The Gateway to Algorithmic and Automated Trading

Going the alpha mile

Published in Automated Trader Magazine Issue 13 Q2 2009

So you’ve beaten the competition. Great.  Now, have you thought about how you’re going to stay ahead?
Phil Perkins and Ben Van Vliet have seven suggestions for high fliers with a fear of falling.

Competition for market opportunities increases every day. Maintaining your edge against market competitors requires as much effort off the trading screen as on it. This can be seen across the industry as firms increase spending on better models, faster technology, and increased cost controls. Developing, updating and replacing the existing portfolio of trading strategies, becomes a critical component of your long-term viability in the markets.

Over the long-term, the returns on your investment portfolio are the direct result of how well you optimize and deliver on investments in front-office strategies and capabilities. The only alternative is to be overtaken by competitors that are better on the screen because they are better off the screen. In this article, we outline seven recommendations for maintaining your edge and creating what we like to call Sustainable AlphaTM.

• Manage your project pipeline like an investment portfolio

Automated trading can learn a lot about sustainability from the pharmaceutical industry. The entire pharmaceutical industry is built on making speculative investments in research on a pipeline of drug candidates. These investments cover high upfront costs and assume a high potential of failure. Scientists at the company develop a hypothesis, perform experiments and ultimately deliver results. With good fortune, some of these drug candidates become commercially viable; most do not.

For pharmaceutical companies to stay competitive, they must improve the flow of new products into the developmental pipeline, cut poor performers quickly, and balance spending to optimize the number of significant launches each year. To do this, the industry takes a portfolio approach and effectively creates an efficient frontier that maximizes the trade-offs between expected rewards and risks of pursuing certain drug candidates.

Consider these comments by Laura Wood, a Senior Manager at Research and Markets regarding a major pharmaceutical company. "In March 2008, the company outlined its growth strategies, which include optimizing its patent-protected portfolio; generating revenue opportunities from established products; accelerating growth in emerging markets; focusing on continuous improvement and innovation; and investing in complementary businesses. The company also re-affirmed its 2008 guidance of a revenue range from US$47 to US$49 billion, with adjusted research and development expenses ranging from US$7.3 to US$7.6 billion."

Can you talk about your firm's results with such clarity?

Managing how and when to make investments in strategies and capabilities is a series of tradeoffs among maximizing the expected economic returns, minimizing risk, and ensuring a diversified approach. The portfolio must be large enough to compensate for failures, but not so large that it overextends resources. Uncertainty and resource limitations make this a challenging problem, because competition among projects for limited resources must not result in extended delivery times on important projects.

To create a portfolio perspective requires taking five steps:

Step Description
Build a project inventory Each business unit builds up an inventory of projects and proposals. Creating a consolidated project inventory across business units gives your firm the ability to determine the overall complexity and risk of the project portfolio. It also allows you to assess how well capital expenditures are matching investment goals
Assess project interdependencies Project interdependencies can cause delays when one project is waiting for another project to complete. There are four types of dependencies to manage:
• Resource dependencies occur when multiple projects compete for the same scarce resources such as the availability of certain skills or access to a testing environment
• Cost dependencies occur when aspects of different projects are combined into a single project
• Opportunity dependencies occur when two strategies target the same market opportunity and will cannibalize each other's returns
• Technical dependencies occurs when technical success on one project affects the technical success of a later project
Assess project value There is no single definition of the 'best' project, but the project valuation method must create the ability to compare project complexity and expected benefits. The valuation method impacts project selection and prioritization.
Quantify project risks Every project has risks, so it is essential to identify mitigation and contingency plans. Identifying project risks gives stakeholders advanced notice of problems that might impact the project's budget, quality, or timelines.
Prioritize and approve projects in the pipeline Projects will have varying short and long term benefits, specifically regarding their synergy with the investment portfolio. The combination of valuations and risk assessments will allow for meaningful selection, comparison, and prioritization of investments in strategies.

• Focus on the capabilities that drive an edge in the marketplace

The project portfolio will be comprised of projects that offer widely differing values but must collectively strive to achieve the same overall strategic objectives. You can only create an operational edge if you are investing in capabilities that give you a direct advantage in the marketplace. In our previous article, "The Five Drivers of Profitability" (Automated Trader, Q1 2009), we defined competitive edge in systematic finance. The five ISOF [Institutue for Market Technology Strategic Operations Framework] goals that make up competitive edge outline the core set of capabilities that your firm and your trading systems exploit in the market. [Find this article at - Ed.]

ISOF goals encompass the things that individual traders may not think about on a daily basis, because a trader's job is to focus on the daily execution of trading decisions. These goals are not bounded by the financial year or changed each time the annual plan is updated. They outline the core set of capabilities that your traders exploit while competing in the marketplace.

In Figure 1, we outline the ISOF profile of four firms. The Market Maker is a risk arbitrager that makes money focusing on Order Execution Speed and readily admits it has no sophisticated Strategic Information Advantage. The Hedge Fund, on the other hand, makes its money through very sophisticated models, but it does not require significantly advanced Order Execution Speed.
Understanding your ISOF profile can focus your project pipeline. If you are a small firm using standard strategies, you may quickly find yourself losing to larger firms with significant resources invested in fast execution. To compete, your smaller firm may need to differentiate by improving its Strategic Information Advantage (i.e. does the strategy have a special sauce?) or Model Portfolio Management (i.e. how well do you reallocate capital to strategies at the time they can best utilize the money?).

Figure 1: The Five Drivers of Profitability (ISOF Profile)
Figure 1: The Five Drivers of Profitability (ISOF Profile)

• Create a Project Investment Committee (PIC)

Well-capitalized firms have investment committees responsible for oversight of the firm's investment portfolio, reviewing investment performance, evaluating investment risk, and managing the firm's capital structure. Managing investments in your project portfolio requires the same oversight. Putting a Project Investment Committee (PIC) in place to coordinate and oversee the firm's portfolio of projects accomplishes the same goal as a traditional investment committee.

The PIC is responsible for monitoring all aspects of trading system development projects. High value projects are clearly the most sought after, but their risks, if too high, may dilute their attractiveness. Conservative projects may quell fears of losing an investment, but if the returns are too low, they can undermine the company's vision. The PIC will not chase the opportunity that looks best in isolation, but consider its impact on your organizations broader portfolio. A PIC's global view will facilitate optimization of resource allocations.
The PIC should include the heads of each trading desk as well as key members of risk management, operations, quantitative research, and IT. The PIC should meet regularly to review the progress of projects in the portfolio as well as the performance of live trading systems against expected returns to validate the models met the needs of the investment portfolio.

• Align the project portfolio with the investment portfolio

The PIC should ensure the project portfolio delivers results that directly sustain the returns targeted by the firm. The PIC should adopt a clear project investment strategy that includes a reasonable set of assumptions about the organization's risk tolerance and the investment portfolio's expected results.
Every part of the organization must then rally behind the PIC's strategy.

Your firm can better enable this rally by understanding and coordinating the interrelationships among organizational structure, business processes, and technology. Figure 2 outlines our Enterprise Architecture Framework and the components that make up your firm.

Enterprise Architecture will give your firm the ability to rapidly improve efficiency and effectiveness, including innovations in the structure of your organization, the centralization or de-centralization of trading activities, the quality and timeliness of market data, or ensuring that money spent on information technology is justified.

Firms that don't manage Enterprise Architecture may find that traders can't accomplish their objectives because IT does not fully understand their needs. If IT reacts to trader requests without understanding the broader goal, time and money may be spent on systems that do not work together.

Figure 2: Enterprise Architecture Framework
Figure 2: Enterprise Architecture Framework

• Align the organization's hierarchy with organizational goals

We recommend structuring your organization to focus on ISOF goals. If your firm is focused on increasing Order Execution Speed, then you should have a top manager responsible for performance in that area. By assigning responsibility and empowering that person to affect change, this manager can ensure that all of the systems are leveraging the best practices available.
Many firms can have greater success introducing horizontal management. As shown in Figure 3, managers with horizontal purviews own business processes and are responsible for the performance they generate. They work across business units, break through organizational barriers to drive a quality end-to-end solution, and ensure a consistent approach to maximizing organization goals. This team must report to the same person as the heads of the trading desks to ensure they have the same support as the other front office units.

• Define a gate-based project approach and valuation methodology
Given an investment portfolio, the financial markets will dictate performance. With a portfolio of trading strategy development projects, management must define its own performance metrics and collect its own data. Evaluating a project must take into account the probability of trading and technical success, contribution to firm-wide alpha, and cost so management can build a realistic picture of the project pipeline.

Figure 3: Horizontal Organizational Structure
Figure 3: Horizontal Organizational Structure

In Figure 4, our K|V project management methodology outlines Design-Test-Implement-Manage (DTIM) stages that identify how and when to collect this data. We like to think of projects as a type of structured product where capital investments are tied to stages in your project management method (See our article "Rapid Model Deployment" Automated Trader Q4 2008). By evaluating each project according to its potential and cost over the short and long term, the PIC is able to build a realistic picture of the project pipeline.

Figure 4: K|V Methodology
Figure 4: K|V Methodology

K|V outlines project gates that identify the critical investment decisions for a project. Gates are predefined PIC meetings where go/kill decisions are made, weak projects are weeded out, and scarce resources are reallocated to more promising prospects. These gates empower the PIC with greater control over the project portfolio.

• Cut bad bets quickly and reallocate capital to good projects

It is common wisdom in trading to cut your losses quickly and let profits run. The same is true of your project portfolio. The odds of an already failing project getting worse are much larger than it getting better.

The project also becomes significantly more expensive as it progresses from one stage to the next (the cost of backtesting a model to prove its performance is small compared to the cost of building a full trading system).

Your firm must manage its project portfolio with the same discipline as a trader managing his trades.

Cut your losses quickly and reallocate capital, labor, and other resources to projects that are succeeding. It can be hard to step away from a project that has already taken up a lot of capital, but the alternative isn't smart investing, it is simply a gamble.

Figure 5 shows six trading system development projects over a period of time. As projects B, C, and E are canceled due to low odds of trading or technical success, resources are reallocated to project A, D, and F. After the fourth month, the portfolio of trading systems has evolved to a more efficient point on the frontier.

Figure 5: Capital re-allocation
Figure 5: Capital re-allocation


Creating a controlled pipeline focused on an ISOF profile is critical to creating Sustainable Alpha™. Your long-term investment portfolio returns are the direct result of how well you optimize investments in front office strategies and capabilities, therefore developing, updating and replacing the existing portfolio of trading strategies becomes one of the most critical components of long-term sustainability.