Caroline Baron, Head of PowerShares ETF Distribution – UK has written a note on how real is the risk of factor crowding.She writes: “For any professional investor, crowded trades are both an occupational hazard and often a cause of sleepless nights.
“A crowded trade means you risk losing your competitive edge in a zero sum game, called investing in the stock market. Worse, it means you’re exposed to a potential liquidity shock if everyone heads for the exit.
“It’s ironic that concerns over crowding have now spread to smart beta, the term commonly used for alternatively weighted indices and the related investment strategies. One key selling point of smart beta has always been that it can help address the concentration risks inherent in the traditional method of weighting index constituents by their size (capitalisation).
“ It’s well known that cap-weighting can lead to indices becoming over-exposed to popular markets or sectors.
“A classic case is Japan’s 44% weighting in the MSCI World index in 1989 at the top of the country’s equity bull market (Japan’s index weighting is now a fifth of that level), or the outsized presence of financial stocks in cap-weighted indices just prior to the 2008 crisis.
“However, some market analysts have recently warned that similar risks apply to smart beta too, pointing out that certain strategies, including several focused on market “factors” like low volatility, momentum and quality, may now be classic crowded trades.
“In particular, certain commentators have argued that high dividend/low volatility smart beta strategies, which have attracted substantial inflows during 2016, are particularly risky.
“Invesco PowerShares provides a variety of smart beta tools via our exchange-traded fund (ETF) range: these include both factor strategies like high dividend/low volatility and ETFs tracking so-called fundamental indices, which screen stocks based upon fundamental accounting metrics. We believe strongly in the merits of both types of smart beta and work closely with professional investors to help them understand how the strategies could behave in different market environments.
“We have therefore followed the debate about smart beta risks with great interest. Nevertheless, we think the warnings of a market-wide crash because of factor crowding are exaggerated. Here’s why.
“First, crowding is particularly dangerous when leverage is also involved. This occurred with the 1998 collapse of hedge fund Long Term Capital Management (LTCM), whose positions in currency and bond markets were amplified via borrowing. Worse, it turned out that the positions had been mirrored by the banks trading with the hedge fund.
“When market conditions turned against LTCM, liquidity vanished and it proved unable to exit its trades. The subsequent collapse led to a central bank bailout of the banks, though LTCM failed.
“Another example of the risks of adding leverage to crowding is the so-called “quant crisis” of August 2007. Here, many leveraged hedge funds had taken long and short positions in the same stocks, creating sharp price moves when a trend reversal took place. Again, several hedge funds were put out of business.
“But we don’t see any evidence that dangerous leverage levels are currently associated with factor strategies like low volatility, quality or high dividends. As far as Invesco PowerShares is concerned, we don’t offer leveraged ETFs in our product range.
“A second contributing factor to the 2007 quant crisis was a lack of diversity in quantitative investment models. It turned out that many of the quants had been using the same software, arriving at the same long-short pair trades.
“In the case of smart beta, there’s significant diversity in index approaches, with competing index firms promoting factor strategies that may have similar names but are often subtly different in methodology. This makes comparisons between indices harder, but we believe it reduces the risk of investors duplicating each other’s positions.
“When warning of a smart beta crash, certain analysts point out the rising valuations of certain smart beta strategies, for example quality and low volatility, as a result of investor inflows. They suggest that a reversal may be imminent.
“If we look at the S&P 500 Low Volatility High Dividend index, which is tracked by our PowerShares S&P500 High Dividend Low Volatility UCITS ETF (HDLV), currently we don’t see cause for alarm over valuations. The index had a forward price/earnings ratio of 15.0 after its latest, July 31 2016 rebalancing, a valuation discount to the broader equity market, represented by the cap-weighted S&P 500 index, at the same date.
“A fourth concern may be that a naïve approach to smart beta index construction could get you into trouble. For example, a company with a record of consistent past dividend payments may represent a so-called value trap—when a high yield masks a weakening balance sheet and growing risk of a dividend cut. A thoughtful approach to selecting value stocks should therefore include some kind of mitigation against such traps.
“In the case of HDLV, the index methodology screens first for the 75 highest-yielding stocks in the S&P 500, then excludes the 25 with the highest historical volatility, providing some mitigation against value traps. A cap on individual sector weights offers potentially an additional safeguard against crowding.
“As an example of how this works, some of the highest dividend yields within the S&P 500 at the end of June 2016 were in Nordstrom (16.6%), the Williams Companies (11.6) and Seagate (9.6 per cent). However, the historical volatilities of these stocks—38 per cent, 89 per cent and 57 per cent, respectively – meant they were excluded from HDLV.
“The investment approach underlying strategies like HDLV can therefore be seen as a potentially sensible compromise. Rather than focusing on the best absolute returns, such strategies seek to compound returns over the long term via a steady stream of above-average dividends, while attempting to filter out potential “falling knives”.
“We don’t want to downplay the importance of crowding risk, especially in an era where central bank interventions have created an intense appetite for yield. But in the case of smart beta, we think crash concerns have been overdone.”