Theory Guides, Experiments Decide.

“Experiment is the only means of knowledge at our disposal.” — Max Planck

Experimentation has been the bedrock of scientific research and progress for centuries. It is at the very heart of the Scientific Method.

Fuelled by new digital tools, the use of controlled experimentation is increasingly widespread in the business world too. For example, a growing number of corporations routinely use experimentation to rapidly test the impact of user interface variations on real customers and prospects, using techniques such as A/B testing.

At ALPIMA, we combine the latest advances in technology with a scientific approach to help our clients improve the way they design products, build portfolios and engage with their customers.

The ALPIMA platform is used by our clients to experiment and evaluate ideas in many ways. Below are a few examples.

(i) Which allocation methodology performed best on our model portfolios historically, with what parameters? Specifically, did Hierarchical Risk Parity produce better results than traditional methodologies?

Providing a comprehensive response to such questions can be very time-consuming. By allowing clients to rapidly iterate, we help them save considerable time and generate valuable insights.

Below is a chart showing various allocation methodologies on a return/risk diagram which took seconds to generate on our platform. By repeating this analysis periodically, one can ensure it remains current as market conditions change.

Chart 1a — Instant comparison of several allocation engines on a multi-asset portfolio

In the same vein, the chart below shows the impact of applying constraints to a particular portfolio.

Chart 1b — Impact of portfolio constraints on the efficient frontier
Chart 2 — Hierarchical Risk Parity (HRP) tested on a global multi-asset portfolio 
In this example, HRP, which is highly effective in certain cases, produces lower risk-adjusted returns.

For complex strategies with a large number of moving parts, we resort to machine-learning to further speed up the process and help our clients quickly identify regions scoring favorably on the metrics of their choice, such as Sharpe ratio, draw-downs, tracking error, or a combination thereof. The results typically look like the chart below, shown here for two parameters for easy visualization.

Chart 3 — Performance scoring visualization across two variables

(ii) What is the impact of changing the re-balancing frequency of our model portfolios?

The trade-off between re-balancing frequency and trading costs is easy to understand, but can be difficult to analyse precisely. The more frequently a strategy re-balances, the better it can adapt to changing market conditions. Re-balancing more frequently, however, can increase trading costs, which erodes performance. The table below, which shows the ex-post impact of different re-balancing methodologies on a multi-asset portfolio took a few minutes to be generated on our platform.

Chart 4 — Table showing the impact of different rebalancing methodologies on a multi-asset portfolio

(iii) What if we added crypto assets to our existing portfolio? What would the impact be on risk and performance ? What would the impact be on sales ?

Since crypto assets have a very short history, any historical analysis must be handled with care and adequately disclaimed. This being said, using the ALPIMA platform, one can easily evaluate the impact of adding crypto-linked securities to a traditional portfolio in seconds. The charts below show the results of such a test conducted on a dynamic portfolio of bonds and equities.

Chart 5 — Table showing the impact of adding crypto-linked assets to a traditional portfolio

Interestingly, by allowing such a test portfolio to be generated quickly with corresponding dashboards and fact sheets, the ALPIMA platform enables businesses to efficiently test new ideas with sales teams and customers too, which helps to assess the overall business impact of adding new assets to traditional portfolios. Based on the gathered feedback, an informed discussion can be had as to how to proceed.

The ALPIMA platform is used in many more ways. Listing them all would not fit in this piece. By enabling personalised strategy design, rapid iteration and intuitive visualisation, the ALPIMA platform allows CIOs, product and client-facing teams to evaluate ideas, old and new, with unmatched clarity and ease.

“It has been just so in all my inventions. The first step is an intuition — and comes with a burst, then difficulties arise” — Thomas Edison

The importance of experimentation in business cannot be overstated, and recent advances in technology and data science make it possible to experiment much faster and better than before.

This has profound implications for the way investment solutions are created and delivered in the digital age as investors, with more and more data at their disposal in their daily lives, demand more evidence from their financial providers going forward.

Contact us to find out how ALPIMA can help you use experimentation to grow your business and transform client service.