Ch8 Style Drifts: Monitoring, Detection & Control

1. Introduction
The detection, monitoring and control of style is perceived as a key preoccupation by hedge fund investors.
- transparency issue: disclosure of strategy drift monitoring
- style drifts should be avoided even though they do not necessarily have a negative impact on performance
- identified relatively easily and measured precisely, either from portfolio holding information or by making use of standard quantitative techniques based on portfolio returns

**Style Drift
- a change in the hedge fund's investment strategy from what was defined in the hedge fund's offering documents
- distort asset allocation and manager selection
- it will be difficult for most investors to detect style drift unless the manager reports its positions as well as its monthly or quarterly profit and losses


2. On Style & Strategies
For the detection of style drifts, investment styles may be identified according to certain parameters and these parameters may be monitored by the investor

**factors driving the performance of hedge funds
- stylistic difference explain a part of the observed performance differences among hedge funds
- return may be interpreted as a premium for an exposure to risk and risk factors that funds are exposed to enables the description of an investment style
- performance-generating process of hedge funds is complex and a linear exposure to factors based on the returns of standard asset classes is not sufficient to describe the risk taken by hedge funds

2-1. Investment Style (0f Hedge Funds)
- long/short position with substantial leverage
- style cannot be determined simply by examining holdings or correlations with indices. traditional investment style analysis must be modified to identify the risk factors to which the hedge fund is exposed
- due to the vast array of hedge fund investing strategies, there are a greater number of specialized or niche hedge fund style versus traditional fund style

3. Style Drift
Hedge fund investing deals primarily with skill-based investing. The potential investor has to understand the particular skill set of a manager and decide whether this skill set is compatible with the investment strategy. In addition, the investor has to assess whether the strategy of the manager fits in with their investment objectives.

Style and/or strategy drift can be defined:
- drift in the exposure to a predefined set of risk factors
- change in the overall quantity

The kinds of style drifts that have to be avoided are the unexpected style drifts that have not been communicated to, and agreed with, the investors for the following reasons:
- from the bottom-up perspective of manager selection and monitoring, the style drifts make part of the analyses conducted and the conclusions drawn during due diligence useless
- from the top-down view of asset allocation and portfolio construction, style drifts invalidate the assumptions made about risk-return profiles and may destroy the expected diversification benefits resulting from style and manager mix.

3-1. Due diligence
Investor should review all factors relevant to making a prudent investment decision.

3-2. Main reasons for style drifts
- poor market environment: poor style market performance
- asset flows & limited capacity of the strategy: excessive cash inflows
- underperformance with respect to their peers: poor manager performance
- large drawdowns: recent losses
- change in key investment professionals: personnel change (leadeship changes)
- change in market structure or regulatory environment

4. Detection, Monitoring & Controlling Style Drifts
- Detection & monitoring: embedded in the ongoing due diligence & risk management processes
- No single analysis gives the investor sufficient comfort to make a definitive judgement

4-1. Approaches for Monitoring & Detecting Style Drift
4-1-1. Monitoring risk factors
- The careful monitoring of the evolution and levels of the identified style- and strategy-specific risk factors = integral part of asset allocation process
- Change in the value of risk factors may precede a change in fund style

4-1-2. Returns-based analysis
- A time series analysis that confirms that the hedge fund has perfromed consistently with its stated investment objective in terms of risk, return, correlation with predefinced asset classes and benchmarks. (assessment of overall hedge fund performance over time & comparability with the strategy of the manager)
- performed during initial due diligence & updated within the context of ongoing due diligence & risk management
- the analysis of time-varying risk & return statistics: a good picture of the dynamic nature of hedge fund strategies
- the extent of style drift can be measured by the extent of changes in the sensitivities of the fund's returns to asset class index returns; significant changes = indication of changes in style and/or risk level of the fund

4-1-3. Analysis of performance attribution & risk exposures
- analysis of profit/loss attribution gives insight into where the performance is coming from
- components of fund return: asset classes, region and currency
- performance attribution analysis provides an insight into potential style drifts
- analysing exposure trends & comparing exposures between managers of the same style provides a cross-sectional and dynamic view of the risks taken by a manager

4-1-4. Peer group comparison
- managers are continuously assessed against their peers and ranked with respect to the various risk, return and performance measure => change in ranking has to be observed regularly
- ideal complement to the returns-based approach
- a cross-sectional analysis of different managers sharing a common investment style in exactly the same market environment
- linear regressions of the single hedge fund returns on the peer group average; alpha = manager's skill, beta = proxy for the level of leverage used relative to the peer group
- underperformance with respect to peers is a leading indicator of a potential style drift
- relative overperformance may be an ex-post indicator of abnormal risk level
- return break-outs (returns lie outside a predefined confidence interval): complementary use of the returns-based and peer group analysis; return break-outs detected during a returns-based analysis may be caused by change in market conditions or by a change in the type and/or level of risk taken by the hedge fund manager

4-1-5. Analysis of Positions
The analysis of single position cannot help to avoid style drifts. However, it may accelerate the detection of drifts and help to validate the accuracy of a manager's reported numbers.
- the style consistency of the fund manager can be examined via a detailed analysis of the separate holdings of the fund (Holdings-based analysis)
- position transparency is necessary

4-1-6. Communication with the fund manager
Ongoing communication with a manager allows warning signals of style drifts to be detected