Introduction

This paper analyses over 4,700 private meetings between a large active asset manager and portfolio firms. Using proprietary data from Standard Life Investments (“SLI”), the study links private meetings with trading decisions across a nine-year period (2007–2015), examining whether meetings generate valuable information and profitable trading opportunities, and how these activities fit within the regulatory framework on fair disclosure.

Prior literature

The paper draws on literature examining investor-firm interactions, proprietary engagement data, information advantages in equity markets, analyst recommendations, and blockholder trading theory.

Opening the active asset manager “Black Box”

Standard Life Aberdeen was the 23rd largest asset manager worldwide and the largest active asset manager in the UK in 2015 (p.9). The UK equities desk comprises 18 fund managers overseeing between 43 and 64 UK funds (p.10), alongside 23 governance specialists (p.11). Fund managers write up meeting notes, rating quality on a scale from 1 (poor) to 5 (excellent). The sample contains 3,423 fund manager meetings and 1,288 governance specialist meetings (p.11). Governance specialists may place companies on an internal “governance health warning” list where significant concerns arise.

Meetings and trading

Fund managers heavily trade portfolio firms on meeting days. The probability of trading on a normal day was 4.7%, rising to over three times higher at 15.7% on a meeting day (p.16). Governance specialist meetings are primarily associated with sell trades; fund manager meetings generate net buying. High-quality meetings (rated 4–5) are associated with aggregate buying, while low-quality meetings (rated 1–2) generate aggregate selling. High-level meetings — where at least two of the CEO, CFO, or Chair of the portfolio firm attend — are associated with consistently higher levels of buying (p.20).

Meeting induced recommendation changes

The sample contains 33,696 internal analyst recommendations: 14,724 “Buy”, 12,944 “Hold”, and 6,025 “Sell” (p.21). There are 1,556 unique downgrades (“Sell Signals”) and 1,433 upgrades (“Buy Signals”) (p.21–22). Of these, 106 sell signals and 141 buy signals are explicitly attributable to meetings in the notes (p.23). Meeting-triggered recommendation changes are associated with trading of comparable magnitude to non-meeting-based recommendation changes.

Trading performance around meetings

A long-short portfolio based on meeting-day trades outperforms the equivalent non-meeting portfolio by 180 bps over 20 trading days and by roughly 260 bps over 40 trading days (p.23). Monthly Sharpe ratios average 0.49 (p.5). Actual trades generate roughly 10 to 20 bps of abnormal returns per position under fast trading assumptions (p.5). Aggregate trading gains total roughly US$420 million (fast trading) and US$170 million (slow trading) (p.27). If non-trading funds traded like the average trading fund, gains would rise to US$1,270 million and US$470 million respectively (p.27–28), suggesting fund managers leave money on the table.

The content of meetings and trading

The tone of meeting notes is significantly related to trading activity. Positive-tone fund manager meetings are associated with much larger buying than negative-tone meetings (p.27). Specific topics appear less reliably linked to directional trading than overall tone, consistent with soft information being the primary driver of trading decisions rather than hard, quantitative information.

Conclusion

Meetings broadly serve to transmit soft information, with tone and attendee seniority more informative than specific topics discussed. Both belief-confirming and belief-changing meetings generate trading opportunities. The majority of fund positions do not trade around meetings, even highly informative ones, suggesting active managers leave economically significant amounts of money on the table through under-trading.