The increased use of models for investing could create herding in risk management strategies, which may be less visible but just as dangerous.
Some stocks and strategies are hard to resist, even for fund managers who claim not to follow the crowd. Yet the hedging of these strategies can be equally crowded, putting pressure on liquidity and capital that can, in turn, create systemic risk.
BAT and FAANG stocks – two ghoulishly sounding acronyms that cover Chinese tech stocks Baidu, Alibaba and Tencent Holdings and the US giants Facebook, Amazon, Apple, Netflix and Google – continue to be the most invested stocks according to the Bank of America Merrill Lynch survey of fund managers from December 2018.
Asset managers’ devotion to the tech behemoths mean they dominate the stock markets. At the end of 2018, the FAANGs represented 28% of the S&P 500 and more than 50% of the S&P tech index.
The motivation to invest in such stocks is strong; technology is the bedrock of equity growth in current markets, the fundamentals are strong, and the future looks bright.
Yet at the end of 2018 FAANGs got walloped after a sharp sell-off. With the exception of Apple, total returns from the most heavily invested stocks since 2017 were negative.
Mahesh Narayan, global head of research and portfolio management at data analytics firm Refinitiv, says: “The downside of overcrowding into a stock or strategy is that not only do valuations get inflated, but also, what happens when there is a change? It creates more systemic challenges.”
Even investors with relatively short memories should be familiar with the consequences of crowding. Momentum strategies saw major drawdowns during the 2009 post-financial crisis recovery, while the more famous bursting of the dot com bubble at the turn of the century painfully highlighted the danger of investor herding.
Tommaso Mancuso, head of multi-asset at Hermes Investment Management, believes the sell-off in 2018 was yet another example of how crowding causes unpleasant market events.
“During 2018 the market experienced air pockets or sudden risk-off episodes,” he says. “While the triggers of these events can be diverse and hard to predict, the ferocity of the market correction is rooted in the crowding of both positioning and investment approaches.”
When this happens the hedging and risk management that investment managers have used will be stressed. The question then is whether the asset management industry is learning from its mistakes or whether there is a systemic failure that is doomed to plague investors until something changes.
Too close for comfort
A 2012 research paper, ‘Risk management issues in European equity funds’ from professors Andrew Clare, Natasa Todorovic and Miguel Corte-Real of Cass Business School, conducted in the aftermath of the financial crisis, identified a number of flaws in risk management across European active equity funds, notably that over 70% of the funds used identical risk management systems, and the authors observed “perhaps too much focus has been put on the quantitative side of risk management and the new generation of complex risk models, without considering the qualitative issues.”
Stefano Carnevale, investment analyst at consultants Quantum Advisory, says there is little divergence between asset managers’ risk models.
“The majority of the time the risk models are very similar… It would not be unreasonable to see a large number of managers with almost the same ideas or approach,” he says.
Carnevale adds that the trend to factor based or smart beta investing has shone an even brighter light on this identikit approach to risk modelling.
The factor-investment industry has enjoyed annual growth rates of 30% in assets under management (AUM) over the prior five years to 2018 and is worth more than US$1 trillion. It is no surprise then that it is seen as one of the most likely areas for herding behaviour which has potential ramifications for the wider investment industry.
In their March 2019 paper Herding in Smart-Beta Investment Productspublished in the Journal of Risk and Financial Management, Eduard Krkoska and Klaus Reiner Schenk-Hoppé, claim that factor-based investors are unaware of the risks from crowding.
They write: “Smart beta products are fashionable as never before and investors do not take into account potential herding into the same strategies, which may become overvalued and near capacity. Moreover, estimations of capacity are varied and by no means exact.”
Krkoska and Schenk-Hoppé are unequivocal in their view that investors must take crowding in smart beta strategies seriously. They argue that ‘due to the severity of its consequences, the risk of herding into an investment style is one which ought to be considered by all practitioners in their risk analysis and portfolio modelling’.
With the pain of 2018’s Q4 downturn still felt sorely by some investors there are signs the investment industry is taking the crowding challenge more seriously.
In July 2018, MSCI released an integrated factor crowding model in which it claims helps investors identify herding behaviour.
MSCI says there is evidence that crowding has contributed to a number of extreme market events and since there is no way of knowing when the next crash or burst bubble will happen, investors ‘may want to be aware of those occasions when crowdedness in a factor has become extreme’.
Raman Aylur Subramanian, head of equity applied research for Americas and EMEA at MSCI, says, “We provide crowding score cards so investors can see any herding mentality that is happening within the factors. It gives them a gauge on where asset managers might be following the crowd.”
Aylur Subramanian says MSCI will release more crowding models this year, notably one that focuses on investment in individual stocks, which should help active investors establish that their managers are piling into popular companies.
The challenge for investors, however, is not so much in the identification of crowding but in finding managers with suitable risk management processes to hedge the impact of a possible fallout.
Refinitiv’s Narayan says, “Investors need to know the extent of their exposure today. They do not want a bunch of metrics that are backwards looking. They want to know their exposure to risk and what the impact would be under certain conditions.”
Standing out from the crowd
It is perfectly possible using today’s technology for fund managers to test exactly how their funds would perform under any number of extreme events.
Sebastjan Smodis, global head of liquidity risk management at State Street Global Advisors, likens the firm’s risk modelling processes to ‘war games’ that test funds and strategies against a vast range of possible events.
“We use a multi-lens approach to find out whether all our funds could withstand a perfect storm. We play through numerous scenarios with investment teams and traders and some of our third parties,” he says.
Smodis says if there is a question mark about a fund or strategy, it is raised through the SSgA governance structure.
Smodis also refutes that all asset managers risk models are the same, arguing that State Street Global Advisors’ assumptions vary considerably from its major competitors.
“Are we all relying on the same assumptions?” he asks. “No, our risk management team would have their own assumptions based on the likely magnitude of a potential liquidity shock, participation rates and how investors would be impacted. We have a good idea of how the other managers use risk modelling and we know we are different.”
It makes sense for asset managers to follow the money; fear of missing out is a clearly a motivation for investing where everyone else is. If investors see that they missed out on returns enjoyed by everyone else, they will rightly put pressure on their managers to find out what went wrong. Yet time and again herding has proven detrimental to individuals, institutions and even global economies.
The end of 2018 showed that some asset managers still have not learnt from the mistakes of the past and it is likely that in 2019 – as investors face ongoing macroeconomic and political uncertainty – managers will need to finally demonstrate that they have the requisite systems and process in place that allow them to stand out from the crowd or at least insulate on the downside if they choose to follow it.