Quantitative investment comes of age

Investors’ desire for diversified sources of return at low costs has powered demand for quantitative investment strategies (QIS). Investors should, however, take pains to understand the nuances of these products and ask challenging questions of providers before selecting a strategy.

This article identifies the characteristics of the most popular QIS iterations, their utility to asset managers and the ways in which they can be implemented within a portfolio. It also elaborates on certain selection criteria that should be considered by institutions that plan on implementing a QIS and highlight the risks they could face on their way.

 

Defining QIS

Buy-side firms are increasingly turning to QIS to complement their portfolios. Why? Because they offer two highly coveted features conventional investments typically lack: diversification and specific risk factor-based returns.

In addition, QIS can be constructed to perform specific functions and encompass multiple asset classes to suit a variety of investor goals. Indeed, the initialism is a catch-all for a series of asset allocation tools, three of which are elaborated on in this article.

 

Smart beta

Smart beta represents an elementary form of QIS. These strategies are designed to systematically select, weight and rebalance a portfolio in line with certain metrics other than an individual security’s market capitalisation.

Their objective is to combine the benchmark-beating performance of a well-crafted, actively managed portfolio with the low expense and simplicity of an index-linked allocation. The strategies are typically long-only and have been embedded in a wide range of exchange-traded funds in recent years.

A common smart beta strategy is equal weighting. Governed by such a methodology, a strategy would invest in each constituent of a chosen index as equally as possible, instead of allocating more to the securities with the highest capitalisation and less to the smallest. This decouples the strategy’s performance from that of the very largest securities – as in a market cap-based strategy – instead responding evenly to the price movements of each asset in the portfolio.

However, in doing so, it increases the exposure to small stocks, which are typically more volatile and vulnerable to sudden drawdowns than their larger peers. The superior returns possible through an equal weighted strategy might come hand-in-hand with higher risks.

This is by design – not accident. All smart beta strategies strive to outperform a traditional allocation methodology rather than subvert it. Thus, their performance remains correlated to the market at large.

Therefore, while they make attractive alternatives to the long-only portion of a portfolio, smart beta strategies cannot offer a high degree of diversification.

 

Alternative risk premia (ARP)

ARP strategies, on the other hand, can offer diversification, yield enhancement and lower risk exposure. These are investments that offer access to systematic sources of return that are well evidenced through empirical research and academic study.

Traditional risk premia are well known as the sources of return over the risk-free rate that can be achieved through a selected asset class. They include the equity risk premia – the returns accessible to an investor through buying into the stock market – and the term risk premia, which rewards those who invest for the long term.

In contrast, ARP returns are linked to other, discrete risk premia inherent across asset classes and markets.

Well-documented ARP, identified through academic research, include:

• Value – captures excess returns to securities that have low prices relative to their fundamental value
• Momentum – draws on securities with stronger past performance as evidenced by their relative returns over a designated look-back period
• Quality – harvests the additional yield characteristic of securities linked to companies with low debt, stable earnings growth and a strong balance sheet
• Curve carry – accesses return by investing at different points of a given futures or forward curve, usually in commodites.

Investing in an ARP strategy offers meaningful diversification as returns are statistically decoupled from traditional market risks such as bonds and equity. This is not to say an ARP allocation does not have its ups and downs – risk premia are subject to cycles of their own – but these may not synchronise with the highs and lows of the market at large.

It’s their capacity as diversifiers that make ARP strategies attractive investments, allowing asset managers to achieve higher expected returns for a given risk budget or similar returns for a lower risk budget.

An asset manager may select a desired risk premia exposure and have that govern the securities it buys and sells within a segment of its portfolio. ARP strategies can incorporate multiple asset classes and – unlike smart beta – seek to maximise factor-based returns through a mix of long and short exposures.

 

Hedging

The third broad category of QIS is systematic hedging strategies. These are programmatic allocation methodologies designed to protect investors from specific risks – market risk, interest rate risk and tail risk – efficiently.

Hedging QIS are not intended as yield-enhancing plays, or a means of diversifying asset managers’ sources of return, as with an ARP strategy. Instead, they can be used to safeguard returns by dynamically rebalancing a portfolio between risky and risk-free assets in response to predefined market signals.

The efficacy of a hedging QIS should be graded against two criteria: its ability to neutralise an investor’s exposure to unwanted specific risk, and its cost of carry – the expense incurred assuming the investment positions dictated by the strategy.

An exorbitant cost of carry could outweigh the hedging benefits of a strategy by depriving an investor of the very returns they wanted to protect in the first place. Efficient hedging QIS therefore attempt to achieve the investor’s desired level of coverage while optimising for cost of carry.

 

Focus on ARP

For those looking to generate diversified and higher expected returns, ARP strategies could prove especially attractive considering current trends, especially when correlations that have held true for a decade or more – and form the basis of investors’ diversification strategies – may be breaking down.

Consider, for example, the equities market shock that first erupted in October 2018 and caught fire in December, when leading bourses entered correction territory. As equities fell, so did bonds – an occurrence that undermined the negative correlation of the two asset classes that dated back to around 1998.

Such findings suggest established correlation assumptions are fraying. For asset managers wedded to a portfolio composed of 60% equity and 40% bond, this spells trouble, as the diversification such a mix intends to provide may not prove robust going forward.

Changes to macroeconomic policy could be responsible for the recent market disruptions, and signal further turbulence ahead.

The US Federal Reserve Board started a process of balance sheet normalisation in October 2017, under which its asset portfolio will shrink by as much as $1.9 trillion over five years. This will draw liquidity out of a market that has become used to easy money, and may increase short-term funding costs for banks.

Central banks are also adjusting low interest rate policies in place since the financial crisis, the effects of which will inevitably roil markets. As of June 2019, the Fed has hiked rates nine times since the end of 2015. The timing and frequency of further shifts, if any, will be determined as the macroeconomic and geopolitical outlooks clear. Other central banks have followed the Fed moves higher, with the Bank of England and the People’s Bank of China adjusting rates up this year. The European Central Bank, though it has kept rates static, has promised to end asset purchases by 2019.

Tighter monetary policy implies lower equity valuations, meaning the engine driving stocks higher worldwide could start to sputter.

It is against this tumultuous backdrop that ARP strategies show their quality. These can supplement any portfolio to provide the diversified returns sorely needed at a time when traditional asset prices appear fragile and correlations unsound.

Indeed, research by Cambridge Associates, a global investment firm, shows that a balanced portfolio with a 30% exposure to ARP funds would have yielded 12% between January 2007 and March 2009 – at the peak of the financial crisis – compared with only 3% for a traditional portfolio without this allocation.

As noted, this kind of performance is possible because ARP strategies diversify away unrewarded specific risks – such as broad market risks – and focus on delivering returns linked to fundamental risk premia.

 

Selecting an ARP strategy

The utility of ARP strategies is clear. But investment should be predicated on a clear understanding of how they are constructed and what drives their returns. A good strategy should be intuitive and readily explainable, and a worthwhile provider transparent on how the chosen premia are defined and accessed.

Take the following example of a carry risk premium strategy. This is designed to maximise the systematic returns investors receive in exchange for holding an asset and assuming its associated risks.

This ARP can be found across asset classes. In rates, for instance, the carry premium is an investor’s reward for taking the duration, interest rate, liquidity and credit risk linked with a defined bond position.

This premium is academically sound. Take the excess performance of a long position in 10-year US Treasuries over the past 20 years. Two-thirds of the performance came from coupon payments to investors – the specific, predefined reward for carrying US government exposure. The other third came from the decline in interest rates, a reward for market risk, which cannot be perfectly known in advance of investing.

A rates-focused ARP strategy seeks to monetise this carry by buying and selling rate exposures across currencies and tenors while controlling for idiosyncratic market risk. Essentially, it makes money by borrowing at a low interest rate and lending at a higher rate.

A practicable implementation of such a strategy would dynamically allocate between a basket of interest rate swaps linked to different currencies at a set tenor – say, the 10-year rate. Those swaps the trading strategy identifies as having the highest carry, after controlling for realised volatility, are bought, and those with the lowest are sold, with the difference between payments sent and received generating the return.

Greater diversification can be achieved by dialling up the number of currencies included in the basket, although this comes at the expense of performance, as the carry premium is eroded by spreading the rate exposure more widely.

This example is founded on a well-known, academically proven risk premium and can be implemented with simple, liquid instruments – in this case, interest rate swaps. These qualities make it readily explainable to an asset manager’s investment board.

 

Steps to implementation

Identifying a reliable ARP should be a prospective investor’s first priority. In recent years, a host of factors have been unearthed by quantitative analysts and academics. However, just a handful can be proven to have produced positive returns. In fact, a 2014 study of 600 factors found that 49% produced zero or negative risk premia.¹

The stable of risk factors with a sufficiently rich history to recommend them to investors is smaller still. For example, leading ARP think-tank Edhec-Risk Institute stands by just six risk factors: value, size, low volatility, high momentum, low investment and high profitability.

The burden of proving a factor’s efficacy lies with the ARP provider, but it is the responsibility of the asset allocator to decide whether a given amount and the quality of evidence is sufficient to greenlight an investment.

It is also the investor’s duty to understand how a given premium is defined by a provider and the process by which related exposures are selected by their provider. ARP with the same names are designed differently by each provider and can exhibit wildly different behaviour.

For example, ‘quality’ strategies rely on a variety of metrics to select appropriate securities – including but not limited to the issuing firm’s leverage, earnings stability, earnings growth, dividend strength and governance. Each provider will use a subtly different selection and weighting taxonomy to extract the ‘quality’ premium, resulting in variable return profiles.

Investors therefore need to be aware of the signals a provider considers when selecting assets to populate an ARP strategy, as well as those used to prompt portfolio rebalancing. Otherwise, they could accrue exposure to unwanted risks or miss out on a premium they want to access. They could also incur excessive trading costs if the chosen signals demand frequent asset turnovers.

 

Backtesting bias

A popular tool used by providers to guide investors’ ARP selection is the backtest. But investors should approach these with a healthy scepticism and scrutinise them for hidden biases.

Testing an ARP using historical data from a specific time period reflecting particular market conditions may show outperformance, but does not guarantee similar results in the here and now.

Take the example of ‘momentum’ strategies. The quantitative methods used may accurately select high-momentum securities over the backtesting period, but could deteriorate in live performance as the signals may have been optimised for that specific timeframe, not designed to be robust in all market conditions.

To avoid buying into ARP strategies that have been overfitted to certain data, investors should check backtests for these kinds of bias, and demand they cover extended timeframes embracing a variety of market regimes.

 

Forms of implementation

Managers have a range of means to gain exposure to ARP strategies, although not every wrapper is appropriate for all.

Cash-rich managers can invest in a dedicated ARP fund, gaining real exposure to the assets captured by the strategy. These are subject to management fees and potentially liquidity fees and redemption gates, which may obstruct an investor’s ability to withdraw their cash at will. In addition, investors do not have the option of altering the strategy once committed as implementation is fully outsourced to the fund manager.

Strategies can also be embedded within structured notes or certificates. These wrappers provide investors with factor exposure under a single Cusip number, and may offer additional benefits such as a principal guarantee or downside protection ‘buffers’.

A structured note wrapper may prove ideal for real money managers prevented from investing in synthetics. Although the return of a note is linked to the performance of the underlying ARP strategy, the investor is never the asset owner. Instead, the provider runs the strategy in-house and transfers its performance to the note by means of a swap.

Money managers permitted to use derivatives, and who are comfortable transacting them, may consider accessing ARP through a total return swap or similar instrument. Here, the investor gets pure exposure to the return of their chosen strategy, minus a set fee.

This structure requires no upfront cash investment, freeing managers to put their cash to work elsewhere. Swaps are generally of short duration, but can be rolled over with changed terms – such as different notionals – offering investors a flexibility they do not get with other implementations.

 

Costs of implementation

An essential consideration when choosing between ARP investments is cost. A central appeal of these strategies is that they can deliver returns above those of the broader market at a lower cost than an active manager or hedge fund allocation. However, returns can be crimped by excessive strategy and wrapper fees.

An ARP provider will typically extract payment for an investor’s right to access its intellectual property and to cover the costs of implementing and managing the investment via a deduction from the strategy’s returns.

Competition between providers has pushed these costs down in recent years, although multifaceted strategies reliant on a high degree of asset turnover and incorporating short positions are typically more expensive. However, an ARP strategy is still considerably cheaper than a typical hedge fund allocation subject to a ‘two and twenty’ fee structure.

The choice of wrapper will also influence an ARP strategy’s all-in cost. A fund allocation will attract management fees and transaction costs if the chosen strategy signals a portfolio rebalancing, whereas a swap implementation will incur costs in the form of a swap maintenance fee to the counterparty, as well as an index fee to cover the expense of the third-party calculation agents.

 

The ARP difference

Prospective investors should be wise to the differences between ARP strategies and traditional hedge fund allocations.

Active managers and ARP strategies both seek to extract returns from selected risk premia. However, the latter do so systematically in accordance with predetermined rules hardcoded into their governing algorithms. Unlike actively managed strategies, an ARP is free from human discretion – and interference – which means an investor always has perfect knowledge of how and why their money is being deployed.

Best practice among ARP providers is to outsource the administration of
a strategy wholesale to an independent third party, further strengthening
its integrity.

Naturally, conflicts of interest arise when a bank owns and operates a strategy that it sells to clients. If a bank’s traders are privy to when an ARP algorithm buys and sells, they have the ability to front-run the transactions and depress the strategy’s returns accordingly, to the detriment of the end-investor.

The presence of an independent calculation agent, alongside the provision of a full strategy disclosure document, can minimise these conflicts.

One other consideration remains for the prospective investor: the choice of a benchmark against which to judge an ARP strategy’s performance. The problem is that few appropriate benchmarks exists. The purpose of these strategies is to offer returns uncorrelated to market risk, meaning traditional benchmarks, including prominent stock and bond indexes, are not fit for purpose.

Furthermore, as each provider’s definition and means of accessing ARP varies, comparing the performance of one strategy with another can prove fruitless. One solution is to use an index consisting of all – or a large sample of – ARP strategies anchored to a particular factor. This way, an investor can determine whether a specific momentum strategy has, for example, outperformed or underperformed the universe of momentum strategies as a whole.

 

Conclusion

QIS unlock a wealth of possibilities for asset managers. ARP implementations in particular offer the attractive opportunity to generate additional returns and inject diversification into a given portfolio.

Potential investors should be wise to the processes that underpin the ARP lifecycle, from inception to construction, and favour strategies that are transparent, simple to explain and underpinned by in-depth, unbiased research documenting their ability to provide outperformance over the long term.

They must also be wary of the implicit and explicit costs of different strategies, and be sensitive to the type of implementation most suited to their investment mandates.

An asset manager that keeps these considerations in mind, however, should find much to celebrate after taking the decision to develop an ARP strategy.

 

Quantitative Investment: Uncovered – Special report 2019
Read more 

Sponsored content