Passive management has transformed the asset management industry. The first index fund was started in 1972. Since then, the US index mutual fund industry has grown to more than $4 trillion, which represents 24% of the industry’s assets under management. This growth was driven by skepticism about the value of active management and a relentless search for lower fees. The next wave of innovation was the emergence of “factor-based” funds, sometimes called “smart,” “strategic,” or “dynamic” beta. These funds sit between active and passive management by following
preset investment rules. They also charge less than active funds by implementing automated quantitative strategies that eschew security-specific analysis. Initially, these funds focused mainly on long equities for example, by tilting the portfolio toward cheap, value stocks. The hope was that such strategies would generate a “risk premium” or
positive long-term return because of some behavioral anomaly or institutional constraint. Alternative risk premia (ARP) funds represent the next logical step. These funds take both long and short positions to navigate within and across asset classes. Such flexibility is the domain of hedge funds. Indeed, ARP funds aim to replicate some of the returns of hedge funds by using automated trading rules. ARP funds are also attractive because of their lower costs relative to hedge funds; management fees are typically below 1% compared with management fees of around 2% for hedge funds, on top of their incentive fees. In addition, ARP funds have better liquidity than hedge funds, often with daily redemption
schedules rather than monthly or longer schedules followed by most hedge funds, which can also have lockups. For these reasons, ARP funds are often called “liquid alternatives.” ARP investments are starting to be widely available and can be accessed through two channels.
The first consists of private funds offered by asset managers to institutional investors, plus a limited range of UCITS (undertakings for the collective investment in transferable securities) funds. The second consists of risk premia products offered by banks (also called “broker/dealers”) in the form of total return swaps (TRS), which is a relatively new market available to institutional investors. The purpose of this article is to provide the first analysis of this bank risk premium (BRP) market and to compare its performance with that of hedge funds.
This market has expanded rapidly; Table 1 describes its current size. According to Albourne (2020), the total market amounted to $704 billion in a notional amount at the end of 2019.1 This amount is split about equally between asset managers and banks. Table 1 also shows a typical count of products within each asset class for bank products. This article focuses on the performance of bank products, for several reasons. First, these bank TRS give direct access to ready-made alternative risk premia. This market is very large, currently around $360 billion, with perhaps 20 active banks and thousands of products. Little empirical research has been published on it.2 Yet, this market is attracting interest. Indeed, the Standards Board for Alternative Investments (SBAI), an industry association, recently issued two “best practices” reports (SBAI 2020a; SBAI 2020b) on broker/dealer products covering product selection and backtesting.
Second, this market created access to a finely detailed set of risk premia across many asset classes and styles. This information should prove useful for understanding—even replicating—some hedge fund returns. Indeed, Fung and Hsieh (2004), who provided a seminal approach to explaining hedge fund returns in terms of only seven factors also surmised, “Research on additional hedge fund styles will probably discover other risk factors” (p. 71). The bank ARP market indeed provides a systematic way to explore additional risk factors and to sort through what has been called the “risk factor zoo.”3 Third, academic research on hedge funds has used factor returns computed without accounting for trading costs, borrowing costs, or market impact.4 This approach unfairly penalizes the performance evaluation of hedge funds because missing transaction costs generate artificially low alphas. In contrast, returns on bank products represent actual net flows to investors, so they necessarily account for all these costs. This accounting is essential for strategies such as trends, in which turnover can be very high, or when shorting single stocks, which can be quite expensive. This focus on the performance of bank products is in line with recent work on mutual funds arguing that omitting transaction costs in benchmarks or factor returns creates a downward bias for measures of skill in performance evaluation.5 For this analysis, I used a hitherto-unexplored data source consisting of BRP indexes constructed by Hedge Fund Research (HFR). These indexes aggregate some 1,100 bank products, providing good coverage of this space. This study evaluated the performance of such bank ARP products and compared them with hedge funds for the 2010–20 period.6 I expected ARP products to explain returns for hedge funds that followed quantitative strategies but to do so less for funds that focused on security-specific analysis. These results should be of interest to investors contemplating investments in ARP and in hedge funds.
Hedge fund investors are motivated by the view that hedge funds can produce alpha with a risk level substantially lower than traditional long equities. Many investors balk at the fees charged by the industry, however, even though what ultimately matters are net returns. The traditional model calls for 2% fixed management and 20% incentive fees. In practice, these fees have gone down steadily over time and are now at 1.4% and 18.4%, respectively.7 Even so, such fees represent a major drag on gross performance. The question is whether ARP investments can provide a good approximation to hedge fund returns. On the one hand, one should expect that the average ARP fund can provide only a fraction of the gross return for a hedge fund with the same broad strategy. For example, a long-short equity hedge fund can devote more resources to the portfolio management process, where value-added is created from both stock selection and factor exposure. Let us call this fractional replication an “efficiency” ratio of, say, r = 50%. On the other hand, the ARP fund benefits from charging a lower management fee—say, 1% flat. The question is, what is the net of these two effects, efficiency versus fees? To illustrate, Figure 1 shows net returns for the two approaches, hedge funds and ARP funds, for various levels of gross hedge fund returns and replication ratios. On the one hand, with perfect replication, r = 100%, the ARP fund must dominate as a result of lower fees. This situation is unlikely because it would quickly lead to the demise of hedge funds. Even with an imperfect replication of r = 75%, however, the ARP fund still dominates for the entire range of gross returns up to 10%. On the other hand, with r = 50%, the hedge fund dominates in the range of gross returns above 2%. For a gross return of +10%, for example, the net returns are 6.4% versus 4.0% for, respectively, the hedge fund and ARP fund. At the other extreme, at r = 25% or below, the ARP fund becomes totally uncompetitive. So, the comparison hinges on the replication ratio. The empirical analysis later will show that correlation coefficient, estimates of this ratio, are on the order of 50% for some hedge fund categories.
Bank ARP products represent the next step, strategies in which both the trading algorithm and execution are outsourced to the bank. For this outsourcing, the return on the trading strategy must be precisely defined in a “rule book”—a confidential, often lengthy, document explaining the precise details of the trading strategy so that it can be exactly reproduced. Valuation is typically carried out using closing prices, from exchanges or third parties, and applying fixed, prespecified transaction costs. This process leads to a time series of daily “benchmark” TRS prices that can be used for mark-to-market valuation and for entering or exiting the product. Ideally, the construction of the benchmark follows the principles of the International Organization of Securities Commissions (IOSCO).10 Such products typically offer daily liquidity. No industry standards have been set for such bank products, unfortunately. Generally, three sources of costs should be considered: (1) costs that are embedded in the index, (2) external swap fees, and (3) entry and exit costs. Banks can profit from charging transaction costs that are higher than actual direct trading costs. They also profit from the “external” swap spread, which covers financing as well as a reward for the banks’ intellectual property. Such spreads typically vary from 0 bps to 100 bps, depending on the product. Banks do list a “stated” spread, but it can usually be negotiated. In some cases, however, the swap fee is embedded in the product return. Therefore, a careful accounting of all costs is required when comparing products. Banks have a comparative advantage in offering such products because the transaction flows from these TRS go into their trading books and maybe crossed against other trades, so they do not bear all actual trading costs. Similarly, trading-intensive strategies, such as daily hedging of options or intraday trading, can be done efficiently within bank trading books.