value proposition of private banking in Switzerland. What is changing today is its accessibility. It should no longer be reserved for ultra-high net worth clients. A broader clientele now expects integrated support and a more global view of their assets, which I believe is a natural evolution of fiduciary responsibility. The role of the CIO is changing accordingly. Wealth is no longer limited to the investment portfolio, and performance alone is no longer enough to measure success. The central question is now whether the client's financial objectives are being met over the entire asset life cycle. This means linking asset allocation to liquidity planning, financing, pensions, tax, real estate, private markets and family governance.
The role of the CIO is thus evolving from purely financial asset management to genuine wealth architecture. Alpha remains important, but a more structural alpha is becoming just as decisive. It reflects the value created by coherent, personalised advice that is fully integrated into all major financial decisions, over and above the selection of investment products.
Traditional private banking models are facing increasing competition from more agile and technological platforms.How do you see the wealth management value chain in Switzerland evolving in the coming years?
Traditional private banking has gradually refocused on the wealthiest segments, while digital platforms have made access to markets cheaper and easier. This has created a structural imbalance. Many clients now have access to financial services and products, but not necessarily to strategic advice.
As portfolio management becomes more technological and largely commoditised, the real differentiation is shifting to the overall understanding of the client's financial situation and the ability to translate this into coherent decisions.
At smzh, our ambition is to close this gap with a family office-inspired model that makes integrated advice covering investments, financing, pensions, insurance, tax and real estate accessible, while lowering the entry threshold.
The threat to traditional private banks comes not just from technology, but above all from positioning. Institutions that have historically focused on product distribution are coming under increasing pressure from platforms that are capable of executing at much lower cost.
The winners will be those that offer genuinely credible advice, independence and a coherent range of expertise. Not all skills need to be brought in-house. Modular partnerships make it possible to integrate specialist expertise only where it adds value, without unnecessarily burdening the system.
Many private investors remain heavily exposed to Swiss property and liquidity.What do you think are the most common allocation biases today?
Concentration on Swiss property is indeed one of the most common biases. For many private investors, directly held real estate represents a significant proportion of their net assets. This is understandable, but it creates a concentration risk that is often underestimated, because property seems familiar, tangible and stable.
The second bias is excess liquidity. Even when real interest rates are negative or close to zero, many investors favour nominal stability over long-term purchasing power. In behavioural terms, this is explained by loss aversion. Liquidity may seem safe, but it entails a significant opportunity cost and, in an inflationary environment, a gradual erosion of real wealth.
These include familiarity bias, home bias and the tendency to focus on recent events. After the weakness of 2022, some investors remained permanently underinvested, while others, buoyed by the equity rally of 2023 to 2025, gradually concentrated their portfolios on technology themes. The Swiss market is highly concentrated by sector and offers limited exposure to the major global growth themes. A purely domestic allocation is therefore not automatically defensive; in many cases, it is simply insufficiently diversified.
Does the fact that SoftBank Group has overtaken Toyota Motor Corporation in market capitalisation reveal anything about the way in which markets now value technological ambition over industrial execution?
This does not mean that industrial execution has lost its importance. Toyota remains one of the most operationally disciplined companies in the world. On the other hand, the market now seems to be placing a very high premium on technological prospects and exposure to the investment cycle in artificial intelligence.
SoftBank is essentially a listed investment holding company with a high exposure to venture capital type strategies. Toyota is an industrial company founded on manufacturing excellence, scale and capital discipline. Comparing their market capitalisations therefore does not simply mean that one is 'better' than the other. Above all, it illustrates the fact that investors today are prepared to pay a high premium for the growth potential associated with artificial intelligence, rather than for excellence in mature industrial systems.
In the light of the expected IPOs of Anthropic and OpenAI, what do you think are the key fundamental success factors that will determine the long-term winner in terms of technological leadership, distribution and monetisation?
We should not necessarily expect a market with a single winner. Several AI platforms may coexist, but with different competitive positions. The key factors are likely to be model quality, integration into customer workflows, trust, cost efficiency and the ability to turn AI use into profitable revenue.
OpenAI appears to have a significant advantage in the consumer segment and general productivity use cases. Anthropic appears to be better positioned for the enterprise, regulated sectors and use cases where reliability and control requirements are high. These are different value propositions. Consumer distribution creates scale, brand recognition and a potential data advantage. Enterprise distribution generates larger contracts, deeper integration into processes and stronger loyalty.