MiFID II and research: Did it change anything?

In 2018, the EU introduced the MiFID II rules, part of which required the unbundling of research revenues from trading revenues, so research analysts wouldn‘t make recommendations just to increase trading volume for their employer.

Well, did it work? Falko Fecht and his colleagues from the Frankfurt School of Finance and the German Bundesbank used the stock recommendations of analysts of German banks and brokers together with the databases of the German Bundesbank to see how research advice has changed since the introduction of MiFID II. They find some tentative signs that with the introduction of the new regulation some things have become better for investors:

• The frequency of analysts’ recommendations and recommendation changes has declined somewhat indicating that there might be less of an incentive for analysts to change ratings or make recommendations to incentivise client trading.

• The accuracy of analyst EPS forecasts has increased somewhat, though it is unclear if this is due to analysts having better quantitative tools at their disposal over the last couple of years or because of the change in incentive structure.

Then there are some mixed blessings:

• Clients of German banks and brokers react more strongly to recommendation changes of these banks since the introduction of MiFID II. This is not because they have realised that the recommendations of their house broker have become better, but because they no longer have access to research from a wide range of brokers. Instead, they often must rely solely on their house broker’s research.

And then there are some concerning developments:

• German banks and brokers are more likely to recommend a stock for purchase to their clients while the firm is selling the same stock in their own portfolios. To be fair, the average correlation between bank research recommendations and a bank’s own portfolio trades is positive and has become bigger after the introduction of MiFID II. On average, banks trade in line with the recommendations they give their clients. But when zooming in on the stocks where the bank recommends a purchase while its portfolio managers are selling, that relationship has become stronger as well, indicating that the bigger influence a bank has on the trading activities of its customers since the introduction of MiFID II (see above) creates an incentive to abuse this influence in favour of the bank’s own P&L.

Overall, MiFID II and the unbundling of research don’t seem to be such an overwhelming success and one unintended consequence seems to be that some bankers in Germany are going back to the pre-2000 days of inflating recommendations to customers while selling the same stocks in their own accounts.