Institutional investors who are traditionally conservative, such as pension funds, are now exploring allocating a larger portion of their assets under management to alternative or non-traditional investments. These include hedge funds, private equity, venture capital, real estate and infrastructure.
Chart 1 underscores the tremendous growth in this asset class. It shows that the hedge fund industry alone had US$3.15 trillion of total assets under management as of end-June 2019, more than double the figure at the end of 2011.
This is not just a US story; it’s a global phenomenon. In Asia, Japan’s Government Pension Investment Fund, the world’s largest pension, began investing in non-traditional asset classes in 2017. The fund targets to allocate 3% of its $1.4 trillion of assets to alternative investments in the next three years.
Malaysia’s Employees Provident Fund (EPF), which has traditionally been an investor in private equity, real estate and infrastructure, is placing up to 10% into real estate and infrastructure alone. The EPF’s investments in real assets include the Guoco Tower in Singapore, the Battersea Power Station development in London, and toll roads in Malaysia.
In Switzerland, according to a comprehensive annual risk study of Swiss pensions by one of its top pension consulting firms, Complementa, as quoted by IPE and top1000funds publications, “allocation to alternative assets has reached double digits for the first time, with investments in insurance-linked securities, private debt and infrastructure behind the growth. Around 9% of all pension funds currently exceed 15% in their allocation to alternatives”.
Canadian pension funds, including the Canada Pension Plan Investment Board and the Ontario Teachers’ Pension Plan, are some of the world’s most sophisticated institutional investors in alternative asset classes. Private equity and hedge funds each make up to around 20% of the average Canadian pension fund’s portfolio.
Some may argue that the current spate of institutional allocations to alternative investments, particularly in private equity, is just “hot money” chasing the asset class. In some cases, it has even been at the expense of underperforming hedge funds. Whatever the case, alternative investments is growing rapidly.
Fortunately, many finance practitioners are also believers in contrarian and behavioural strategies. In the popular factor investing space, the valuation factor (a.k.a. the Warren Buffet factor), which measures the cheapness of a security relative to its industry median, hasn’t been performing well over the last few years. It is a bread and butter factor in many quantitative equity funds.
Yet academic studies and empirical evidence to date indicate that the contrary is true in the long run. Perhaps this is now the time to be a contrarian and invest in beaten-up quantitative equity funds that bet on valuation.
Similarly, long/short hedge funds have been getting the rap for being underperformers of late but yet expensive. Adding to its woes, hedge fund strategies utilising high frequency, algorithmic or statistical trading techniques have been accused of being black boxes, and hence lacking transparency.
However, algorithmic trading-based hedge funds created by those with combined training in financial economics, investments and computer science do merit a closer look.
Most long-term investors, particularly asset owners like pension funds who are looking for good hedge fund managers to invest in, express a preference not only for low correlations to traditional asset classes such as equity markets, but also stability in investment strategy, risk-adjusted performance and human capital.
Chart 2 shows that the broad-based hedge fund index has had low volatility and correlation compared to the S&P 500 index. And the 36-month moving average performance in chart 3 reinforces our observation regarding the broad-based hedge fund index having lower volatility than the S&P 500.
Our proposition here is that an algo-driven hedge fund, which successfully marries skills in financial investments and computer science along with big data, could be the solution to the institutional investor’s quest for low correlations and stability in both returns and human capital.
The “robo CIO”
An algo fund is attractive for a number of reasons. Programme-driven strategies are systematic and disciplined. Those within scope are strategies based on sound academic research and science and which have been backtested and stress-tested over multiple market cycles to yield stable risk-adjusted returns, or alpha.
Such strategies preclude the “key man” risk issue, where the loss of a key person could affect confidence in – and hence result in the outflow of assets from – the hedge fund. All that matters for the firm’s business continuity is that a team of well-trained financial economists, computer scientists and big data analysts can translate investment research into a set of profitable hedge fund trading strategies using computer programmes.
Once this is achieved, there will also be consistent and systematic application of the algorithmic process globally, irrespective of investment universe, region, industry and/or employee location.
This may sound like a scene out of a Hollywood science fiction movie, but the chief investment officer (CIO) of such a hedge fund is also an algo. The “robo-CIO” dispassionately and scientifically selects the most optimal risk-adjusted strategies, constructs portfolios, controls leverage, creates the lowest-cost execution and trading strategies using state-of-the-art order management systems, and manages portfolio risk. Human intervention is only a last resort.
Meanwhile, humans are freed up to continually test new research ideas and strategies, refine and update the process, ensure compliance with the regulatory authorities, and perform client-facing activities of a humanoid nature.
Customisation is easy
An attendant benefit of algos is that fund subscriptions and redemptions – and therefore, fund openings and closures – are driven by systematic, process-driven programmes and algorithms. An algo fund, hence, will never become “too large” or take on opportunities unless the risk-adjusted alpha still exists. In other words, it will not be a consequence of capital inflows.
Finally, the customisation of portfolios to an investor’s desires and needs is a cinch using algos. This includes incorporating client-specific considerations concerning sustainable and responsible investing, be it environmental, social and governance, principles for responsible investments, sustainable development goals, taxes, compliance with shariah law in the case of Islamic bonds or sukuk, and so on.
For owners of long-term capital looking for hedge fund strategies that are systematic and disciplined, that can be customised for lower volatility and leverage, and automatically screened for quality, sustainability, “greening” and other criteria they care about deeply, the time of algo-driven hedge funds has probably arrived.
* Joseph Cherian is practice professor of finance at Singapore’s NUS Business School and academic adviser to Asia Asset Management. Manish Sansi is co-founder of Xen Capital and founder of Algante, a quant artificial intelligence research firm.