Best Buys and Own Brands: Investment Platforms’ Recommendations of Mutual Funds

Howard Jones is an Associate Professor of Finance at University of Oxford Saïd Business School. This post is based on a recent paper, forthcoming in The Review of Financial Studies, by Professor Jones; Gordon Cookson, Associate Director at KPMG UK; Tim Jenkinson, Professor of Finance at the University of Oxford Saïd Business School; and Jose Vicente Martinez, Assistant Professor of Finance at the University of Connecticut.

Retail investors in mutual funds are faced with a bewilderingly wide choice of products. Traditionally, they would be guided by their broker, but increasingly they are investing in mutual funds through online investment platforms, or ‘fund supermarkets’. These platforms produce recommendations of funds to help investors make their choice. Using a unique, largely non-public, dataset sourced from the United Kingdom financial regulator, the Financial Conduct Authority (FCA), we address three questions about these ‘best-buy’ lists— how they are drawn up, whether investors follow them, and whether they add value. We investigate whether platforms’ recommendations are tilted towards two categories of funds from which they are especially likely to benefit: funds affiliated with the platforms themselves and those which share a large part of their own commission revenues with the platforms.

Background and Data

In the United States, the distinction between brokers and investment platforms is not clear-cut, as many brokers also operate platforms and do not break out the flows through different channels—nor do they publish their fund recommendations in a systematic way. Prior research working with U.S. data has found that broker-mediated mutual funds have higher fees and worse performance (Bergstresser, Chalmers, and Tufano 2009; Chalmers and Reuter 2012; Del Guercio and Reuter 2014), and that the incentives of brokers intermediating mutual funds, notably revenue sharing, drive flows into those funds (Christoffersen, Evans, and Musto 2013). However, these studies have not been able to separate recommendations from other, unobservable benefits which brokers may provide to retail investors, such as advice on the appropriate mix of asset classes. The U.K. platforms we study are intermediaries which limit their advice to recommendations, which they make freely available across a wide range of asset classes and regions. Moreover, our regulator-sourced dataset includes non-public details of the fees charged by asset managers, the fraction of these fees shared with platforms and, for two of the platforms, the investor flows into mutual funds that were channeled through the platforms. We are therefore able to make a more direct link between recommendations on the one hand, and incentives, flows, and fund performance on the other than has been possible hitherto.

Drivers of recommendations

We find that funds which are affiliated with the platform are more likely to be recommended by platforms than others. Moreover, for that part of our sample period when commission-sharing was permitted, funds were more likely to be recommended if they shared a higher proportion of their revenues with the platforms. A cross-check with funds rated by the financial-services company Morningstar revealed that Morningstar’s recommendations were not biased towards these two categories of funds. As detailed below, there is in general a large overlap between the recommendations of platforms and those of Morningstar, and the absence of overlap for the two categories of funds that platforms are incentivized to recommend suggests that platforms recommend them out of favoritism.

Impact of recommendations on flows

Our analysis shows that the impact of platforms’ recommendations on flows is significant and substantial. For every month recommended, a fund experiences an average annualized inflow of 1.92% of the fund’s total assets under management. When we interact recommendations with affiliation, we find that flows are less responsive to the recommendation of a fund affiliated with a platform, suggesting that retail investors are wary of such recommendations. On the other hand, the flows to recommendations do not vary significantly as a function of revenue-sharing, suggesting that investors do not discount platforms’ biases in this respect. The different responses likely reflect the fact that fund affiliation is obvious to investors (notably because the affiliated fund and the platform are branded similarly), whereas in the U.K., revenue-sharing agreements are not disclosed.

Value added by recommendations

Over our ten-year sample period, we find that a portfolio of all the funds recommended by platforms yielded average net returns of 0.08% per year in excess of a broad benchmark. Non-recommended funds accessible via the same platforms underperformed their benchmark by 0.86% per year. Running the analysis using narrower benchmarks, which more closely match the investment category, yields similar results. However, while the value added by platform recommendations is significant and positive, it should be noted that these recommendations are closely correlated with recommendations by Morningstar and do not significantly outperform them. Moreover, when platform recommendations do not match those of Morningstar, as occurs especially in the case of affiliated funds and those with high revenue shares, those recommended funds do not outperform. These findings are further evidence that platform recommendations are driven by conflicts of interest.

Effects of a ban on commission-sharing

In 2014, towards the end of our sample period, a regulatory change in the U.K., the Retail Distribution Review (RDR), banned asset managers from sharing commissions with platforms; instead, platforms earned fees by charging retail investors directly for their services. Following RDR, we find that platforms increased the number of affiliated funds they included on their recommendation lists—consistent with platforms increasing their recommendations in one conflicted category, affiliated funds, to compensate for the loss of another source of revenues, commission-sharing. Despite this, the value added of platform recommendations did increase when compared with the value of those of Morningstar, lending further support to the argument that platform recommendations were driven by a conflict of interest.

In line with Stoughton, Wu, and Zechner (2011), RDR also led to a fall in the costs to retail investors of investing in mutual funds through platforms. We find that fees and charges to retail investors in mutual funds fell by 33 bps, on average, partly because the same funds lowered the commissions they charged and partly because platforms replaced more expensive funds with less expensive ones. In this way, RDR had an impact not only on which funds were recommended by platforms but also on which funds were offered at all.


We find that platform recommendations are biased in favor of two categories of fund from which they particularly gain, affiliated funds and those sharing high levels of revenues with them. The removal of the latter conflict by the separation of advisory and asset management charges can lead to a reduction in costs to investors. At a time when bans on commission-sharing have been recently implemented elsewhere (e.g. in the Netherlands and Australia) or are on the way to being implemented (e.g. in Switzerland, Sweden, South Africa, and Canada), our findings are consistent with the potentially beneficial effects of greater disclosure of otherwise hidden incentives.

The complete paper is available here.

Both comments and trackbacks are currently closed.