With new trading technologies providing non-discriminatory access to the world’s leading equity markets from any geographical location, you might think that it no longer matters where a trader is based. Certainly, that is becoming the conventional wisdom in the international investment community.

But a recent CEPR Discussion Paper by Harald Hau indicates that in the case of Germany at least, locals still have an advantage. He has uncovered powerful new evidence that traders in German equities who are located outside Germany in non-German-speaking cities make significantly smaller profits on their proprietary trading than their German counterparts. But language and culture seem to be the key factors rather than geography per se: German-speaking traders do equally well wherever they are located.

Hau’s research explores the relationship between a trader’s geographical location and the informational advantages that may offer. For example, do traders in foreign countries underperform relative to domestic traders when it comes to proprietary trading of domestic equities? If so, do language barriers explain the performance differences? Do traders in a global financial centre like Frankfurt make more out of DAX stocks – the German blue chips – than traders in regional centres like Munich? Do traders in Munich do better than traders in Frankfurt with stocks where the corporate headquarters is in Munich? And do larger financial institutions with many traders outperform smaller institutions because of scale economies in the production of information?

These questions are related to what determines the distribution of equity holdings across the wider investment community. Empirical research has established that most portfolios are dominated by domestic equities, a phenomenon known as ‘home bias’. This behaviour implies that investors forgo the benefits of better risk diversification that a truly international portfolio would confer. Initial explanations for this phenomenon focused on barriers to international investment, such as government restrictions on foreign and domestic capital flows, taxes, tariffs and fees. But recent evidence shows that even within the United States, investors tend to prefer the stocks of local companies. Local informational advantages seem to be the main explanation.

But the ultimate test of whether local information matters is superior trading performance, not simply portfolio bias in favour of more familiar stocks. Pursuing this idea, Hau has gathered data on more than 1,300 traders who use Xetra, the electronic trading system of the German stock market. He concentrates on the 451 largest traders; those that make at least 100 transactions in any of 11 randomly selected DAX stocks over a four-month period.

These large traders are located in 23 different cities and eight different European countries. Yet in spite of their wide geographical distribution, they all have equal access to the electronic trading platform. The absence of institutional discrimination makes Xetra data ideal for tracing differences in information – ‘asymmetries’ – across the trader community.

Hau measures the profitability of the traders over different time periods, adjusting for risk and differentiating between intra-day or high frequency profits, weekly or medium frequency profits and monthly or low frequency profits. This distinction is useful since the data spans only four months and intra-day profits can be observed more frequently than weekly or monthly profits. The high frequency profits therefore provide a more powerful statistical measure of informational asymmetries than lower frequency ones.

So what role does location play in determining individual trading profits at the three frequencies? The analysis reveals that a trader based in Frankfurt enjoys no comparative profit advantage over other German traders, which suggests that there are no special benefits from being located in a world financial centre. In contrast, being based in a foreign non-German-speaking city has a strongly negative correlation with profitability, a shortfall that Hau documents for intra-day, weekly and monthly profits. At the same time, foreign traders in other German-speaking countries – Austria and Switzerland – do not make significantly lower profits. This suggests that it is linguistic and cultural barriers rather than geographical distance per se that hold the key to the informational advantages identified in the data.

The underperformance by non-German-speaking foreign traders is not only statistically significant; it is also of considerable economic magnitude, averaging 20 D-Marks per market transaction. That is the equivalent of a relative quarterly profits loss of approximately 1.2 million D-Marks per actively traded account.

Are there are informational advantages that derive from being physically close to the companies whose stocks are being traded? Hau examines whether traders operating no more than 100 kilometres from one of the nine corporate headquarters outside Frankfurt (two are in Frankfurt) have superior trading performance in those stocks. He finds that local proximity is positively correlated with intra-day profits, indicating local informational advantages. But the correlation is not significant at lower frequencies.

The data also reveal no evidence that large financial institutions have a better trading performance than smaller ones. There appear to be no ‘informational scale economies’ in proprietary equity trading.

Hau’s findings, notably the profit difference between domestic and foreign traders, confirm that language, culture and geography continue to be important in determining how information is distributed across financial market participants. They suggest that market making in the equity business may be considerably less global than is often claimed.

This article discusses research reported in ‘Information and Geography: Evidence from the German Stock Market’, CEPR Discussion Paper No. 2297 (November 1999) by Harald Hau. Hau is at ESSEC in Cergy-Pontoise and a Research Affiliate in CEPR’s International Macroeconomics programme.