Alternative data and hedge funds - Creating new sources of alpha

The hunt for alpha is never-ending with one of the new battlegrounds being the acquisition and development of alternative data. Don't think of the development of new strategies, think of new information. A recent survey by Lowenstein Sandler, "Alternative Data = Better Investment Strategies, But Not Without Concerns" outlines the strong pivot to alternative data. 

82% of hedge funds surveyed are using alternative data and 42% of funds say their usage is extensive. Most firms are expecting to increase their research budgets to alternative data with close to 75 percent expecting to increase budgets by at least 11 percent.  

Their use of this data is wide ranging. Bigger firms are gathering this information in-house and not just buying from vendors. While alternative data is used to enhance existing research, larger firms are using new information to develop new trading strategies based on their new data. 

There are a few methods for generating alpha from research. All involve improving return predictions. Managers could gain alpha through better manipulation of existing data. The race to use machine learning could be considered the frontier for data manipulation. The other method for gaining an edge is through new sources of information that is unavailable to others. Since statistical techniques are in the public domain, alternative data can be viewed as a more effective way of creating a barrier to entry and unique advantage.

Craftsmanship alpha with trend-following - Providing the right amount of complexity

Where is the value-added for trend-following strategy? Many believe that the strategy has been commoditized, but there is a difference between simple trend-following and more complex strategies. Put differently, there is the difference between generic trend-following and skill-based trend-following. I have more than once argued that trend-following signals could be a flat fee business, but investors should pay a premium for risk management and portfolio construction.

This difference between generic and complex trend-following is craftsmanship alpha, the excess return that is generated from properly assessing and managing a portfolio of trends.

This value-added or craftsmanship alpha was nicely shown in a short paper called, "Protecting the Downside of Trend When It Is Not Your Friend" in the Journal of Portfolio Management. Simple additions of complexity can add significantly to return and risk control. The authors show that a little complexity to signals and better portfolio construction will lead to meaningful gains in the Sharpe ratio. Of course, as complexity increases the core trend-following risk premia can There is obviously little value-added if there are no trends; however, even in those cases, a well-diversified portfolio may limit losses. Adding some complexity will avoid momentum crashes when there are sharp trend reversals.  

The craftsmanship alpha for trend-following can be divided into two key areas, signal and portfolio construction. The simplest trend signal can be a single moving average across all markets. The simplest mechanism for portfolio structuring is to just form an equal volatility weighted portfolio. The choices made by the firm defines its personality and skill. This would be the craftsmanship of the manager.

From very simple cases, there can be increasing complexity. For example, a number of different timeframes for trends can be added. The style of identifying trends can also be diversified.  There also can be mechanisms for profit-taking or identification of periods susceptible for mean reversion. Trades can also be ranked or conditioned based different criteria. Risk management can be added to offset losses before a reversal signal is generated. 

In the case of portfolio construction, there can be a move from equal volatility to equal risk contribution. Managers can also account for the quality signals, correlation, and liquidity. A large number of tradable futures does not mean that all should be included in a portfolio.

There will be diminishing returns to complexity. Complexity will also increase trading costs. Investors pay for a manager's skill at making these trade-off decisions. The result can be measured, but the choices require a craftsman.

Portable Alpha - It can be generated with alternative risk premia

Portable alpha strategies are fairly old and have been used extensively to generate excess returns versus a benchmark with limited tracking error. Recent developments in the delivery of alternative risk premia through swaps make this concept fresh and offers investors a new opportunity to enhance their equity portfolios.

Portable alpha can come in a number of forms, but one simple but effective approach is to replicate a benchmark through the futures market but use alternative collateral as a a means of return enhancement. Assume that the SPX can be replicated with futures. $100 mm in SPX exposure can be obtained through futures and the collateral can be invested in fixed income assets that exceed the risk-free rate of return. The added yield will "enhance" the return on the index minus the cost of index replication. The investor now has an enhanced return with almost no tracking error versus the benchmark. However, there are more choices available to investors.

Given that alternative risk premia can be executed through total return swaps, their return streams can be blended with an equity benchmark to create a portable alpha strategy. The steps for the portable alpha strategy are straight-forward.

• Find an appropriate benchmark which can be traded in the futures. This is not a requirement but provides the simplest case.
• Build an alternative risk premia portfolio that has a low correlation with the benchmark and can be executed through total return swaps. 
• Blend the benchmark with the ARP swap portfolio to target an excess volatility or tracking error against the benchmark. A low correlation between the benchmark and ARP portfolio will allow the ARP portfolio to potentially add return without significantly increasing the benchmark risk. 
• Measure the excess return versus the benchmark against the added volatility to find an appropriate excess return versus portfolio tracking error.
• Measure the collateral needed for the swap portfolio and find the margin necessary to ensure that there is still full exposure to the benchmark.
• Form a structure that will allow for the joint holding of swaps, futures, and cash equity exposure.
• Track and monitor portfolio for deviations from expected return and risk.

There are obvious more details with implementing any portable alpha strategy but the concept of blending uncorrelated assets through swaps and futures is an appropriate way of improving a benchmark portfolio return profile.

Invesco Factor Survey 2019 - More investors are using and actively trading factors

The Invesco Global Factor Investing Study 2019 shows further increases in factor investing alongside active and passive strategies. Factors are not being used just to monitor risk, but to increase returns. Investors are making market beta decisions but also factor decisions like carry, value, and momentum. Managers are also not just investing in factor exposures but making tactical trading decisions on which factors make sense at any given time. 

Factor investing is not as large as fundamental active or market weighted investing, but the allocations are meaningful. Investors are dropping their fundamentally active exposures to hold these factor exposures.  Investors are making active timing choices between value, carry, and momentum and not just passively making allocations to core factor strategies.

There are still challenges for investors who use factor strategies, but the need is not education but analysis on timing. The greatest challenges are determining when to add or subtract a factor, forming return expectations, and monitoring the risk from factor exposures. 

Investors are accepting the scheme that asset returns can be decomposed into factors which can be exploited directly. The challenge is not determining whether to decide on factor exposures but how to make adjustment in exposures and determining when to change exposures. Factors are time varying and investors want to exploit factor opportunities no different than buying (selling) cheap (rich) assets.