2018年-FCA英国金融行为监管局_ms17_1_2_annex_5_16页_1022kb

Summary of MS17/1.2: Annex 5 - Model Portfolio Review July 2018

Core Content

This report is part of the Investment Platforms Market Study and focuses on the model portfolio review conducted by the Financial Conduct Authority (FCA) in July 2018. It aims to assess whether non-advised investors can make informed investment decisions based on model portfolios and whether the naming conventions used by platforms help investors compare and understand the risk levels of these portfolios.

Main Issues Analyzed

• Risk naming conventions: Whether the labels used (e.g., 'cautious', 'balanced', 'aggressive') accurately reflect the risk level of the model portfolios.
• Charges and net returns: How much investors pay to use model portfolios and whether higher charges correlate with better or worse net returns.
• Asset allocation and risk categorization: The relationship between asset allocations and the implied risk level of the portfolios.

Key Findings

1. Growth in Model Portfolios

• Model portfolios have grown significantly, especially on adviser platforms, where AUA increased 25 times from £0.9bn in 2011 to £22.5bn in 2017.
• D2C platforms also saw growth, with AUA increasing three-fold from £4.5bn to £14.1bn.
• In consumer research, 17% of non-advised investors used model portfolios.

2. Risk Naming Conventions

• There is no consistent industry-wide standard for categorizing portfolios by risk.
• Similar labels (e.g., 'cautious', 'balanced', 'aggressive') often correspond to different risk levels, leading to confusion among investors.
• For example:
  • Portfolios labeled 'adventurous' or 'aggressive' had IR categories ranging from 3 to 6.
  • Portfolios labeled 'moderate' or 'balanced' had IR categories from 1 to 5.
  • Portfolios labeled 'cautious', 'conservative', or 'defensive' had IR categories from 1 to 3.
• The overlap in IR categories across similar labels indicates that naming conventions may not reliably reflect the actual risk level.

3. Asset Allocation and Risk

• The FCA assigned Imputed Risk (IR) categories (1–6) based on asset allocation data.
• Asset allocation variations were significant, especially in total equity and total bond.
• For instance, medium-risk labeled portfolios showed a wide range in bond and equity allocations, with some falling into low or high risk categories.
• Around 20% of medium-risk portfolios were classified as low risk (IR 1 or 2), and 42% of comparable firms' medium-risk portfolios were also classified as low risk.

4. Charges and Net Returns

• A wide range of average charges was observed, from 0.05% to 2.5%.
• Passive model portfolios had lower average charges (0.15–0.85%) compared to active ones (0.9%).
• Unitised portfolios generally had higher charges than non-unitised portfolios.
• Net returns were negatively correlated with average charges, with higher charges associated with lower risk-adjusted returns (measured by the Sharpe ratio).

5. Risk-Adjusted Returns

• The Sharpe ratio showed that low-charges portfolios had higher net risk-adjusted returns than high-charges portfolios.
• The relationship was statistically significant at 99% confidence level.
• Non-unitised portfolios also had higher net returns than unitised ones, consistent with the lower charge levels.

6. Limitations of the Analysis

• The analysis is based on historical data from 2012 to 2017, which reflects a stable economic environment.
• The benchmark approach was not used due to lack of consistent benchmark data across portfolios.
• The analysis focused on specific naming conventions and did not cover portfolios with labels like 'growth' or 'income'.

Sample Size and Data Sources

• Total portfolios analyzed: 797 (including both platforms and comparable firms).
• Data sources:
  • Firm submissions provided general information, AUA, charges, and asset allocations.
  • Morningstar Direct provided net returns and charges for unitised portfolios.
• Exclusions:
  • Bespoke non-unitised portfolios.
  • Non-unitised portfolios rebalanced by the platform.
  • Third-party model portfolios hosted on platforms.
  • Portfolios with insufficient data.

Future Work

• The FCA plans to:
  • Explore whether investors can assess risk without relying on naming conventions.
  • Investigate the main drivers of charge differences between model portfolios on platforms and those from comparable firms.

Conclusion

• Model portfolio naming conventions are inconsistent and may mislead investors.
• Higher charges are associated with lower net risk-adjusted returns.
• Unitised portfolios tend to have higher charges and lower net returns compared to non-unitised portfolios.
• The FCA recognizes the need for further analysis to understand how risk can be assessed without naming conventions and to determine the factors influencing charges.