Home sweet home – even in your pension fund portfolio? 

Investors who invest according to basic theories of finance use the market portfolio as a guide. It contains every investment in proportion to their market value and offers the highest reward for the risk being taken. Broadly diversified, it completely neutralizes the idiosyncratic risks that are not compensated by the financial market and has a beta of 1.

However, a look at the asset allocation of Swiss pension funds shows a significant structural deviation from this theoretical concept, especially regarding geographical allocation. In late March 2020, the Pension Fund Index showed that 24.4% of total assets were invested in foreign bonds, real estate, and equities.

In the market portfolio, the optimum share for the home market is calculated using the ratio of Switzerland's market capitalization to the other countries in the reference index. Accordingly, the amount allocated to foreign bonds, real estate, and equities ought to be considerably higher. The average institutional portfolio shows a significant bias toward Switzerland compared to the market portfolio.

Distinct home bias in all asset categories

As the figure below shows, this home bias exists in every major asset category. The share of foreign securities in the bond portfolio is 16% according to the Pension Fund Index instead of 99% based on the market weighting, and has even decreased slightly in the past few years. By contrast, investors have increased their allocation to foreign real estate. The amount is still extremely low at a mere 12% (instead of 99%). The proportion of foreign equities in the stock portfolio of the Pension Fund Index is 56% and has hardly changed over the past ten years.

According to the 2020 Global Pension Assets Study, the equity and bond portfolios of foreign pension funds also show a clear preference for their respective home markets. By international comparison, Switzerland displays an average level of home bias. Besides tax-related issues, currency risks, and asymmetric information, the reasons for the tendency to show home bias also include aspects of behavioral economics such as (supposedly) better market knowledge through geographical proximity or optimism about the returns in the home market. Switzerland's special position as a safe harbor can also help explain why it takes up a relatively large share of the portfolio. Particularly in times of crisis, the Swiss franc serves as a safe-haven currency for investors. It has been appreciating for decades for structural reasons.

From a regulatory point of view, the only limit imposed on Swiss pension funds when it comes to the proportion of foreign assets they can hold is in the real estate asset category. That limit is set at one-third of the maximum real estate allocation (Ordinance on Occupational Retirement, Survivors' and Disability Pension Plans 2, Art. 55(C)), which corresponds to 10% of total assets. The regulator limits the share of unsecured investments in foreign currencies to 30%.

Does the dominance of the Swiss home market make investment strategies irrational?

Portfolio theory teaches that such a significant home bias leads to a diversification deficit and, accordingly, to portfolios with suboptimal risk-reward ratios. Is there nevertheless a justification for the dominance of the home market?

In actual fact, portfolios containing a disproportionate share of Swiss securities have managed to post better returns during the current coronavirus pandemic. They also suffered significantly fewer losses during the dot-com bubble and the financial crisis.

The figure below shows the performance of a portfolio based on the current allocation in the Pension Fund Index of 48% equities, 46% bonds, and 6% liquid assets. The investments in the "home bias" portfolio are invested according to the actual allocation between Swiss and foreign securities within those asset categories. The market" portfolio, by contrast, invests in accordance with Switzerland's share of the broader market. The observation period ran from January 2002 to March 2020, and the "home bias" portfolio achieved a return of 3.7% p.a., higher than the "market" portfolio (3.1% p.a.), and it did so with lower volatility. Both portfolios suffered their greatest loss in February 2009. However, the "home bias" portfolio's decline of 26.4% was slightly lower (28.4% for the "market" portfolio). Viewed historically, the home bias has generated better risk-adjusted returns in roughly the past 20 years.

However, the chart illustrates that this is not true for every market phase. There is no discernible pattern during upward swings. For example, the "market" portfolio produced better returns during the ten-year bull market after the financial crisis, while the "home bias" portfolio performed better during the upswing from May 2004 to October 2007. In times of crisis, by contrast, the "home bias" portfolio always fared better, meaning its losses were not as severe. This pattern repeats itself with different allocations. For example, the same thing happens to a portfolio with 25% of investments allocated to real estate, 35% to equities, 34% to bonds, and 6% to liquid assets. This time, the observation period extended from January 2007 to December 2019. Here as well, the upswings on the stock markets did not show a clear pattern, but the "home bias" portfolio consistently outperformed during the downward phases.

The "home bias" portfolio in the observation period chiefly owes its attractiveness to the strong performance of Swiss equities versus the world index. They posted fewer losses during the observation period, particularly during the three major collapses. With regard to bonds, the yield differential between domestic and foreign investments was relatively small by comparison.

How the structural deviation from the market portfolio will affect the yield and the risk to the pension funds' portfolios going forward depends on the yield performance of Swiss securities compared to foreign ones as well as on the correlations in the portfolio context and could, in turn, vary based on the period examined and the indices used.

Market portfolio falls short

This shows that it is not enough to simply base one's investments on the market portfolio dictated by financial theory. The conventional wisdom of the market portfolio says that ideal allocation corresponds to the market-weighted total of all investments. However, it is nearly impossible to implement that in practice. Broad-based capital-weighted indices do indeed make it possible to invest in a manner that is approximately proportional to market capitalization. Nevertheless, they always track only a part of the investment universe. For instance, the broadly diversified MSCI ACWI Index with over 3,000 securities reflects only about 85% of global equities and neglects securities such as frontier markets, unlisted companies, and companies with a small free float. The market portfolio, in contrast, contains every security regardless of size or free float. In reality, it is impossible for pension funds to invest in a portfolio with the (theoretically) optimal risk-reward ratio.

It is important to consider for each asset category which market characteristics and expectations justify giving the home market a higher weighting and to carefully set the long-term strategic asset allocation as part of an asset and liability management analysis. Short- to medium-term expectations can be incorporated by means of tactical investments while implementing that strategy. This enables individual countries in the portfolio to be under- or overweighted.

Swiss bonds allow secure alignment with liabilities

If we look more closely at the bond portfolio, we see that excessive bias toward the Swiss market results in concentration risks. That is because the market for CHF bonds is dominated by issuers from the financial industry, making it more poorly diversified than the global market. At the same time, its liquidity is significantly lower compared to other markets. Including bonds denominated in foreign currencies but hedged in CHF can counteract this situation. Due to the high correlation between yields on CHF bonds and FC bonds (hgd)¹ (0.92 over the past ten years), we consider them equal to CHF bonds on the broader market over the long term.

Bonds, however, also play a key role in the risk management of Swiss pension funds. That is because their liabilities are denominated exclusively in Swiss francs and can be covered by bonds offering a high level of security. Bonds denominated in CHF allow for the safest cash flow matching since all payments are made directly in Swiss francs.

Particularly in the case of pension funds that have a high proportion of retirees and a correspondingly low structural risk ability, there is certainly justification for a higher weighting of domestic bonds.

Equities bring uncompensated risks

When investing in equities, taking asset and liability management into account makes it difficult to argue for home bias. One particular reason for that is the poor diversification of the Swiss equity market. The heavyweights, Nestlé, Novartis, and Roche, make up more half of the Swiss Performance Index, and the largest ten securities account for roughly 70%. What's more, the industry structure is extremely unbalanced, with approximately two-thirds of the market being driven by the healthcare sector. This concentration at both the stock and sector levels leads to security-specific risks for which investors are not compensated.

Home bias would be justified only if Swiss equities earned more structurally attractive returns. That was the case in the recent past. For example, the real annual return on Swiss equities in Swiss francs from 2000 to 2019 was 3.1% p.a., which was 1% above that for global equities (2.1% p.a.). Yet, if we look at the periods from 1970 to 2019 or even from 1900 to 2019, we see that global equities performed better.² Over the long term, we expect comparable risk premiums for domestic and foreign securities. That is because there is not an economic explanation for why the Swiss equity market ought to outperform the global market in the long run.

In the Pension Fund Index, 28.8% of assets are currently allocated to equities, with 16.2% in domestic and 12.6% in foreign securities. Based on market weighting, the proportion of foreign equities would need to be 27.7% of the total allocation. The home bias is least obvious in the area of equities, but it is still clearly visible. However, in Switzerland's unique case, the fact that major, internationally oriented corporations make up the lion's share of the Swiss equity market puts it into perspective somewhat. They may be listed in Switzerland, but those companies earn most of their revenue outside Switzerland, thereby offering de facto global exposure.

Structural advantages of the Swiss real estate market

When it comes to real estate, the dominance of the home market in the portfolios of Swiss pension funds is especially pronounced. There are concrete reasons for that. First, a portion of the home bias has grown over time since, in their early years of employee benefits insurance, pension funds provided their insured with housing. Second, investing in foreign properties is highly challenging and is carried out primarily through investment funds. Then there are additional hurdles in the form of limited marketability, tax aspects, and transaction costs.

Due to the limited size of the Swiss market and its relative homogeneity, the clear home bias leads to very concentrated portfolios. They are disproportionately exposed to Switzerland's framework conditions in the areas of tenancy law, migration, demographics, and building codes.

Nevertheless, a certain amount of bias toward the home market can be justified by the structure of the Swiss real estate market. The domestic real estate market has a relatively high percentage of residential properties. The strong legal protection for renters makes their yields stable, and their performance is significantly less dependent upon economic developments than the returns on commercial properties. So, there is a weaker correlation between them and Swiss equities.

Nonetheless, the opportunities provided by the diverse global real estate market should not be ignored. With a correlation of -0.15,³ foreign real estate funds perform largely independently of Swiss funds, enabling not only diversification across properties and regions but also through different business cycles.

The need to distinguish between investments by asset category

The average pension fund portfolio has a significant bias toward the Swiss market in every asset category, causing it to deviate from the theoretical market portfolio considerably. In hindsight, home bias has proved to be helpful, especially in crisis situations.

However, some distinctions need to be made when examining the individual asset categories. With regard to bonds and real estate, overweighting domestic securities may be justifiable from a standpoint of risk management and structural considerations. As for equity investments, home bias is less pronounced in the current allocation of the Pension Fund Index. Even though it can have a positive effect on returns in individual phases, it is hard to explain over the long term because, especially on the highly concentrated Swiss market, the lack of international diversification leads to non-systematic risks.