Improvements to the inbestMe Investment Comparison Service

At the beginning of 2021 (4 and a half years ago), we were pioneers in launching our Investment Comparison Service, which remains operational today.

The conclusions we drew after the first year remain, in general, still valid today:

• most of the portfolios analyzed from our clients could be improved: inbestMe portfolios offered a 5% higher return (CAGR), better risk-adjusted returns, and total costs up to 3.8 times lower.

Over the past year, we have made improvements to this service (V2), which we summarize below.

Automation of the analysis and the report

The most significant effort has focused on making this service sustainable and ensuring an acceptable response time.

To achieve this, an interdisciplinary team—specifically an investment analyst and an inbestMe developer—worked closely together to automate the process.

This required many hours of analysis and testing.

Currently, the tool is capable of generating a report without human intervention in approximately one minute.

To do this, our process collects portfolio data and connects to the Bloomberg API to automate the entire data extraction process, ensuring reliability. We have made a major effort to ensure data reliability. Specifically, we never use averages; we use real data from the analyzed funds combined with the information provided by the client.

Improvement in determining the costs of the analyzed portfolio

An important aspect is determining the costs borne by the portfolio.

In this regard, we collect the real TERs of the funds in the analyzed portfolio whenever possible. If these are not available, we use management fees as an alternative (which are usually slightly lower). There may be slight inaccuracies in the nominal estimation of portfolio costs, but this does not affect the return calculation, since the NAV (Net Asset Value or price) already reflects these costs.

We have also added a field where the client can enter other portfolio costs: here they should include management fees, custody fees, and any other costs borne by the portfolio.

By default, a cost of 1.5% is applied, estimated based on prior comparison experience, unless the client modifies it. It is advisable for the client to confirm or change this figure, as it can significantly affect the analysis.

Ability to run a dual analysis comparing by volatility or by equity exposure

After dozens of analyses, we have found that basing the comparison solely on equity exposure could lead to inaccurate results.

The automatic execution that the user or client can run will report a comparison based on matching by volatility. We believe this comparison yields more homogeneous results.

In any case, the client may contact our customer service team, who will be happy to generate the alternative report and discuss both in detail (this complementary service will only be available for portfolios ≥ €50,000).

Removal of minimum thresholds

Thanks to these improvements and, above all, the automation of the service, we can now make it available to all users regardless of portfolio size.

Therefore, we have removed the €50,000 minimum to perform the analysis, provided that the client initiates the execution themselves.

They simply need to access the Investment Comparison Service page and manually enter the funds (ISIN and amount, up to a maximum of 5) or, preferably, upload a file with this information.

The automation of the service has forced us to eliminate pension plan analysis due to the lack of reliable and systematic access to information. We will explore whether it can be reinstated in the future.

Improvements to the data collection page and report format

In line with the above, we have introduced minor improvements to the page to make data entry more intuitive.

Improving this user interface will become a priority over the coming months.

We are aware that there is still work to be done in this area.

We are also aware of the difficulties our clients face in capturing the required information, and we are thinking about solutions to this problem. Nevertheless, we believe that the user/client who truly wants to analyze their positions should be able to extract information as simple as (with or without help from their current provider):

• ISIN/Amount
  or
• ISIN/Weight (in this case, a €100,000 portfolio will be assumed)

The automation of the process requires that the client be able to extract and report this information.

We have also taken the opportunity to align the format with the new rebranding carried out in September 2024.

Future improvements

In addition to the above, we have identified other potential improvements, initially defined as:

• V2: will recover and integrate metrics that had to be removed in V1.5 (regional weights, currencies, Sharpe ratio)
• V3: will integrate user interface improvements

We expect these versions to be operational in early 2026.

In the meantime, if you wish to analyze your portfolios, do not hesitate to use the service.

Important note. Limitations of the Investment Comparison Service

During the analysis and development of this new version, some limitations of the comparison process have been identified or confirmed, which we summarize below:

1. Static analysis:
   
   The analysis is based on a snapshot of the portfolio at a specific point in time. If the portfolio has changed over time (improving or worsening), those changes are not reflected in the report.
   
2. Limited time horizon:
   
   The analysis only covers the period for which data is available for all funds. For example, if a fund has only two years of history, the report will be limited to those two years.
   
3. Fund costs:
   
   We use the TER provided by Bloomberg. If it is not available, the management fee is used. If Bloomberg data is incorrect, this will affect the reported figure but not the return, which is calculated using real prices (NAV).
   
4. Management and custody costs:
   
   If the client does not enter these costs, a default 1.5% is assumed. It is recommended to enter them correctly, as they can significantly influence the result.
   
   Institutions are required to send an annual report detailing these costs (usually in the first quarter of the following year).
   
   Important: do not include fund costs in this figure, as they are already deducted from the NAV and are reflected separately.
   
5. Unaccounted biases:
   
   Currently, the analysis is not adjusted for portfolio biases (for example, SRI). The comparison is made against the closest standard inbestMe portfolio.
   
   This can be useful (objective view) but may also limit comparability (non-homogeneous portfolios).
   
6. Other limitations:
   
   Other limitations may exist that have not yet been identified. Therefore, we recommend reviewing the results carefully and with a critical mindset.
   
   If you have more than €50,000, you can request a meeting with our team to discuss the report.

Academic research driving the evolution of the comparator

As a complement to these practical improvements, recent research developed internally at inbestMe has been published in a leading international academic conference. The work, titled “Analyzing Similarity Across Investment Portfolios: A Multi-dimensional Approach” and published in the book Innovative Perspectives on Computational Intelligence and Data Science, proposes a multi-dimensional framework to analyze similarity across investment portfolios beyond traditional approaches based solely on equity exposure or volatility.

By using advanced metrics that capture both portfolio composition and behavior in terms of risk, return, and market sensitivity, this research opens the door to future extensions of the fund comparator toward richer and more accurate analyses. While the work is primarily academic and experimental in nature, it reinforces our commitment to evolving our tools based on robust, validated methodologies with real-world application potential to help our clients invest better.